The FAA invites you to send any written relevant data, views, or arguments about this proposal. Send your comments to an address listed under ADDRESSES . Include “Docket No. FAA-2023-1987; Project Identifier MCAI-2023-00807-T” at the beginning of your comments. The most helpful comments reference a specific portion of the proposal, explain the reason for any recommended change, and include supporting data. The FAA will consider all comments received by the closing date and may amend this proposal because of those comments.

Except for Confidential Business Information (CBI) as described in the following paragraph, and other information as described in 14 CFR 11.35, the FAA will post all comments received, without change, to regulations.gov, including any personal information you provide. The agency will also post a report summarizing each substantive verbal contact received about this SNPRM.

CBI is commercial or financial information that is both customarily and actually treated as private by its owner. Under the Freedom of Information Act (FOIA) (5 U.S.C. 552), CBI is exempt from public disclosure. If your comments responsive to this SNPRM contain commercial or financial information that is customarily treated as private, that you actually treat as private, and that is relevant or responsive to this SNPRM, it is important that you clearly designate the submitted comments as CBI. Please mark each page of your submission containing CBI as “PROPIN.” The FAA will treat such marked submissions as confidential under the FOIA, and they will not be placed in the public docket of this SNPRM. Submissions containing CBI should be sent to Shahram Daneshmandi, Aviation Safety Engineer, FAA, 1600 Stewart Avenue, Suite 410, Westbury, NY 11590; telephone: 206-231-3220; email: shahram.daneshmandi@faa.gov. Any commentary that the FAA receives which is not specifically designated as CBI will be placed in the public docket for this rulemaking.

The FAA issued AD 2021-17-02, Amendment 39-21685 (86 FR 48490, August 31, 2021) (AD 2021-17-02), for all ATR—GIE Avions de Transport Régional Model ATR42-200, -300, and -320 airplanes. AD 2021-17-02 was prompted by an MCAI originated by EASA, which is the Technical Agent for the Member States of the European Union. EASA issued AD 2021-0024, dated January 19, 2021, to correct false activation of the stall warning system due to wiring damage on the wire bundle between an angle of attack (AOA) probe and the crew alerting computer. AD 2021-17-02 requires a one-time inspection for discrepancies of the wire bundles between the left- and right-hand AOA probes and the crew alerting computer, and, depending on findings, applicable corrective actions. AD 2021-17-02 also requires, for certain airplanes, modifying the captain stick shaker wiring, and for all airplanes, revising the existing aircraft flight manual (AFM) and applicable corresponding operational procedures to incorporate procedures for the stick pusher/shaker. AD 2021-17-02 superseded AD 2020-23-13, Amendment 39-21330 (85 FR 73407, November 18, 2020).

Since the FAA issued AD 2021-17-02, EASA superseded AD 2021-0024, dated January 19, 2021, and issued EASA AD 2023-0134, dated July 5, 2023 (EASA AD 2023-0134), to correct an unsafe condition for all ATR—GIE Avions de Transport Régional Model ATR42-200, -300, and -320 airplanes. EASA AD 2023-0134 states that final modification instructions of the affected wiring were developed.

The FAA issued an NPRM to amend 14 CFR part 39 by adding an AD to supersede AD 2021-17-02 that would apply to all ATR—GIE Avions de Transport Régional Model ATR42-200, -300, and -320 airplanes. The NPRM published in the Federal Register on October 5, 2023 (88 FR 69102). In the NPRM, the FAA proposed to retain all of the requirements of AD 2021-17-02. The NPRM also proposed to require installing a new AOA power supply unit and removing the AFM amendment.

Since the FAA issued the NPRM, EASA superseded EASA AD 2023-0134 and issued EASA AD 2023-0191, dated November 2, 2023 (EASA AD 2023-0191) (also referred to as the MCAI), to correct an unsafe condition for all ATR—GIE Avions de Transport Régional Model ATR42-200, -300, and -320 airplanes. The MCAI states that since EASA AD 2023-0134 was issued, the manufacturer published ATR Service Bulletin ATR42-31-0121, Revision 2, dated September 13, 2023, which specifies additional wiring modification for certain airplanes.

The FAA is proposing this AD to address false activation of the stall warning system, which could result in loss of control of the airplane during take-off and landing phases. You may examine the MCAI in the AD docket at regulations.gov under Docket No. FAA-2023-1987.

The FAA received one comment, from Air Line Pilots Association, International (ALPA), who supported the NPRM without change.

The FAA determined that the exception specified in paragraph (h)(4) of the proposed AD (in the NPRM) is not necessary because it would have required repairing all discrepancies before further flight, which is more restrictive than the MCAI. The FAA also determined that the exception specified in paragraph (h)(5) of the proposed AD (in the NPRM) is not necessary because it would have required amending “applicable corresponding operational procedures,” which the FAA has determined should not be an AD requirement. The FAA has removed both exceptions from this proposed AD and renumbered subsequent paragraphs accordingly.

Although this proposed AD does not explicitly restate the requirements of AD 2021-17-02, this proposed AD would retain all of the requirements of AD 2021-17-02. Those requirements are referenced in EASA AD 2023-0191, which, in turn, is referenced in paragraph (g) of this proposed AD.

EASA AD 2023-0191 specifies procedures for a one-time inspection for discrepancies of the wire bundles between the left- and right-hand AOA probes and the crew alerting computer, and, depending on findings, applicable corrective actions (repair). EASA AD 2023-0191 also specifies procedures, for certain airplanes, for modifying the captain stick shaker wiring, and for all airplanes, revising the existing AFM to incorporate procedures for the stick pusher/shaker. Finally, EASA AD 2023-0191 specifies procedures for installing the AOA power supply unit, removing the AFM amendment, and accomplishing additional modification of the affected wiring. This material is reasonably available because the interested parties have access to it through their normal course of business or by the means identified in ADDRESSES .

This product has been approved by the aviation authority of another country and is approved for operation in the United States. Pursuant to the FAA's bilateral agreement with this State of Design Authority, it has notified the FAA of the unsafe condition described in the MCAI referenced above. The FAA is issuing this SNPRM after determining that the unsafe condition described previously is likely to exist or develop in other products of the same type design.

Certain changes described above expand the scope of the NPRM. As a result, it is necessary to reopen the comment period to provide additional opportunity for the public to comment on this SNPRM.

This proposed AD would retain all of the requirements of AD 2021-17-02. This proposed AD would require accomplishing the actions specified in EASA AD 2023-0191 described previously, except for any differences identified as exceptions in the regulatory text of this proposed AD.

EASA AD 2023-0191 requires operators to amend the AFM to “inform all flight crews” of revisions to the AFM; and thereafter to “operate the aeroplane accordingly.” However, this proposed AD would not specifically require those actions as those actions are already required by FAA regulations. FAA regulations require that operators furnish to pilots any changes to the AFM (for example, 14 CFR 121.137), and to ensure the pilots are familiar with the AFM (for example, 14 CFR 91.505). As with any other flightcrew training requirement, training on the updated AFM content is tracked by the operators and recorded in each pilot's training record, which is available for the FAA to review. FAA regulations also require pilots to follow the procedures in the existing AFM including all updates. 14 CFR 91.9 requires that any person operating a civil aircraft must comply with the operating limitations specified in the AFM. Therefore, including a requirement in this proposed AD to operate the airplane according to the revised AFM would be redundant and unnecessary.

In the FAA's ongoing efforts to improve the efficiency of the AD process, the FAA developed a process to use some civil aviation authority (CAA) ADs as the primary source of information for compliance with requirements for corresponding FAA ADs. The FAA has been coordinating this process with manufacturers and CAAs. As a result, the FAA proposes to incorporate EASA AD 2023-0191 by reference in the FAA final rule. This proposed AD would, therefore, require compliance with EASA AD 2023-0191 in its entirety through that incorporation, except for any differences identified as exceptions in the regulatory text of this proposed AD. Using common terms that are the same as the heading of a particular section in EASA AD 2023-0191 does not mean that operators need comply only with that section. For example, where the AD requirement refers to “all required actions and compliance times,” compliance with this AD requirement is not limited to the section titled “Required Action(s) and Compliance Time(s)” in EASA AD 2023-0191. Service information required by EASA AD 2023-0191 for compliance will be available at regulations.gov under Docket No. FAA-2023-1987 after the FAA final rule is published.

The FAA estimates that this AD, if adopted as proposed, would affect 26 airplanes of U.S. registry. The FAA estimates the following costs to comply with this proposed AD:

The FAA has received no definitive data on which to base the cost estimates for the on-condition actions specified in this proposed AD.

Title 49 of the United States Code specifies the FAA's authority to issue rules on aviation safety. Subtitle I, section 106, describes the authority of the FAA Administrator. Subtitle VII: Aviation Programs, describes in more detail the scope of the Agency's authority.

The FAA is issuing this rulemaking under the authority described in Subtitle VII, Part A, Subpart III, Section 44701: General requirements. Under that section, Congress charges the FAA with promoting safe flight of civil aircraft in air commerce by prescribing regulations for practices, methods, and procedures the Administrator finds necessary for safety in air commerce. This regulation is within the scope of that authority because it addresses an unsafe condition that is likely to exist or develop on products identified in this rulemaking action.

The FAA determined that this proposed AD would not have federalism implications under Executive Order 13132. This proposed AD would not have a substantial direct effect on the States, on the relationship between the national Government and the States, or on the distribution of power and responsibilities among the various levels of government.

For the reasons discussed above, I certify this proposed regulation:

(1) Is not a “significant regulatory action” under Executive Order 12866,

(2) Would not affect intrastate aviation in Alaska, and

(3) Would not have a significant economic impact, positive or negative, on a substantial number of small entities under the criteria of the Regulatory Flexibility Act.

Accordingly, under the authority delegated to me by the Administrator, the FAA proposes to amend 14 CFR part 39 as follows:

I. Background A. Statutory Provisions

Enacted in 1970, the Currency and Foreign Transactions Reporting Act, generally referred to as the Bank Secrecy Act (“BSA”), is designed to combat money laundering, the financing of terrorism, and other illicit finance activity, and to safeguard the national security of the United States. 1 The Secretary of the Treasury (“the Secretary”) delegated the authority to implement, administer, and enforce the BSA and its implementing regulations to the Director of FinCEN. 2

Section 326 of the Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (“USA PATRIOT Act”) of 2001 added a subsection to the BSA, subsection ( l ) to 31 U.S.C. 5318, in order to facilitate the prevention, detection, and prosecution of international money laundering and the financing of terrorism. Subsection 31 U.S.C. 5318( l ) requires the Secretary to “prescribe regulations setting forth the minimum standards for financial institutions and their customers regarding the identity of the customer that shall apply in connection with the opening of an account at a financial institution.” The regulations implementing section 326 must, at a minimum, “require financial institutions to implement, and customers (after being given adequate notice) to comply with, reasonable procedures for—(A) verifying the identity of any person seeking to open an account to the extent reasonable and practicable; (B) maintaining records of the information used to verify the person's identity, including name, address, and other identifying information; and (C) consulting lists of known or suspected terrorists or terrorist organizations provided to the financial institution by any government agency to determine whether a person seeking to open an account appears on any such list.”  3 These programs are referred to as Customer Identification Programs (“CIPs”) and are long-standing, foundational components of a financial institution's anti-money laundering program.

As enacted, section 326 applies to all “financial institutions.” This term is defined broadly in the BSA to encompass a variety of entities, including commercial banks; agencies, and branches of foreign banks in the United States; thrift institutions, credit unions, and private bankers; trust companies; securities brokers and dealers registered with the Commission; investment companies; futures commission merchants; insurance companies; travel agencies; pawnbrokers; dealers in precious metals, stones, and jewels; check-cashers; certain casinos; and telegraph companies, among others. 4 The BSA also grants authority to the Secretary to define, by regulation, additional types of businesses as financial institutions where the Secretary determines that such businesses engage in any activity “similar to, related to, or a substitute for” those in which any of the businesses listed in the statutory definition are authorized to engage. 5 As part of the implementation, administration, and enforcement of the BSA, this authority has been delegated to the Director of FinCEN. 6

On February 15, 2024, the Secretary, through FinCEN, proposed to designate certain investment advisers as “financial institutions” under the BSA and subject them to anti-money laundering/countering the financing of terrorism (“AML/CFT”) program requirements and Suspicious Activity Report (“SAR”) filing obligations, as well as other BSA requirements (“AML/CFT Program and SAR Proposed Rule”). 7 Although the Investment Advisers Act of 1940 (“Advisers Act”) and the rules thereunder apply to a wide range of investment advisers, 8 the AML/CFT Program and SAR Proposed Rule—and the rule proposed in this joint NPRM as well—would only apply to a narrower subset of persons meeting the Advisers Act definition of “investment adviser”:  9 advisers registered or required to be registered with the SEC (referred to as “registered investment advisers,” or “RIAs”), as well as those exempt from registration under sections 203(l) or 203(m) of the Advisers Act and applicable rules thereunder (referred to as “exempt reporting advisers,” or “ERAs”).

In prescribing regulations for financial institutions implementing section 326, 31 U.S.C. 5318(l)(3) directs the Secretary to “take into consideration the various types of accounts maintained by various types of financial institutions, the various methods of opening accounts, and the various types of identifying information available.”  10 Further, 31 U.S.C. 5318(l)(4) requires that implementing regulations for certain types of financial institutions—which would include the set of investment advisers proposed to be added to the definition of “financial institution” through the AML/CFT Program and SAR Proposed Rule—be prescribed jointly with the appropriate Federal functional regulator (as defined in section 509 of the Gramm-Leach-Bliley Act). 11 The appropriate Federal functional regulator for investment advisers is the SEC. 12 Thus, FinCEN and the SEC are issuing this proposed rule jointly.

While investment advisers have not been previously subject to CIP requirements, in certain circumstances, some investment advisers already obtain and conduct verification of customer identity information. 13 For example, some investment advisers may implement CIP requirements if the entity is also a registered broker-dealer  14 or a bank ( i.e., a dual registrant), or is an operating subsidiary of a bank;  15 other investment advisers are affiliates of banks or broker-dealers, which may implement an enterprise-wide AML/CFT program that includes a CIP. In addition, some investment advisers have already implemented voluntary AML/CFT programs that may include CIP measures. 16

This proposed rule is generally consistent with existing rules requiring other financial institutions, such as brokers or dealers in securities, open-end investment companies (such as mutual funds), 17 credit unions, banks, and other financial institutions, to adopt and implement CIPs. 18 The similarity between this proposed rule and those rules reflects the importance that FinCEN and the SEC (“the Commission”) assign to the harmonization of CIP requirements, including for the purposes of increasing effectiveness and efficiency for investment advisers that are affiliated with other financial institutions, such as banks, broker-dealers, or open-end investment companies (such as mutual funds) that are already subject to CIP requirements. CIP requirements also support the application of other AML/CFT measures by making it more difficult for persons to use false identities to establish customer relationships with investment advisers for the purposes of laundering money, financing terrorism, or engaging in other illicit finance activity.

Under the proposed rule, the substantive requirements of the joint proposed rule would be codified with other BSA regulations as part of Treasury's proposed regulations in 31 CFR part 1032.

Section 1032.100(a) Account. The proposed rule would define “account” for the purposes of investment advisers' CIP obligations as any contractual or other business relationship between a person and an investment adviser under which the investment adviser provides investment advisory services. 20 The proposed definition excludes an account that an investment adviser acquires through an acquisition, merger, purchase of assets, or assumption of liabilities. Customers do not “open” such transferred accounts, and, therefore, these accounts do not fall within the scope of section 326. 21 Such accounts, however, may still be subject to other AML/CFT requirements applicable to advisory activities, including activities within the scope of the AML/CFT Program and SAR Proposed Rule, to the extent it is adopted. 22 Additionally, the definition of account would include accounts opened for the purpose of participating in an employee benefit plan established pursuant to the Employee Retirement Income Security Act of 1974 (“ERISA”). While ERISA accounts are excluded from the definition of “account” in the CIP rules applicable to mutual funds, 23 they are not being excluded here to harmonize the applicability of this proposed rule with the AML/CFT Program and SAR Proposed Rule, which would require RIAs and ERAs to apply AML/CFT program and SAR reporting requirements to all of their accounts, including accounts opened for the purpose of participating in an employee benefit plan established pursuant to ERISA.

Section 1032.100(b) Commission. The proposed rule would define “Commission” to mean the United States Securities and Exchange Commission.

Section 1032.100(c) Customer. The proposed rule would define “customer” for the purposes of investment advisers' CIP obligations as a person—including a natural person or a legal entity—who opens a new account with an investment adviser. This means the person identified as the accountholder, except in the case of an individual who lacks legal capacity, such as a minor, and non-legal entities, in which case the customer would be the individual who opens the new account for a minor or non-legal entity. Under this proposed rule, an investment adviser would not be required to look through a trust or similar account to its beneficiaries and would only be required to verify the identity of the named accountholder. 24

The proposed rule's definition of “customer” would not include individuals with authority or control over the accounts, if such persons are not the accountholders. In addition, the definition would not include persons who fill out the account opening paperwork or provide information necessary to set up an account but are not the accountholder. Instead, as described below, section 1032.220(a)(2)(ii)(C) of the proposed rule separately would require an investment adviser's CIP to address situations where, based on the investment adviser's risk assessment of a new account opened by a customer that is not an individual, the investment adviser will need to obtain information about individuals with authority or control over the account in order to verify the customer's identity.

The proposed definition of “customer” would also not include a financial institution regulated by a Federal functional regulator or a bank regulated by a State bank regulator; certain government entities; certain persons (other than banks) that are publicly listed on U.S. securities exchanges or certain subsidiaries of persons listed on U.S. securities exchanges;  25 or persons that have an existing account with the investment adviser, provided the investment adviser has a reasonable belief that it knows the true identity of the person. These exemptions are being included to be consistent with CIP requirements for other financial institutions. 26

Section 1032.100(d) Financial institution. The proposed rule includes a definition of “financial institution” that cross-references the BSA's definition of “financial institution” in 31 U.S.C. 5312(a)(2) and (c)(1), and its implementing regulations, which is currently codified at 31 CFR 1010.100(t). 27 The proposed rule includes this definition to avoid any ambiguity about the meaning of “financial institution” in proposed § 1032.220(a)(6). Proposed § 1032.220(a)(6) would allow investment advisers to rely on certain other financial institutions' performance of their CIP procedures under specific circumstances, as described below. Accordingly, and as described below, an investment adviser would be able to rely on such performance by other “financial institutions” as defined in 31 U.S.C. 5312(a)(2) and (c)(1) and its implementing regulations to fulfill those aspects of its CIP obligations.

Section 1032.100(e) Investment adviser. The proposed rule includes a definition of “investment adviser” that is the same as the proposed definition of investment adviser in the AML/CFT Program and SAR Proposed Rule. 28 In this way, both this proposed rule and the AML/CFT Program and SAR Proposed Rule would apply to the same group of persons. The proposed definition in the AML/CFT Program and SAR Proposed Rule—and thus the definition proposed in this NPRM—is “[a]ny person who is registered or required to register with the SEC under section 203 of the Advisers Act (15 U.S.C. 80b-3(a)), or any person that is exempt from SEC registration under section 203(l) or 203(m) of the Advisers Act (15 U.S.C. 80b-3(l), (m)).”  29 In other words, under this proposed definition, an investment adviser would be any RIA (those registered or required to register with the SEC) or ERA (those exempt from SEC registration under the listed provisions). 30 We anticipate that any change to the scope of the AML/CFT Program and SAR Proposed Rule, as finalized, would also be reflected in this rule, to ensure that the scope of both rules remain consistent.

Section 1032.220(a)(1) In general. Section 326 requires the Secretary and, where relevant, the appropriate Federal functional regulator (here, the SEC) to prescribe regulations requiring financial institutions to implement, and customers (after being given adequate notice) to comply with, “reasonable procedures” for verifying the identity of any person seeking to open an account, “to the extent reasonable and practicable”;  31 for maintaining records associated with such verification; and for consulting lists of known terrorists and terrorist organizations. 32 Proposed § 1032.220(a)(1) accordingly would require that each investment adviser establish, document, and maintain a written CIP as part of the AML/CFT program under 31 U.S.C. 5318(h). 33 This proposed requirement is intended to make clear that the CIP is not a separate program, but rather would be incorporated into an investment adviser's overall AML/CFT program. The proposed rule would require that the CIP be appropriate for its size and business that, at a minimum, includes each of the requirements of paragraphs (a)(1) through (a)(5) of proposed section 1032.220.

The investment adviser may deem these requirements satisfied for any mutual fund it advises if the mutual fund has developed and implemented a CIP that is compliant with CIP requirements applicable to mutual funds under the relevant provision of this subpart. FinCEN and the SEC believe that this exemption is appropriate because of the regulatory and practical relationship between mutual funds and their investment advisers. As a practical matter, we believe that any CIP requirement imposed on an RIA to a mutual fund is already addressed by the existing CIP requirements imposed on the mutual fund itself. 34 Consequently, we are proposing not to require investment advisers to mutual funds to include those mutual funds within the investment advisers' own CIP programs, as doing so would be redundant. This exemption is permissive and not mandatory; an investment adviser could decide to include the mutual funds it advises in complying with the investment adviser's CIP requirements.

Section 1032.220(a)(2) Identity verification procedures. Proposed § 1032.220(a)(2) would require that an investment adviser's CIP include risk-based procedures for verifying the identity of customers, to the extent reasonable and practicable, and that such verification occur within a reasonable time before or after the customer's account is opened. The inclusion of “before or after” account opening is intended to offer flexibility to an adviser in complying with the requirements of the proposed rule during the process of creating an advisory relationship with a customer. The procedures must enable the investment adviser to form a reasonable belief that it knows the identity of each customer.

A person becomes a customer each time the person opens a new account with an investment adviser. Therefore, upon the opening of each account, the verification requirements of this proposed rule would apply. However, if a customer whose identification has been verified previously opens a new account, the investment adviser would generally not need to verify the customer's identity again, provided the investment adviser (1) previously verified the customer's identity, to the extent required, in accordance with procedures consistent with the proposed rule, and (2) continues to have a reasonable belief that it knows the true identity of the customer based on the previous verification.

Under this proposed rule, the procedures must be based on the investment adviser's assessment of the relevant risks, including those presented by the various types of accounts maintained by the investment adviser; the various methods of opening accounts provided by the investment adviser, the various types of identifying information available and the investment adviser's size, location, and customer base. Other relevant risk factors could include, for example, the types of money laundering and terrorist financing activities present in the respective jurisdiction; whether account opening occurs in-person or online; the types of services and transactions offered or performed by the investment adviser; and the reliance on third-party firms (including other investment advisers, broker-dealers, or funds) for identity verification procedures.

Thus, in developing and updating CIPs, investment advisers would be required to consider the type of identifying information available for customers and the methods available to verify that information. While paragraph (a)(2)(i) of this proposed rule would require certain minimum identifying information to be obtained, and paragraph (a)(2)(ii) discusses certain suitable verification methods, as described below, investment advisers should consider on an ongoing basis whether other identifying information or verification methods are appropriate, particularly as they become available in the future.

Section 1032.220(a)(2)(i) Customer information required. Pursuant to the proposed rule, an investment adviser's CIP must require the investment adviser to obtain, at a minimum, certain identifying information with respect to each customer before or after an account is opened for the customer. Specifically, the investment adviser must obtain with respect to each customer: (1) name; (2) date of birth for an individual or the date of formation for any person other than an individual; (3) address;  35 and (4) identification number. 36 Under proposed § 1032.220(a)(2)(i), the term “name” would refer to a customer's full legal name, and the investment adviser should consider collecting any aliases or assumed names as well ( e.g., “doing business as” or “DBA” names). For persons other than an individual, the date of formation may be available on the certificate of formation or incorporation (or other document used to create a legal person), as well as any amendments to those documents.

Proposed § 1032.220(a)(2)(i)(A) would require only that this minimum identifying information be obtained. Investment advisers, in assessing the risk factors in paragraph (a)(2), however, would also be required to determine whether other identifying information is necessary to enable the investment adviser to form a reasonable belief that it knows the true identity of each customer. There also may be other circumstances that make it appropriate to obtain additional information. 37 For example, under proposed section 1032.220(a)(2)(ii)(C), an investment adviser must set forth guidelines in its CIP for situations where, based upon a risk-based assessment of a customer that is not an individual, additional information should be obtained about the individuals with authority or control over the customer's account. The CIP generally should include guidelines for collecting additional information in other situations where the investment adviser determines in the course of examining the nature of its business and operations that additional information should be obtained, consistent with a risk-based CIP, in order to enable the investment adviser to form a reasonable belief that it knows the true identity of the customer. Such guidelines generally should indicate the types of additional information needed and the circumstances when it would be obtained.

Proposed § 1032.220(a)(2)(i)(B) includes an exception from the requirement to obtain a taxpayer identification number from a customer opening a new account. As proposed, this exception would allow an investment adviser to open an account for a person that has applied for, but has not received, a TIN. In this case, the CIP would be required to include procedures to confirm that the application was filed before the person opened the account and to obtain the TIN within a reasonable period of time after the account is opened.

Moreover, under proposed § 1032.220(a)(2)(i), when opening an account for a non-U.S. person that is not an individual and that does not have an identification number, the investment adviser would be required to request alternative government-issued documentation certifying the existence of the customer. In contrast to the CIP requirements for mutual funds and broker dealers regarding a non-U.S. person that is not an individual, this specific requirement is being included here to account for changes in how financial institutions now routinely verify the identity of non-U.S. persons that are not individuals.

Section 1032.220(a)(2)(ii) Customer verification. Under proposed § 1032.220(a)(2)(ii), after obtaining identifying information with respect to a customer, the investment adviser would be required to follow risk-based procedures to verify the accuracy of that information in order to reach a point where it can form a reasonable belief that it knows the true identity of the customer. The proposed rule would require that verification procedures be undertaken within a reasonable time before or after a customer's account is opened. This flexibility would have to be exercised in a reasonable time, given that verifications too far in advance may become stale and verifications too long after the fact may provide opportunities to launder money or engage in other relevant illicit finance activity while verification is pending. The amount of time it will take an investment adviser to verify the identity of a customer may depend on the type of account opened, whether the customer opens the account in person, and the type of identifying information available. For example, an investment adviser may choose to place limits on the account, such as temporarily limiting advisory-related activities in an account until the customer's identity is verified in which case the adviser should inform the accountholder. Therefore, the proposed rule would provide investment advisers with the flexibility to use a risk-based approach to determine when the identity of a customer must be verified relative to the opening of an account.

Proposed § 1032.220(a)(2)(ii) would provide for two methods of verifying identifying information: verification through documents and verification through non-documentary means. This proposed provision would require that an investment adviser's CIP address both methods of verification. The CIP would have to set forth risk-based procedures describing when documents, non-documentary methods, or a combination of both will be used. These procedures should be based on the investment adviser's assessment of the factors described in paragraph (a)(2) of the proposed rule.

The risk that an investment adviser will not have a reasonable belief that it knows a customer's true identity will be heightened for certain types of accounts, such as accounts opened in the name of a corporation, partnership, or trust that is created, or conducts substantial business, in jurisdictions designated as primary money laundering concerns or designated as non-cooperative by an international body, or jurisdictions that are otherwise considered high-risk for money laundering or terrorist financing with respect to their compliance with relevant international standards. 38 Obtaining sufficient information to verify a given customer's identity can reduce the risk an investment adviser will be used as a conduit for money laundering and terrorist financing. An investment adviser's identity verification procedures must be based on its assessments of the factors in paragraph (a)(2). Accordingly, when those assessments suggest a heightened risk, the investment adviser should modify its verification measures accordingly ( e.g., by utilizing additional measures).

Section 1032.220(a)(2)(ii)(A) Customer verification through documents. Proposed § 1032.220(a)(2)(ii)(A) would require an investment adviser's CIP to contain procedures that set forth the documents that the investment adviser will use for verification, based on a risk-based analysis of the types of documents that it believes will enable it to verify customer identities. The proposed rule includes a list of identification documents, though an investment adviser would be allowed to use other documents, provided they allow the investment adviser to establish a reasonable belief that it knows the true identity of the customer. For individuals, these documents may include unexpired government-issued identification evidencing nationality or residence and bearing a photograph or similar safeguard. For other persons, suitable documents would include documents showing the existence of the entity, such as certified articles of incorporation, a government-issued business license, a partnership agreement, or a trust instrument. The investment adviser's procedures must take into account circumstances in which there may be problems authenticating documents and the inherent limitations of certain documents as a means of identity verification. 39 These limitations would affect the types of documents that would be necessary to establish a reasonable belief that the investment adviser knows the true identity of the customer and would require the use of non-documentary methods in addition to documents under some circumstances.

Under proposed § 1032.220(a)(2)(ii)(A), once an investment adviser obtains and verifies the identity of a customer through a suitable document, the investment adviser would not be required to take steps to determine whether a document has been validly issued. An investment adviser generally would be allowed to rely on an unexpired government-issued identification for verification purposes;  40 however, if a document has indicators of fraud, the investment adviser would have to consider that factor in determining whether it could form a reasonable belief that it knows the customer's true identity. 41

Section 1032.220(a)(2)(ii)(B) Customer verification through non-documentary methods. Proposed § 1032.220(a)(2)(ii)(B) would require an investment adviser's CIP to describe non-documentary verification methods and when such methods will be employed in addition to, or instead of, verification through documents. The proposed rule would permit the exclusive use of non-documentary methods because some accounts may be opened by telephone, mail, or over the internet in ways that may make sole reliance on documentary verification difficult or burdensome. 42 However, even if the customer presents identification documents, it may be appropriate to use non-documentary methods as well. Under this provision, the investment adviser would be ultimately responsible for employing verification methods that enable the adviser to form a reasonable belief that it knows the true identity of the customer.

Proposed § 1032.220(a)(2)(ii)(B) would set forth certain non-documentary methods that would be suitable for verifying identity. These methods may include contacting a customer; obtaining a financial statement; comparing the identifying information obtained with respect to the customer against relevant fraud, bad check databases to determine whether any of the information is associated with known incidents of fraudulent behavior; comparing the identifying information with information available from a trusted third-party source, such as a credit report from a consumer reporting agency or an account verification database; and checking references with other financial institutions. This list is not intended to exhaust all methods that may be suitable, however. For example, the investment adviser also may wish to analyze whether there is logical consistency between the identifying information provided, such as the customer's name, street address, ZIP code, telephone number (if provided), date of birth, and social security number.

Proposed § 1032.220(a)(2)(ii)(B) also would require an investment adviser's CIP to address situations in which (1) an individual is unable to present an unexpired government-issued identification document that bears a photograph or similar safeguard; (2) the investment adviser is not familiar with the types of documents presented; (3) the investment adviser does not obtain documents to verify the identity of the customer; (4) the investment adviser does not meet face-to-face with a customer who is a natural person; and (5) the investment adviser is otherwise presented with circumstances that increase the risk the investment adviser will be unable to form a reasonable belief that it knows the true identity of a customer through documents.

FinCEN and the SEC recognize that identification documents, including those issued by a government entity, may be obtained illegally and may be fraudulent. In light of the recent increase in identity theft, investment advisers would be encouraged to use non-documentary methods as well, even when an investment adviser has received identification documents from the customer. Additionally, investment advisers are encouraged to consider using both documentary and non-documentary verification methods and should consider on a regular basis whether their procedures for identity verification are appropriate.

Section 1032.220(a)(2)(ii)(C) Additional verification for certain customers. Proposed § 1032.220(a)(2)(ii)(C) would require that an investment adviser's CIP address circumstances in which, based on the investment adviser's risk assessment of a new account opened by a customer that is not an individual, the investment adviser will obtain information about individuals with authority or control over such accounts in order to verify the customer's identity. This requirement would apply only when the investment adviser cannot verify the true identity of a customer that is not an individual using the verification methods described in paragraphs (a)(2)(ii)(A) and (B) of the proposed rule. 43

While investment advisers may be able to verify the majority of customers adequately through the documentary or non-documentary verification methods described above, there may be circumstances when the investment adviser cannot form a reasonable belief that it knows the true identity of a customer using such methods. The risk that the investment adviser will not know the customer's true identity may be heightened for certain types of accounts, such as an account opened in the name of a corporation, partnership, or trust that is created or conducts substantial business in a jurisdiction that has been designated by the United States as a primary money laundering concern or by an international body as non-cooperative, or jurisdictions that are otherwise considered high-risk for money laundering or terrorist financing with respect to their compliance with relevant international standards. As a result of this, FinCEN and the SEC are proposing to require (1) that an investment adviser identify customers that are not individuals that pose a heightened risk of not being properly identified and (2) that an investment adviser's CIP prescribe additional measures that may be used to obtain information about individuals with authority or control over the account to verify the customer's identity when standard documentary or non-documentary methods prove to be insufficient.

Section 1032.220(a)(2)(iii) Lack of verification. Proposed § 1032.220(a)(2)(iii) would require that an investment adviser's CIP include procedures for responding to circumstances in which the investment adviser cannot form a reasonable belief that it knows the true identity of a customer. These procedures should describe (1) when the investment adviser should not open an account, (2) the terms under which the investment adviser may provide advisory services to the customer while the investment adviser attempts to verify the customer's identity, (3) when the investment adviser should close an account after attempts to verify a customer's identity fail, and (4) when the investment adviser should file a SAR in accordance with applicable law and regulation. 44

Section 1032.220(a)(3) Recordkeeping. Proposed § 1032.220(a)(3) would require that an investment adviser's CIP include procedures for making and maintaining a record of information obtained under procedures implementing proposed paragraph (a), as discussed in greater detail in the following paragraphs. This proposal is consistent with the requirement of 31 U.S.C. 5318(l)(2)(B) that CIPs include procedures for maintaining records of the information used to verify a person's identity, including name, address, and other identifying information.

Section 1032.220(a)(3)(i) Required records. Proposed § 1032.220(a)(3)(i) would require that an investment adviser's CIP include procedures for making and maintaining records related to verifying customer identity, as well as procedures for how to do so. Records would have to include the identifying information about each customer under proposed (a)(2)(i) and a description of any document that the investment adviser relied on to verify the identity of the customer (noting the document type, any identification number contained therein, the place of issuance, and the date of issuance and expiration as applicable and relevant) under proposed (a)(3)(i)(B). Proposed § 1032.220(a)(3)(i)(C) would require records to include a description of the methods and results of any measures undertaken to verify the identity of the customer. This description would include any relevant non-documentary methods and additional verification for certain customers used to verify identity under proposed § 1032.220(a)(2)(ii)(B) and (C). Finally, proposed § 1032.220(a)(3)(i)(D) would require investment advisers to record a description of the resolution of each substantive discrepancy discovered when verifying the identifying information obtained.

An investment adviser would be allowed to use electronic records to satisfy the requirements of this proposed rule.

Section 1032.220(a)(3)(ii) Record retention. Proposed § 1032.200(a)(3)(ii) would prescribe a bifurcated record retention schedule that is consistent with a general five-year retention requirement. Under this proposed provision, an investment adviser would be required to retain the information obtained about a customer pursuant to proposed paragraph (a)(3)(i)(A) ( i.e., identifying information about the customer) while the account remains open and for five years after the date the account is closed. 45 The remaining records required under proposed paragraphs (a)(3)(i)(B), (C), and (D) ( i.e., information regarding the verification of a customer's identity), however, would only have to be retained for five years after the record is made.

Section 1032.220(a)(4) Comparison with Government Lists. Under 31 U.S.C. 5318(l)(2)(C), a CIP must include reasonable procedures for determining whether a customer appears on any list of known or suspected terrorists or terrorist organizations provided by any government agency. Proposed § 1032.220(a)(4) accordingly would require that an investment adviser's CIP include reasonable procedures for determining whether a customer appears on any such list provided by any Federal Government agency that is designated as such by Treasury in consultation with the Federal functional regulators, and that an investment adviser make such a determination within a reasonable period of time after the account is opened, or earlier if required by another Federal law, regulation, or directive issued in connection with the applicable list. This requirement would apply only with respect to lists circulated, directly provided, or otherwise made available by the Federal government and designated as such by Treasury in consultation with the Federal functional regulators. In addition, proposed § 1032.220(a)(4) would state that the procedures must require investment advisers to follow all Federal directives issued in connection with such lists. Because Treasury and the Federal functional regulators have not yet designated any such lists for the purposes of CIP, the proposed rule cannot be more specific with respect to the lists that investment advisers must check. However, investment advisers would not have an affirmative duty under this rule to seek out all lists of known or suspected terrorists or terrorist organizations compiled by the Federal government. Instead, investment advisers would receive separate notification regarding the lists that they must consult for purposes of this provision.

Many investment advisers already have procedures for determining whether customers' names appear on some federal government lists, including lists that identify known terrorists and terrorist organizations. For example, under current law, there are substantive legal requirements associated with the lists circulated by Treasury's Office of Foreign Assets Control (“OFAC”). Failure to comply with these requirements may result in criminal or civil penalties.

Section 1032.220(a)(5) Customer Notice. Section 5318(l)(2) also provides that financial institutions must give their customers adequate notice of their identity verification procedures. Therefore, proposed § 1032.220(a)(5) would require that an investment adviser's CIP include procedures for providing customers with adequate notice that the firm is requesting information to verify their identities. The proposed rule would state that this notice is adequate if the investment adviser generally describes the identification requirements of the proposed rule and provides such notice in a manner reasonably designed to ensure that a prospective customer is able to view the notice, or is otherwise given notice, before opening an account. Under proposed § 1032.220(a)(5), depending on how an account is opened, an investment adviser could post a notice on its website, include the notice in its account applications, or use any other form of written or oral notice. 46 The sample notice included in the proposed rule, if appropriate, would be deemed adequate notice to an investment adviser's customers when provided in accordance with the other requirements described in this section.

Section 1032.220(a)(6) Reliance on another financial institution. There may be circumstances in which an investment adviser could rely on the performance by another financial institution of some or all of the elements of the investment adviser's CIP. However, the investment adviser would remain responsible for ensuring compliance with the proposed rule 1032.220(a)(6), and therefore would be required to actively monitor the operation of its CIP and assess its effectiveness. Proposed § 1032.220(a)(6) would provide that an investment adviser's CIP may include procedures that specify when the investment adviser will rely on the performance by another financial institution (including an affiliate) of any procedures of the investment adviser's CIP, and thereby satisfy the investment adviser's obligations under the proposed rule. Under proposed § 1032.220(a)(6), reliance would be permitted if a customer of the investment adviser is opening an account or has opened or has established an account or similar business relationship with the other financial institution to provide or engage in services, dealings, or other financial transactions, provided that: (1) such reliance is reasonable under the circumstances, (2) the other financial institution is subject to a rule implementing the AML/CFT compliance program requirements of 31 U.S.C. 5318(h) and is regulated by a Federal functional regulator, and (3) the other financial institution enters into a contract with the investment adviser requiring it to certify annually to the investment adviser that it has implemented an AML/CFT program and will perform (or its agent will perform) the specified requirements of the investment adviser's CIP. This last element could be satisfied by a reliance letter or other similar documentation. The investment adviser would not be held responsible for the failure of the other financial institution to fulfill adequately the adviser's CIP responsibilities, provided that the investment adviser can establish that its reliance was reasonable and that it has obtained the requisite contracts and certifications. The SEC and FinCEN emphasize that the investment adviser and the other financial institution upon which it relies would have to satisfy all of the conditions set forth in this proposed rule. If they do not, then the investment adviser would remain solely responsible for applying its own CIP to each customer in accordance with this rule. 47

Section 1032.220(b) Exemptions. Proposed § 1032.220(b) would provide that the SEC, with the concurrence of the Secretary, may by order or regulation exempt any investment adviser or any type of account from the requirements of this section. Proposed § 1032.220(b) would also provide that the Secretary, with the concurrence of the Commission, may exempt any investment adviser or any type of account from the requirements of this section. In issuing such exemptions, the SEC and the Secretary would have to consider whether the exemption is consistent with the purposes of the BSA and in the public interest, and they may consider other necessary and appropriate factors.

Section 1032.220(c) Effective Date. FinCEN and SEC anticipate that the effective date of the proposed rule will be 60 days after the date on which the final rule is published in the Federal Register . In order to provide time for investment advisers to come into compliance, section 1032.220(c) states the compliance date by which an investment adviser would be required to comply with this section. Specifically, under this proposed rule, an investment adviser would be required to develop and implement a CIP that complies with the requirements of this section on or before six months from the effective date of the regulation, but no sooner than the compliance date of the AML/CFT Program and SAR Proposed Rule, if adopted. We believe that six months strikes an appropriate balance between providing advisers with sufficient time to develop and implement a CIP while not overly delaying CIP implementation across the investment adviser industry.

Section 1032.220(d) Other requirements unaffected. The proposed rule would include a provision, proposed § 1032.220(d), parallel to that in CIP rules previously adopted for other financial institutions, stating that nothing in the rule shall be construed to relieve an investment adviser of its obligations to comply with any other provision of this chapter, including provisions concerning information that must be obtained, verified, or maintained in connection with any account or transaction. 48

FinCEN and the SEC invite comment on all aspects of the proposed regulation, and specifically seek comment on the following issues:

1. Whether the proposed definition of “account” is appropriate and unambiguous, and whether other examples of accounts should be added to the rule text.

a. Should an account opened for the purpose of participating in an employee benefit plan established under ERISA be excluded from the CIP account definition?

b. Are there types of accounts that should be exempted from CIP obligations?

2. The proposed definition of “account” would exclude an account that an investment adviser acquires through an acquisition, merger, purchase of assets, or assumption of liabilities, given that customers do not “open” transferred accounts, and, therefore, the accounts do not fall within the scope of section 326. As discussed above, advisers may be required to apply other sanctions and export compliance and AML/CFT requirements to those accounts. Are there circumstances in which advisers should be required to fulfill identity verification requirements for some transfers?

a. Should the rule require advisers to re-verify a customer's identity after a certain period of time ( e.g., every year, every other year, or every five years)?

3. Should the definition of “account” refer to the activities enumerated in 15 U.S.C. 80b-2(a)(11) for the definition of investment adviser? Or is the reference to “investment advisory services” sufficient?

4. Is the proposed definition of “customer” appropriate? Should other examples of customers be added to the rule text?

5. Should the definition of investment adviser apply to non-U.S. advisers registered or required to register with the SEC (for RIAs) or that report to the SEC on Form ADV (for ERAs), as proposed? What would be the logistical challenges of this approach?

6. Should terms defined elsewhere within 31 CFR chapter X, such as “U.S. Person”, “Non-U.S. Person”, and “Taxpayer Identification Number” be defined in the proposed rule as well or are those terms well-understood for CIP purposes?

7. To what extent do RIAs and ERAs already require customer identification and verification or otherwise have procedures in the manner proposed in the course of regular business or under other, existing regulatory obligations?

a. To what extent do the customer identification and verification procedures currently implemented by RIAs and ERAs resemble or differ from those required by the proposed rule?

8. Are there other categories of entities that, like mutual funds, should be exempted from an investment adviser's CIP program. Why or why not?

9. Should the exemption for mutual funds be dependent on the nature of the relationship between the investment adviser and its mutual fund customer and the ability of the investment adviser to meet CIP obligations?

10. Should closed-end registered funds, wrap fee programs, or other types of accounts advised by investment advisers be, on a risk-basis, exempted from an investment adviser's CIP program?

11. FinCEN also requests comment on the money laundering, terrorist financing, and other illicit finance risks faced by closed-end funds, and how entities with existing CIP requirements, such as banks and broker-dealers, apply those requirements to activity involving closed-end funds.

12. How would an investment adviser apply the identification and verification requirements at proposed § 1032.220(a)(2) to a private fund customer? What type of information would the adviser use to ask identification questions? We expect that advisers would likely already have this information in respect of private funds that they manage. Do commenters agree?

13. Proposed § 1032.220(a)(2) would require that an investment adviser verify customer identity within a reasonable time before or after the customer's account is opened. To what extent would an investment adviser provide advisory services prior to verifying customer identity? How much time would an investment adviser reasonably need to verify customer identity ( e.g., 30 days)?

14. How do investment advisers currently collect identity information for non-U.S. customers that are not individuals, such as foreign legal entities or other legal persons and legal arrangements?

15. Are the provisions in section 1032.220(a)(6) sufficient to permit an adviser to rely on another financial institution to perform its CIP requirements? Would there be any challenges for advisers with the proposed approach? Do commenters agree that an investment adviser should be required to actively monitor the operation of its CIP and assess its effectiveness in order to rely on another financial institution, or should the adviser not be held responsible by showing it reasonably relied on another financial institution that satisfied all of the conditions set forth in this proposed rule?

16. Is the proposed requirement for the other financial institution to enter into a contract with the investment adviser feasible? Does it depend on the size of the investment adviser and its negotiating power? Should we modify this requirement? For example, should we remove or modify the requirement for the other financial institution to certify that it will perform specified requirements of the investment adviser's CIP?

17. Does the proposed compliance date (six months after the final rule is issued) give advisers sufficient time to comply with the requirements of the proposed rule? Should the compliance date be staggered based on adviser size?

18. If an investment adviser cannot form a reasonable belief that it knows the true identity of a customer, should the investment adviser be able to engage in advisory activities on behalf of the customer prior to verifying the customer's identity?

FinCEN  49 and the SEC are sensitive to the economic effects that could result from the proposed rule and have accordingly considered certain likely effects and reasonable alternatives. Section 326 of the USA PATRIOT Act requires Treasury to prescribe regulations setting forth minimum standards for financial institutions regarding the identities of customers when they open an account. It also provides that the regulations issued by Treasury and the SEC must, at a minimum, require financial institutions to implement reasonable procedures for: (1) verification of the identity of any person seeking to open an account, to the extent reasonable and practicable; (2) maintenance of the information used to verify the person's identity, including name, address, and other identifying information; and (3) consulting lists of known or suspected terrorists or terrorist organizations provided to the financial institution by any government agency to determine whether a person seeking to open an account appears on any such list. 50

Under the BSA, FinCEN recently published the AML/CFT Program and SAR Proposed Rule, which would include certain investment advisers in the definition of financial institutions. 51 If that rule is adopted and “investment adviser” is thereby added to FinCEN's definition of “financial institution” at 31 CFR 1010.100(t), covered investment advisers would be financial institutions for purposes of section 326. As a consequence, FinCEN and the SEC would be required to jointly prescribe rules that establish minimum standards for covered investment advisers regarding the identities of customers when they open an account, which are proposed in this release. 52 This proposed rule is designed to align the requirements for investment advisers with existing rules for other financial institutions, such as broker-dealers, mutual funds, credit unions, banks, and others, to adopt and implement CIPs.

Evaluating the effectiveness of AML/CFT regimes is difficult because there is no precise method to determine either the actual number or magnitude of money laundering and terrorism financing crimes that occur, since some of these crimes go undetected. In addition, it is impossible to infer either the number or magnitude of such crimes that would have occurred absent the regime or under some alternative enforcement regime. To our knowledge, there are no academic studies that specifically assess the efficacy of CIP provisions as a part of AML/CFT regimes. However, there is some empirical evidence that points toward the effectiveness of the U.S. AML/CFT regime more broadly. 53

The scale of money laundering in the United States is large. Specifically, in fiscal year 2022, offenders in 1,001 money laundering cases were sentenced in the Federal system according to the United States Sentencing Commission (“USSC”). 54 These cases involved a median loss of approximately $300,000 and approximately 17.3 percent of these cases involved a loss of greater than $1.5 million. 55 USSC does not provide data that would allow us to determine what percentage of these offenses involved investment advisers. However, a Treasury-led review of SARs filed between 2013 and 2021 found that approximately 15.4 percent of RIAs and ERAs were associated with or referenced in at least one SAR ( i.e., they were identified either as a subject or in the narrative section of the SAR) during this time. 56 Further, the number of SAR filings where an RIA or ERA was referenced increased by approximately 400 percent between 2013 and 2021—a disproportionately higher increase than the overall increase in SAR filings during that time, which was approximately 140 percent. 57

According to Treasury, in its 2024 National Money Laundering Risk Assessment (“NMLRA”), “[m]oney laundering enables criminal activity and is necessary to disguise ill-gotten gains. It facilitates crime, distorts markets, and has a devastating economic and social impact on citizens. It also threatens U.S. national security as money laundering allows drug traffickers, fraudsters, human trafficking organizations, and corrupt officials, to operate and expand their criminal enterprises.”  58 Money laundering distorts markets because the incentives for criminals' use of the financial system differ from those of the broader market. Illicit funds also have a probability of seizure that could negatively impact the broader market, as it could increase the rate of return investors demand as compensation for risk and thus firms' cost of capital. 59

Money laundering also provides the appearance of legitimacy to proceeds of international corruption. By requiring that investment advisers verify the identity of their customers, the proposed rule would make it more difficult for money launderers to use investment advisers as an entry point into the U.S. financial system, reducing money launderers' ability to launder the proceeds of these criminal enterprises and thereby decreasing incentives to engage in these crimes. It would also help address the illicit finance risks identified in NMLRA. 60 As a result, the proposed rule would reduce both monetary as well as nonmonetary costs associated with money laundering involving investment advisers. 61

Given the overall scale of money laundering in the United States, preventing cases involving investment advisers could have substantial benefits. The NMLRA, has identified several vulnerabilities facing investment advisers and highlights some cases involving investment advisers. 62 The NMLRA cites ERAs, RIAs that are not dually registered as or affiliated with a bank or broker-dealer, and investment advisers managing private funds as the highest-risk types of investment advisers. 63

AML/CFT regimes can lower the amount of money laundering that occurs by creating barriers to these transactions. Economic theory would suggest that as more entities and transactions are subject to an AML/CFT regime, the deterrent effect of any particular regulation will increase: Illicit dollars attempting to access U.S. financial markets will seek entry via methods that are outside of, or at the weakest point of, an AML/CFT regime. As the number of possible entryways shrinks, these illicit dollars would be funneled into fewer and fewer channels. As the difficulty of laundering illicit dollars thus increases, the marginal cost of using these channels increases at an accelerated rate, further deterring their use. In targeting money laundering involving customers of investment advisers, the proposed rule thus seeks to fill a current gap in the U.S. AML/CFT regime, as recognized by the Financial Action Task Force (“FATF”). 64

The baseline against which the costs, benefits, and the effects on efficiency, competition, and capital formation of the proposed rule are measured consists of the current U.S. AML/CFT statutory framework, its regulatory implementation, and current AML/CFT practices of investment advisers and their related parties.

The AML statutory framework in the United States is commonly known as the BSA. 65 Under this framework, many types of financial institutions currently are required to enact AML programs that include a CIP. The SEC has jointly enacted rules with FinCEN that specifically impose CIP requirements on broker-dealers and mutual funds. 66

While investment advisers are not currently defined as financial institutions under the BSA and are not subject to CIP requirements, certain investment advisers already perform AML/CFT functions, including those associated with a CIP, as a result of existing requirements. 67 Specifically, some RIAs and ERAs may perform certain AML/CFT functions, including those associated with a CIP, if the entity is also a registered broker-dealer or a bank ( i.e., a dual registrant), or is an operating subsidiary of a bank; other investment advisers are affiliates of banks or broker-dealers, which may implement an enterprise-wide CIP-compliant AML/CFT program that would include investment advisers. Some investment advisers perform these functions via contract with a broker-dealer ( e.g., if the investment adviser performs CIP functions for joint customers) or other financial institutions.

In addition, certain investment advisers already obtain identifying information with respect to some accounts or customers. For example, U.S. investment advisers, like all U.S. persons, must comply with OFAC sanctions and U.S. export controls, so they are prohibited from engaging in transactions that violate foreign economic and trade sanctions and export controls imposed by the U.S. government and may engage in due diligence to ensure that they remain compliant with such sanctions and export controls. 68 As another example, advisers may be subject to non-U.S. AML and CIP laws, such as those applicable to private funds organized in the Cayman Islands. 69

Since the USA PATRIOT Act was passed, multiple rules have been proposed that would have required some investment advisers to apply AML/CFT requirements. While the substantive requirements contained in these proposals are not part of the baseline for the present rulemaking, some investment advisers have developed AML/CFT measures consistent with these prior proposals, as discussed in the next section. Specifically, on September 26, 2002, FinCEN published an NPRM proposing to require that unregistered investment companies, to include private funds, establish AML programs. 70 This was followed by the May 5, 2003, NPRM proposing to require certain investment advisers to establish AML programs. 71 On September 1, 2015, FinCEN published an NPRM “to prescribe minimum standards for . . . [AML] programs to be established by certain investment advisers and to require such investment advisers to report suspicious activity to FinCEN pursuant to the . . . BSA” (“Second Proposed Investment Adviser Rule”). 72 This proposed rule would have included RIAs within the definition of “financial institution” under the BSA and required them to maintain AML programs, report suspicious activity, and comply with other travel and recordkeeping requirements, but would not have included ERAs in the scope of the rule nor would it have established minimum CIP requirements.

Some financial institutions are required to establish a CIP that would include procedures for determining whether a customer appears on lists of known or suspected terrorists or terrorist organizations issued by any Federal government agency and designated as such by Treasury in consultation with the Federal functional regulators. 73 While no such lists have been designated by Treasury for any financial institution, our understanding is that some financial institutions, including some investment advisers, already check their customers against OFAC's Specially Designated Nationals and Blocked Persons List (“SDN List”). 74

While not legally required, some investment advisers currently have voluntary AML/CFT programs, which may be CIP-compliant. 75 Investment advisers also collect identifying information to perform operational tasks such as distinguishing between customer accounts, or contacting their customers for the purposes of sending administrative, regulatory, or other notices.

The 2016 Investment Management Compliance Testing Survey (“2016 IMCTS Survey”) collected information from approximately 700 RIAs on their existing implementation of AML/CFT measures. 76 According to this survey, as of 2016, approximately 40 percent of RIAs had already adopted AML/CFT policies consistent with the Second Proposed Investment Adviser Rule. An additional 36 percent of RIAs adopted some AML/CFT policies and procedures, but those were generally not in line with the Second Proposed Investment Adviser Rule. Therefore, according to the 2016 IMCTS Survey, approximately 76 percent of RIAs have at least some AML/CFT measures in place. In particular, 49 percent had annual employee AML/CFT training, 24 percent had a designated AML/CFT compliance officer, and 40 percent performed independent testing of their AML/CFT program annually. Similar information was not available for ERAs. While this survey did not ask a question about CIPs specifically, it is possible that some advisers did have a CIP as part of their AML/CFT policies and procedures.

Some investment advisers currently outsource some or all of the work needed for investment advisers or other parties to comply with regulatory requirements. 77 A variety of third-party firms ( e.g., fund administrators) exist that assist investment advisers in complying with their regulatory responsibilities and contractual obligations.

As of October 5, 2023, there were 14,914 RIAs, with roughly $114 trillion assets under management and 931,000 employees. 78 There were also 5,546 ERAs with additional gross assets of $5.2 trillion (ERAs do not report the number of employees). 79 RIAs had approximately 51.5 million natural person customers and 2.9 million legal entity customers. 80

The provisions added to the BSA from section 326 of the USA PATRIOT Act facilitate the prevention, detection, and prosecution of money laundering and the financing of terrorism. Section 326 requires financial institutions to establish CIP programs. If the AML/CFT Program and SAR Proposed Rule is adopted, investment advisers will be financial institutions under the BSA and in such event the BSA would require specifying how an investment adviser is to establish and execute a CIP program.

Obtaining and verifying the identity of account holders or responding to circumstances in which the investment adviser cannot form a reasonable belief that it knows the true identity of a customer would reduce the risk of terrorists and other criminals accessing U.S. financial markets to launder money, finance terrorism, or move funds for other illicit purposes. Comparing customer identities to those on government lists of known or suspected terrorists or terrorist organizations would assist investment advisers in identifying and preventing criminal activity. 81 Maintaining records would enhance investment advisers' internal compliance efforts and aid investment advisers in detecting and taking measures to prevent potential illegal activity and in identifying customers who have newly been added to such government lists. For example, in the event that an investment adviser's customer is flagged by screening software, maintaining records as required by the proposed rule would assist investment advisers in determining whether this flag was a false positive or whether the customer was truly added to a relevant government list. Establishing a CIP would help investment advisers systematize, and in some cases automate, practices that would facilitate detection of attempted financial crimes and would help ensure that investment advisers have practices that are as effective as possible at deterring financial crimes.

In circumstances where investment advisers are or could be performing CIP activities for certain entities that already have CIP obligations, the obliged entities (such as banks and broker-dealers) may not necessarily have a direct relationship with the customer. In such cases, investment advisers may be able to more efficiently perform CIP obligations such as collecting the required information from these customers because they have a more direct relationship with these customers. The proposed rule would aim to harmonize investment adviser CIP obligations with those of other obliged entities, which could enhance the benefits to the public and reduce the total costs imposed on the industry of these CIP obligations since investment advisers and other obliged financial institutions can decide by contract which party is most efficiently able to execute the CIP and the current disparity in CIP requirements may be distorting these negotiations. To the extent that investment advisers already have practices consistent with the requirements of the proposed rule either because of these extant obligations or for operational efficiency, 82 the benefits of the proposed rule described above would be mitigated.

The proposed rule would only require investment advisers to collect and verify the identity of customers that directly open and hold accounts (as defined in the proposed rule) with the adviser. The proposed rule's benefits would thus only apply in cases of money laundering activity involving those customers and not other individuals or entities. For example, an investment adviser may have a private fund as a customer. In this case, the proposed rule would require that the investment adviser collect the identifying information of the private fund and, in some cases, individuals with authority or control over such private fund, 83 but not that of those invested in such fund. In certain contexts, an investment adviser may itself be the individual with authority or control over the private fund.

Similarly, the benefits of the proposed rule would also be lessened to the extent that an investment adviser's customer holds accounts for purposes other than accessing financial markets (for example, if the customer holds an account only to receive investment research services). 84 In such cases, the benefits associated with protecting financial markets would not directly apply, although the other benefits discussed above would apply. 85

It is difficult to estimate how much economic loss the requirements would prevent. Neither the SEC nor FinCEN has data that would allow the quantification of how much money laundering would be reduced as a result of the proposed rule, or how much other illegal activity would be curbed by this reduction in money laundering. 86 Money laundering and other illicit financing is related to human trafficking, drug trafficking, terrorism, public corruption, the proliferation of weapons of mass destruction, fraud, and other crimes and illicit activities that cause substantial monetary and nonmonetary damages. 87 By reducing money laundering, and by extension its associated crimes, the proposed rule would reduce those harms to the extent that investment advisers are being used to facilitate such unlawful activity.

While certain provisions of the proposed rule specify minimum requirements, such as the pieces of information required to be obtained and verified, many aspects of the proposed rule require an investment adviser to establish and implement its CIP according to its specific circumstances. For example, under the proposed rule, the CIP must be based on factors specific to each investment adviser, such as size, customer base, and location. Thus, the analysis and detail necessary for a CIP would depend on the complexity of the investment adviser and its operations. Highly complex firms have more risk factors to consider, given, for example, their number of offices, variety of services and products offered, and range of customers. However, many of these firms already have some AML/CFT procedures in place and investment advisers already collect some identifying information that they would be required to collect under the proposed rule. 88

Generally, these requirements are similar to those for other financial institutions with which investment advisers engage. Many advisers may already bear the cost of these similar CIP requirements for other financial institutions in certain lines of business in ways that would reduce the costs of complying with the proposed rules, and in such cases the proposed rules would create minimal additional costs. Some RIAs and ERAs may have reduced costs because they may already perform certain AML/CFT functions, including those associated with a CIP, because they are dual registrants or affiliated with a bank or broker-dealer.

Under the proposed rule, RIAs that are dual registrants or affiliated advisers would not be legally required to establish a separate CIP for their advisory activities, provided that an existing comprehensive CIP-compliant AML/CFT program covers all the entity's legal and regulatory obligations under the proposed rule. RIAs would also be exempt from having to apply most of the proposed requirements with respect to the mutual funds they advise, as mutual funds have their own CIP requirements and are otherwise required to comply with the other reporting and recordkeeping requirements included in the proposed rule. Certain RIAs and ERAs may also already collect and verify certain information provided by customers via contract for a joint customer with another financial institution or through a voluntary AML/CFT program. 89

Some investment advisers may have similarly reduced costs even if they do not currently directly perform CIP-related AML functions. In particular, investment advisers that use broker-dealers on behalf of their customers but that do not perform the procedures required by the broker-dealer's CIP may currently already bear some or all of the proposed rule's costs indirectly. Specifically, these investment advisers could bear such costs in the form of higher charges for the broker-dealer's services, since these broker-dealers are already required to comply with similar CIP requirements related to their joint customers. In such cases, investment advisers would face new costs associated with the proposed rule, but these may be offset at least in part by reduced costs for broker-dealer services.

The proposed rule would only require investment advisers to collect and verify the identity of customers that directly open and hold accounts (as defined in the proposed rule) with the adviser. This scope of the rule would mitigate the proposed rule's costs just as it would mitigate the proposed rule's benefits as described above. 90

In addition, investment advisers may deem the requirements of the proposed rule for any mutual fund to be satisfied if the customer ( i.e., the mutual fund it advises) has developed and implemented a CIP that is compliant with the investment company's CIP requirements. This provision further lowers the aggregate cost of the proposed rule by negating or minimizing the cost associated with customers that are mutual funds.

RIAs and ERAs would have to establish or in some instances modify a CIP to comply with the requirements of the proposed rule unless they currently have in place a CIP consistent with the proposed rule's requirements. Creating or modifying the policies and procedures detailed in the CIP would entail costs for these advisers. However, investment advisers may already have procedures in place for obtaining identifying information of customers and some investment advisers may have already implemented voluntary AML/CFT programs that are CIP-compliant or that could serve as a framework for a CIP that is consistent with the minimum requirements of the proposed rule. 91 Some investment advisers may have already implemented voluntary AML/CFT programs that are CIP-compliant. In particular, certain investment advisers that use broker-dealers or other financial institutions on behalf of their customers may have these programs in place, as these programs assist those financial institutions to comply with their CIP obligations. 92 Accordingly, such investment advisers may already have written policies and procedures for conducting these or similar activities. The existing infrastructure related to extant practices would reduce the cost of complying with the proposed rule.

Establishing a written CIP would result in additional costs for some investment advisers to the extent they do not have policies and procedures that meet the minimum requirements in the rule. This includes investment advisers that would need to augment their policies and procedures to make them compliant, and costs associated with programming and testing automated systems. FinCEN and the SEC estimate that the average internal time cost for an investment adviser to establish, document and maintain a written CIP as described above would be $1,169.30, with most investment advisers incurring additional ongoing external costs of $584. 93 These estimates imply $23,923,878 in aggregate industry internal costs and $8,961,480 in aggregate industry annual external costs. 94

The proposed rule would require an investment adviser's CIP to contain procedures that specify the identifying information that will be obtained with respect to each customer. This information must include, at a minimum, the name, date of birth (or date of formation), address, and identification number of customers opening new accounts. Investment advisers already obtain from customers identifying information, such as their names and addresses, since most investment advisers need to distinguish their customers operationally and these particular forms of personally identifiable information are common ways of doing so.

Despite this, we estimate that there would be some new costs for investment advisers because some may not be obtaining all the information required by the proposed rule or doing so consistently. These investment advisers would face additional costs in collecting this information and updating their account opening applications or account opening websites to insert line items requesting that customers provide the required information.

The proposed rule would further require an investment adviser's CIP to include procedures to verify the identity of each customer and would provide investment advisers with multiple possible methods to do so. For example, depending on the procedures implemented based on the investment adviser's assessment of the relevant risks, customers that open accounts with an investment adviser can simply provide an unexpired government-issued identification evidencing nationality or residence and bearing a photograph or similar safeguard, such as a driver's license or passport, or if the customer is not an individual, provide a copy of any documents showing its existence as a legal entity ( e.g., certified articles of incorporation, government-issued business licenses, partnership agreements, or trust instruments and any amendments to such documents). Alternatively, investment advisers may, for example, obtain a financial statement from the customer or compare the information provided by the customer with information obtained from a consumer reporting agency or public database.

The documentary and non-documentary verification methods set forth in the rule to verify the identities of customers are not meant to be an exclusive list of the appropriate means of verification. Other reasonable methods may be available now or in the future. The purpose of making the rule flexible in this regard would be to allow investment advisers to select verification methods that are, as section 326 would require, reasonable and practicable. Methods that are appropriate for an investment adviser with a localized customer base may not be sufficient for a different firm with customers from many different countries. The proposed rule recognizes this fact and, therefore, would allow investment advisers to employ such verification methods as would be suitable to form a reasonable belief that it knows the true identities of its customers.

The SEC and FinCEN recognize that obtaining and verifying the identity of each customer would result in incremental costs for many investment advisers if these firms currently do not use verification methods or do not verify identities in a way that is consistent with the proposed rule's requirements. According to the PRA analysis in section V, the average cost of an ERA with two customers (the median number of ERA customers) to obtain and verify the identifying information as described above would be $212.60 in internal cost burdens with most ERAs facing an additional $46.72 in annual ongoing external costs. 95 Similarly, the average internal cost burden for an RIA with 100 customers (the median number of RIA customers) would be $10,630, with most RIAs facing ongoing external annual costs of $2,336. 96 These estimates are based on averages and do not reflect the fact that costs will vary between investment advisers for myriad reasons. In particular, ERA customers are limited to venture capital funds and other private funds. These customers would likely have a smaller per-customer cost than natural person customers.

The proposed rule would also require an investment adviser's CIP to include procedures for responding to circumstances in which the investment adviser cannot form a reasonable belief that it knows the true identity of a customer. While the direct costs of this requirement are included in the estimate above, this requirement may create an additional unquantifiable indirect cost. Specifically, to the extent that any customers who are not intended to be targeted by the proposed rule may be unable to have their identities verified, and thus be subjected to the consequences of this failure to identify (for example, being unable to receive services from the investment adviser), there would be costs associated with temporarily (or possibly in unusual unforeseen circumstances, permanently) losing or having diminished access to financial markets.

The proposed rule would require an investment adviser's CIP to include reasonable procedures for determining whether a customer appears on any list of known or suspected terrorists or terrorist organizations issued by any Federal government agency and designated as such by Treasury in consultation with the Federal functional regulators. Treasury and the Federal functional regulators have not yet designated any such lists. However, for purposes of this economic analysis, we nonetheless estimate the costs of complying with this provision if such lists were to be designated. Our understanding is that some investment advisers and other financial institutions already check their customers against the SDN List. Since the SDN List is often checked in practice and since the creation of such lists that could be provided to investment advisers is a reasonable consideration given this provision in the proposed rule, we are estimating the costs of complying from the current screening practices using the SDN List. We assume, based on staff experience with firms that already check against government lists, that for most accounts this process would be automated and conducted on a batch-file basis, though with significant manual intervention to address false positives. We estimate that the average cost to an ERA with two customers to check such lists would consist of $170.08 in internal time costs with no additional ongoing external costs. 97 Similarly, for the average cost of an RIA with 100 customers would be $8,504 in internal costs with no additional ongoing annual external costs. 98 These estimates are based on averages and do not reflect the fact that costs will vary between investment advisers for myriad reasons. In particular, ERA customers are limited to venture capital funds and other private funds. These customers are exceedingly unlikely to be placed on government lists, and so the costs of compliance with this provision will be lower for ERAs or other types of advisers that solely have funds as customers.

The proposed rule would require an investment adviser's CIP to include procedures for providing their customers adequate notice that the investment adviser is requesting information to verify their identities. Notice would be considered adequate under the proposed rule if the investment adviser generally describes the identification requirements in the proposed rule and provides such notice in a manner reasonably designed to ensure that a customer is able to view the notice, or is otherwise given notice, before opening an account. For example, if an account is opened electronically, such as through an internet website, the investment adviser may provide notice electronically. We estimate the average internal cost burden of an ERA with two customers to provide notice to customers to be $17, with most ERAs facing annual ongoing external costs of $23.36. 99 Similarly, the average internal cost burden of an RIA with 100 customers would be $850, with most RIAs facing additional ongoing annual external costs of $1,168. 100 These estimates are based on averages and do not reflect the fact that costs will vary between investment advisers for myriad reasons.

The proposed rule would require an investment adviser's CIP to include procedures to make and retain records of customers' identifying information for five years after the date of closing of the account and records regarding the verification of a customer's identity for five years after the record is made. We estimate that many of the records required by the rule are already made and maintained by investment advisers. As discussed above, investment advisers already obtain some of the minimum identifying information specified in the proposed rule, and this information is retained for use in firms' operations. 101 We estimate that the recordkeeping requirement could result in additional costs for some investment advisers that currently do not maintain certain of the records for the prescribed time period. We estimate that the average cost to an ERA with two customers to make and maintain the required records would be an internal cost burden of $106.30 with no additional ongoing external costs. 102 Similarly, the average internal cost burden to an RIA with 100 customers would be $5,315, with no additional annual external costs. 103 These estimates are based on averages and do not reflect the fact that costs will vary between investment advisers for myriad reasons.

The proposed rule allows an investment adviser to, under certain circumstances, rely on another financial institution to perform any of the procedures associated with the adviser's CIP. This provision would generally lessen the direct compliance cost of the rule since it would allow these procedures to be done by the party most efficiently positioned to do so and would decrease the likelihood that multiple parties will perform duplicative tasks to comply with regulations affecting different entities. While there may be costs associated with entering or modifying a contract with another financial institution to ensure that the contract's terms have language required by this provision, and there may be monitoring costs to ensure compliance, investment advisers can generally choose to not rely on another financial institution instead if those costs are greater than the cost mitigation that comes from relying on said financial institution.

We recognize that the actual costs associated with establishing a CIP will vary from the estimates above depending on the size of the investment adviser, its lines of businesses, the relevant risks to be addressed by the investment adviser's CIP, and the extent to which the investment adviser's current practices would need to be modified to comply with the requirements. We estimate that the average total cost to an ERA with two customers to comply with the proposed rules would be an internal cost burden of $1,675, with most ERAs facing total annual ongoing external costs of $654. 104 Similarly, the average total internal cost for an RIA with 100 customers would be $26,468, with most RIAs facing total ongoing annual external cost burdens of $4,088. 105 We further estimate total aggregate industry costs of: $404,045,339 in internal time costs and $ 48,446,970 in annual external time costs. 106

We also recognize that these costs would not necessarily be borne solely by investment advisers. Some of these costs could be passed on to the funds and other customers managed by investment advisers. The extent to which these costs would be passed on to customers depends on the interplay of relevant market forces and thus is impossible to predict with accuracy.

The proposed rule provides that the SEC, with the concurrence of the Secretary, may by order or regulation exempt any investment adviser or any type of account from the requirements of this section, or that the Secretary, with the concurrence of the SEC, may exempt any investment adviser or type of account. In issuing such exemptions, the SEC and the Secretary will consider whether the exemption is consistent with the purposes of the BSA, and in the public interest, and may consider other necessary and appropriate factors. This could provide another way to mitigate costs in unforeseen circumstances. For example, if the SEC and the Secretary determine that it is not in the public interest for certain types of accounts to be subject to the requirements of the proposed rule, they may exempt these accounts.

Investment advisers would be required to develop and implement a CIP that complies with the requirements of the proposed rule on or before six months from the effective date of the regulation. Because the overall development burdens are relatively low, 107 we do not believe that this timeline would impose additional costs beyond the direct costs of compliance as quantified in the PRA.

We expect that the requirements would have minimal impact on efficiency, competition, and capital formation. Relative to the size of investment markets, the magnitude of assets that are intended to be targeted are relatively small. Nasdaq estimates that $3.1 trillion in illicit funds entered the global financial system in 2023. 108 State Street Global Advisors, by contrast, has estimated the global market portfolio (the value of all investable capital assets) to be $179 trillion as of December 31, 2021. 109 These estimates suggest that the proposed rule could, at a maximum, impact 1.7 percent of global market funds using an average investment holding period for illicit funds of one year and making the extreme assumption that all illicit funds that enter the global financial system do so through investment advisers. 110 The actual impact is likely to be much lower because investment advisers do not facilitate all funds in the global financial system, and the average time of investment for funds used in money laundering is likely to be shorter than one year. 111 Further, the costs associated with compliance are small enough relative to assets managed by investment advisers so as not to have a significant impact on competition in the investment adviser market. 112

To the extent that the rule would be effective at preventing illicit assets from entering financial markets—for example, if it deters money launderers from attempting to do so or assists investment advisers and law enforcement in discovering these activities—there may be impacts on market efficiency. Specifically, those that engage in money laundering or finance terrorism are likely to have incentives for investment unrelated to the expected return or risk of the asset that differ from the broader market. As a result, their investments change the equilibrium of expected asset risks and returns from what would exist in a market without these illicit funds. For example, money launderers could have very different time horizons for investment and thus could have different liquidity preferences, and so their investments could drive up the premium for liquidity. They likely also have a greater desire to keep their identity hidden and so may choose assets based on this feature. To the extent that money launderers invest based on preferences different from those of the broader market, asset values could, as a result, be distorted relative to what would be efficient for the broader market. Accordingly, it is possible that removing those funds from financial markets would increase market efficiency. The extent to which market efficiency would increase depends on how different money launderers' investment preferences are from those of other investors and how much capital would be effectively prohibited from entering financial markets. We would generally expect these effects to be small, however, given that the magnitude of assets that are intended to be targeted are relatively small compared to the size of investment markets, as discussed above.

Competition may decrease because of the additional compliance costs associated with the proposed rule. However, the relatively small magnitude of estimated costs of the proposed rule, as compared to total assets under management of advisers, suggests that this effect is unlikely to be significant.

The proposed rule is unlikely to have a significant effect on capital formation. To the extent that investors choose to invest more in financial markets because they believe that the proposed rule would reduce the risk of investing by removing illicit funds from the market, capital formation could increase.

FinCEN and the SEC seek comment on all aspects of the economic analysis of the proposed rule, including whether the analysis accurately characterizes the costs and benefits of the minimum requirements set forth by the proposed rule, and whether the specific form of the requirements creates costs or benefits that are not attributable to the statute. To the extent possible, we request that commenters provide supporting data and analysis. In particular, we ask commenters to consider the following questions:

(1) In section IV.B, we state that we do not know what percentage of money laundering crimes sentenced involve investment advisers. Are there sources of data that estimate this percentage? In what ways would this figure be useful for assessing the benefits associated with the proposed rule that is not achieved by the available Treasury analysis of SAR reports?

(2) In section IV.C.1, we state that some investment advisers may already check their customers' accounts against the SDN List. To what extent do investment advisers currently check the SDN List and what factors lead an investment adviser to do so?

(3) In sections IV.C.2 and IV.D.2., we state that investment advisers already collect identifying information required under the proposed rule, either because of contractual obligations or out of operational considerations. Under what circumstances do investment advisers not already collect or retain records of some or all of this information?

(4) In section IV.D.1, we state that we do not have data that would allow us to assess how much money laundering would be reduced as a result of the proposed rule, or how much other illegal activity would be curbed by this reduction in money laundering. What, if any, data exists regarding this activity

(5) In section IV.D.2, we state that some investment advisers may already bear some of the costs of this rule as a result of similar extant requirements on other financial institutions. Is this assumption correct, and if so, how prevalent is this practice?

(6) In Section IV.D.2, we state that allowing an investment adviser to, under certain circumstances, rely on another financial institution to perform any of the procedures associated with the adviser's CIP would generally lessen the direct compliance cost of the rule. To what extent is any anticipated reduction of costs likely to occur, considering the costs of relying on such institutions?

(7) To what extent do investment advisers currently rely on third-party service providers to perform functions related to AML/CFT responsibilities, particularly those associated with a CIP? Would the proposed rule increase or decrease investment advisers' reliance on third-party service providers to perform these functions? If the proposed rule increased investment advisers' reliance on third-party service providers to perform these functions, what additional costs would result?

(8) In places where the costs of the proposed rule are estimated, are these estimates reasonable? Are there any data that could inform the cost estimates of complying with any provisions of the proposed rule? To what extent are there important determinants of costs that could vary between investment advisers that we have not considered in this analysis?

(9) In section IV.G, we discuss the possibility of requiring the Legal Entity Identifier (“LEI”) as the identifier for non-natural person customers. Should the final rule require advisers to use the LEI as the identifier for such customers?

(10) If the adviser knows or has reason to know a customer's assets are maintained at a financial institution that performs CIP requirements because the financial institution is subject to BSA obligations, what would be the benefits and costs of an adviser being required to comply with the proposed rule?

The proposed rule allows various forms of identification for non-natural person customers. We considered whether investment advisers should be required to use the LEI or some other uniform standard as the identifier for such customers.

The LEI is an identification number based on the International Organization for Standardization (“ISO”) 17442-1 standard that uniquely identifies a legal entity. 113 It can facilitate the automatic processing of financial transactions and is used in financial regulatory reporting. For example, the SEC requires an adviser to provide an LEI, if it has one, on Item 1.P on Form ADV. 114

Using the LEI would assist investment advisers and enforcement agencies in detecting money laundering more effectively than using a broad array of identifiers because of its uniformity and relative ease of analysis ( e.g., minimizing any need to map disparate jurisdictional identifiers), though this may be mitigated by the proposed rule's scope which is limited to the direct customers of an adviser and not its beneficial owners. However, this is balanced against the flexibility provided to investment advisers to comply with the rule's requirements. Further, since natural persons could not use this identifier, rule compliance would already necessitate collecting different types of identifiers for these individuals. Moreover, omitting an LEI requirement from the proposed rule would be consistent with the existing rules for broker-dealers and mutual funds, and notwithstanding the absence of an LEI requirement, customers could still provide their LEIs to help advisers satisfy their obligations under the proposed rule. Finally, because legal names and associated LEIs are publicly available, bad actors could use such information to impersonate legitimate entities in their submissions to an investment adviser's CIP and reduce the reliability of the LEI as an identification tool.

The proposed rule would require an investment adviser's CIP to apply to all types of accounts, 115 though the adviser may consider the type of account in determining what procedures are appropriate. We considered whether accounts of customers that do not use the investment adviser to access financial markets, such as those that only receive investment research services, should be excluded from the definition of account.

The benefits of the rule related to such accounts are lower than for other accounts because the benefits relating to protecting financial markets do not directly apply. However, other benefits discussed above, including those relating to identifying criminal activity and preventing illicit proceeds from being legitimized still apply. Further, criminal networks could still use investment adviser products to inform their investment decisions or use investment advisers as ways of appearing legitimate to broker-dealers or other financial institutions. For example, criminals could potentially gain information on how to place or layer illicit funds, even if the investment adviser is not actually doing the placing or layering. An investment adviser's CIP must assess the relevant risks and enact procedures that take account of these risks.

Excluding or otherwise excepting these accounts would eliminate the benefits of the rule for these accounts, but could also reduce the costs of compliance with the rule, if the costs of differentiating these accounts are not higher than the costs of complying with the rule for these accounts. However, we do not believe that these cost savings would be large because: (1) the cost per account of the rule is relatively low; (2) to the extent that these accounts create less risk than do other types of accounts, the compliance cost of these accounts under the proposed rule could be even lower than for other accounts if investment advisers enact procedures with lower costs than those they would establish for riskier types of accounts; and (3) differentiating these accounts may be relatively costly, as investment advisers would need to create new systems to identify which customers only have accounts that would fit the exclusion.

Certain provisions of the proposed rule contain “collection of information” requirements within the meaning of the Paperwork Reduction Act of 1995 (“PRA”). 116 The proposed rule would include new information collection burdens. The title of new collection of information we are proposing is “Amendment to 31 CFR part 1032 under the USA PATRIOT Act.” OMB has not yet assigned a control number for this title. We are submitting the proposed collections of information to the Office of Management and Budget (“OMB”) for review in accordance with the PRA. 117 An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB control number. We discuss below the collection of information burdens associated with the proposal.

If FinCEN's proposed AML/CFT Program and SAR Proposed Rule  118 is adopted, section 326 of the USA PATRIOT ACT would require Treasury and the Commission to prescribe regulations setting forth minimum standards for investment advisers regarding the identities of customers when they open an account. Section 326 also provides that the regulations issued by Treasury and the Commission must, at a minimum, require investment advisers to implement reasonable procedures for: (1) verification of the identity of any person seeking to open an account, to the extent reasonable and practicable; (2) maintenance of the information used to verify the person's identity, including name, address, and other identifying information; and (3) determination of whether the person appears on any lists of known or suspected terrorists or terrorist organizations issued by any government agency. 119 These requirements are referred to as Customer Identification Program (“CIP”) regulations and are long-standing, foundational components of the United States' AML/CFT regime. Under this proposed rule, the CIP must be based on the investment adviser's assessment of the relevant risks, including, at a minimum, those presented by the various types of accounts maintained by the investment adviser, the various methods of opening accounts provided by the investment adviser, the various types of identifying information available and the investment adviser's size, location, and customer base. 120

Under this proposed rule, an investment adviser would be required to retain (1) the identifying information obtained from a customer while the customer's account remains open and for five years after the date the account is closed and (2) the records pertaining to the verification of a customer's identity for five years after the record is made. 121 Each requirement to disclose information, offer to provide information, or adopt policies and procedures constitutes a “collection of information” requirement under the PRA. 122 The respondents to these collection of information requirements would be RIAs and ERAs. As of October 5, 2023, there were approximately 14,914 RIAs and approximately 5,546 ERAs. 123 This collection of information is found at 31 CFR 1032.220 and is mandatory. All RIAs and ERAs would be subject to the requirements of the proposed rule. Responses provided to the Commission in the context of its examination and oversight program concerning the proposed rule would be kept confidential subject to the provisions of applicable law.

Investment adviser implementation of CIPs and reasonable procedures related thereto under this proposed rule would make it easier to prevent, detect, and prosecute money laundering and the financing of terrorism by (i) specifying the information investment advisers must obtain from or about customers that can be used to verify the identity of the customers, (ii) requiring investment advisers to maintain and retain records of the information used to verify the customer's identity, and (iii) requiring investment advisers to determine whether the customer appears on any lists of known or suspected terrorists or terrorist organizations provided by any Federal government agency and designated as such by Treasury in consultation with the Federal functional regulators. This would make it more difficult for persons to use false identities to establish customer relationships with investment advisers for the purposes of laundering money or moving funds to effectuate illegal activities, such as financing terrorism.

We have made certain estimates of the burdens associated with the proposed rule solely for the purpose of this PRA analysis. The table below summarizes the initial and ongoing annual burden and cost estimates associated with the proposed rule.

We request comment on whether these estimates are reasonable. Pursuant to 44 U.S.C. 3506(c)(2)(B), FinCEN and the Commission solicit comments in order to: (1) evaluate whether the proposed collection of information is necessary for the proper performance of the functions of the Commission, including whether the information will have practical utility; (2) evaluate the accuracy of FinCEN and the Commission's estimate of the burden of the proposed collection of information, including whether the estimates are too high or too low; whether the median number of clients is an appropriate figure to use; whether certain costs, such as verification costs, should be lower for certain customers (such as private funds if the adviser forms the private fund) and higher for other types of customers (such as in separately managed account relationships); (3) determine whether there are ways to enhance the quality, utility, and clarity of the information to be collected; and (4) determine whether there are ways to minimize the burden of the collection of information on those who are to respond, including through the use of automated collection techniques or other forms of information technology.

Persons wishing to submit comments on the collection of information requirements of the proposed rule should direct them to the OMB Desk Officer for the Securities and Exchange Commission, MBX.OMB.OIRA.SEC_desk_officer@omb.eop.gov, and should send a copy to Vanessa A. Countryman, Secretary, Securities and Exchange Commission, 100 F Street NE, Washington, DC 20549-1090, with reference to File No. S7-2024-02. OMB is required to make a decision concerning the collections of information between 30 and 60 days after publication of this release; therefore, a comment to OMB is best assured of having its full effect if OMB receives it within 30 days after publication of this release. Requests for materials submitted to OMB by the Commission with regard to these collections of information should be in writing, refer to File No. S7-2024-02, and be submitted to the Securities and Exchange Commission, Office of FOIA Services, 100 F Street NE, Washington, DC 20549-2736.

The SEC and FinCEN have prepared the following Initial Regulatory Flexibility Analysis (“IRFA”) in accordance with section 3(a) of the Regulatory Flexibility Act (“RFA”). 124 It relates to the proposed rule that would amend 31 CFR part 1032 and be issued pursuant to section 326 of the USA PATRIOT Act as amended and codified at 31 U.S.C. 5318( l). 125

The reasons for, and objectives of, the proposed rule are discussed in more detail in sections I and II, above. The burdens of these requirements on small advisers are discussed below as well as above in sections IV and V, which discuss the burdens on all advisers subject to the proposed rule. Sections II through V also discuss the professional skills that compliance with the proposed rule would require.

If FinCEN's proposed AML/CFT Program and SAR Proposed Rule  126 is adopted, section 326 of the USA PATRIOT Act requires Treasury and the Commission to prescribe regulations setting forth minimum standards for investment advisers regarding the identities of customers when they open an account. The statute also would provide that the regulations issued by Treasury and the Commission must, at a minimum, require investment advisers to implement reasonable procedures for: (1) verification of the identity of any person seeking to open an account, to the extent reasonable and practicable; (2) maintenance of the information used to verify the person's identity, including name, address, and other identifying information; and (3) determination of whether the person appears on any lists of known or suspected terrorists or terrorist organizations issued by any government agency. 127 The objective of the proposed rule is to make it easier to prevent, detect and prosecute money laundering and the financing of terrorism. The proposed rule seeks to achieve this goal by requiring investment advisers to establish a CIP with procedures that include obtaining identifying information from customers that can be used to verify the identity of the customers. This will make it more difficult for persons to use false identities to establish customer relationships with investment advisers for the purposes of laundering money or moving funds to effectuate illegal activities, such as financing terrorism. The proposed rule is designed to align the requirements for investment advisers with existing rules for other financial institutions, such as broker-dealers, mutual funds, credit unions, banks, and others, to adopt and implement CIPs.

We are also proposing to revise 31 CFR 1032.100 to provide numerous definitions for purposes of proposed 31 CFR 1032.220. This aspect of the proposed rule has no independent substantive requirements or economic impacts.

The proposed rule is being promulgated pursuant to the BSA, which mandates that FinCEN and the Commission issue a regulation setting forth minimum standards for financial institutions and their customers regarding the identity of the customer that shall apply in connection with opening of an account at the financial institution. 128

The proposed rule would affect investment advisers that are small entities. Under Commission rules, for the purposes of the RFA, an investment adviser generally is a small entity if it: (1) has, and reports on Form ADV, assets under management having a total value of less than $25 million; (2) did not have total assets of $5 million or more on the last day of the most recent fiscal year; and (3) does not control, is not controlled by, and is not under common control with another investment adviser that has assets under management of $25 million or more, or any person (other than a natural person) that had total assets of $5 million or more on the last day of its most recent fiscal year (“small adviser”). 129 The proposed rule would not affect most small advisers, because generally small advisers are registered with one or more state securities authorities and not with the Commission pursuant to section 203A of the Advisers Act. As a result of section 203A, most small advisers are prohibited from registering with the Commission because an investment adviser generally must have more than $25 million of assets under management or be an adviser to a registered investment company. 130 Based on data from the Investment Adviser Registration Depository system (“IARD”), we estimate that as of October 5, 2023, approximately 276 RIAs and 113 ERAs are small entities under the RFA. 131 As discussed above in section IV, FinCEN and the Commission estimate that based on IARD data as of October 5, 2023, approximately 14,914 RIAs and approximately 5,546 ERAs, including all of the approximately 276 RIAs and 113 ERAs that are small entities under the RFA, would be subject to the proposed rule.

The proposed rule would impose certain notification and compliance requirements on investment advisers, including those that are small entities. All RIAs and ERAs, including small entity advisers, would be required to comply with the proposed rule's CIP requirements, which are summarized in this IRFA. All of these requirements are also discussed in detail, above, in sections I and II, and these requirements and the burdens on respondents, including those that are small entities, are discussed above in sections IV and V and below. The professional skills required to meet these specific burdens are also discussed in sections II through V.

There are different factors that would affect whether a smaller adviser incurs costs relating to these requirements that are higher or lower than the estimates discussed in section V. For example, we would expect that smaller advisers may not already have CIP programs, or they may not already have CIP programs that meet certain of the elements that would be required under the proposed rule. Also, while we would expect larger advisers to incur higher costs related to this proposed rule in absolute terms relative to a smaller adviser, we would expect a smaller adviser to find it more costly, per dollar managed, to comply with the requirements because it would not be able to benefit from a larger adviser's economies of scale.

As discussed above, there are approximately 276 RIAs and 113 ERAs that are small entities, and we estimate that 100 percent of these are subject to the proposed rule. As discussed above in section V, the proposed rule, which would require advisers to, among other things, adopt and implement procedures to verify the identity of any customer, would create a new annual burden of approximately 249 hours per RIA and 15.76 hours per ERA, or 70,504.88 hours in aggregate for small advisers (1,780.88 hours for ERAs and 68,724 hours for RIAs). We therefore would expect the annual monetized aggregate cost to small advisers associated with the proposed rule to be approximately $7,494,668.74 ($7,305,361.20 for RIAs and $189,307.54 for ERAs). 132

Investment advisers generally do not have obligations under the BSA specifically for customer identification programs. 133 As a result, we have not identified any federal rules that would duplicate, overlap, or conflict with the proposed rule. If FinCEN's proposed AML/CFT Program and SAR Proposed Rule  134 is adopted, section 326 of the USA PATRIOT Act requires Treasury and the Commission to prescribe regulations setting forth minimum standards for investment advisers regarding the identities of customers when they open an account. This congressional directive cannot be followed absent the issuance of a new rule.

The RFA directs FinCEN and the Commission to consider significant alternatives that would accomplish our stated objective, while minimizing any significant economic effect on small entities. We considered the following alternatives for small entities in relation to the proposed rule: (1) exempting advisers that are small entities from all or part of the proposed rule; (2) establishing different requirements, to account for resources available to small entities; (3) clarifying, consolidating, or simplifying the compliance requirements under the proposed rule for small entities; and (4) using design rather than performance standards.

Regarding the first and second alternatives, FinCEN and the SEC currently believe that establishing different requirements for small advisers, or exempting small advisers from the proposed rule, or any part thereof, would likely be inappropriate under these circumstances. Moreover, FinCEN and the Commission do not believe that those alternatives are appropriate given the flexibility built into the rule to account for, among other things, the differing sizes and resources of advisers, as well as the importance of the statutory goals and mandate of section 326. As discussed above, implementation of CIPs and reasonable procedures related thereto under this proposed rule is intended to assist in preventing, detecting, and prosecuting money laundering and the financing of terrorism by specifying the information investment advisers must obtain from or about customers that can be used to verify the identity of the customers. We assess that this proposed rule would make it more difficult for persons to use false identities to establish customer relationships with investment advisers for the purposes of laundering money or moving funds to effectuate illegal activities, such as financing terrorism. Establishing different conditions for large and small advisers even though advisers of every type and size must open accounts for customers would negate these benefits.

Regarding the third alternative, we believe the rule as proposed is clear and that further clarification, consolidation, or simplification of the compliance requirements is not necessary. As discussed above, the proposed rule would require advisers to, among other things, adopt and implement procedures to verify the identity of any customer, to the extent reasonable and practicable; maintain and retain records of the information used to verify the customer's identity; and determine whether the customer appears on any lists of known or suspected terrorists or terrorist organizations provided by any Federal government agency. 135 The proposed rule would serve as an explicit requirement for firms to adopt and implement a comprehensive CIP.

Regarding the fourth alternative, we determined to use performance standards rather than design standards. Performance standards allow for increased flexibility in the methods firms can use to achieve the objectives of the requirements. Design standards specify the behavior or manner of compliance that regulated entities must adopt. Although the proposed rule would require policies and procedures that are reasonably designed to address a certain number of elements, we do not place certain conditions or restrictions on how to adopt and implement such policies and procedures. The general elements are designed to enumerate core areas that advisers must address when adopting and implementing a CIP. As discussed above, given the number and varying characteristics of advisers, firms would need the ability to design their CIPs in a manner appropriate for their size and business. The proposed rule therefore would allow advisers to address the general elements based on the types of accounts they maintain, the various methods of opening accounts, and the types of identifying information that are available. The proposed rule would also provide flexibility for advisers to determine the personnel who would implement and oversee the effectiveness of their CIPs.

FinCEN and the Commission encourage written comments on the matters discussed in this IRFA. We solicit comment on the number of small entities subject to the proposed rule. We also solicit comment on the potential effects discussed in this analysis; and whether this proposal could have an effect on small entities that has not been considered. We request that commenters describe the nature of any effect on small entities and provide empirical data to support the extent of such effect.

For purposes of the Small Business Regulatory Enforcement Fairness Act of 1996, or “SBREFA,”  136 we must advise OMB whether a proposed regulation constitutes a “major” rule. Under SBREFA, a rule is considered “major” where, if adopted, it results in or is likely to result in (1) an annual effect on the economy of $100 million or more; (2) a major increase in costs or prices for consumers or individual industries; or (3) significant adverse effects on competition, investment, or innovation. We request comment on whether this proposal would be a “major rule” for purposes of the SBREFA. We also request comment on the potential effect of the proposed rule on the U.S. economy on an annual basis; any potential increase in costs or prices for consumers or individual industries; and any potential effect on competition, investment, or innovation. Commenters are requested to provide empirical data and other factual support for their views to the extent possible.

Executive Orders 12866, 13563, and 14094 (that is, E.O. 12866 and its amendments) direct agencies to assess the costs and benefits of available regulatory alternatives and, if regulation is necessary, to select regulatory approaches that maximize net benefits (including potential economic, environmental, and public health and safety effects; distributive impacts; and equity). 137 E.O. 13563 emphasizes the importance of quantifying both costs and benefits, reducing costs, harmonizing rules, and promoting flexibility. E.O. 13563 also recognizes that some benefits are difficult to quantify and provides that, where appropriate and permitted by law, agencies may consider and discuss qualitatively values that are difficult or impossible to quantify. 138

FinCEN has designated this proposed rule a “significant regulatory action;” accordingly, it has been reviewed by the Office of Management and Budget (“OMB”).

FinCEN believes that the primary costs of complying with the proposed rule are considered in the Analysis of the Costs and Benefits Associated with the Proposed Rule described in detail in section IV and the Paperwork Reduction Act (44 U.S.C. 3507(d)) burden estimates described in detail in section V, which amount to a new annual aggregate burden (RIAs and ERAs) of 3,800,990 hours with $404,045,339.05 in internal time costs, and $48,446,969.76 in estimated total new annual external cost burden.

As discussed above in sections IV and V, benefits of this proposed rule are expected to include reduced money laundering and terrorist financing occurring through the U.S. financial system. Overall, the proposed rule would benefit law enforcement by improving their ability to investigate, prosecute and disrupt the financing of international terrorism and other priority transnational security threats, as well as other types of transnational financial crime. Obtaining and verifying the identity of account holders or responding to circumstances in which the investment adviser cannot form a reasonable belief that it knows the true identity of a customer would reduce the risk of terrorists and other criminals accessing U.S. financial markets to launder money, finance terrorism, or move funds for other illicit purposes. The proposed rule would also help investment advisers to identify and prevent criminal activity including by allowing investment advisers to identify high risk customers. While it is difficult to estimate the economic losses that would be prevented by reducing money laundering and other financial crimes through this rule, the prevention of such crimes would reduce the monetary and nonmonetary harms they cause.

As an alternative to the proposed rule, as discussed in section IV, FinCEN considered requiring investment advisers use the LEI or some other uniform standard as the identifier for such customers. While using the LEI would assist investment advisers and law enforcement agencies to detect money laundering than using a number of identifiers, the proposed rule's application to direct customers rather than beneficial owners limits the benefit of using LEIs. Further, natural persons could not use this identifier, meaning compliance with the proposed rule would require the collection of different types of identifiers.

Regarding costs, as noted above in sections IV and V, in accordance with section 326 of the USA PATRIOT ACT, the proposed rule would require an investment adviser to establish and implement its CIP according to its specific circumstances and do not set inflexible requirements for all advisers. Further, the proposed requirements are similar to those for other financial institutions with which investment advisers engage; complying with the proposed rule may therefore create minimal additional costs in certain lines of business. Some RIAs and ERAs may have reduced costs because they may already perform certain AML/CFT functions because they are dual registrants or affiliated with a bank or broker-dealer. Finally, per the analysis above in sections IV and V, investment advisers may deem the requirements of the proposed rule with respect to its business relationship with a mutual fund to be satisfied if the customer ( i.e., the mutual fund that it advises) has developed and implemented a CIP that is compliant with the investment company's CIP requirements.

The costs incurred by the proposed rule would arise through the following requirements: establishing a CIP; obtaining and verifying identifying information; determining whether customers appear on a federal government list; providing notice to customers; recordkeeping; and reliance on another financial institution. Overall, FinCEN estimates that the average total cost to an ERA with two customers to comply with the proposed rules would be an internal cost burden of $1,675, with most ERAs facing total annual ongoing external costs of $654, while the average total internal cost for an RIA with 100 customers would be $26,468, with most RIAs facing total ongoing annual external cost burdens of $4,088.

Given the analysis included in the preceding sections, FinCEN believes that the benefits of this rule would exceed the costs.

FinCEN has analyzed the rule under the factors set forth in the Unfunded Mandates Reform Act (“UMRA”) (section 202(a)). Under this analysis, FinCEN considered whether the proposed rule includes any Federal mandate that may result in the expenditure by State, local, and tribal governments, in the aggregate, or by the private sector, of $100,000,000 or more (adjusted annually for inflation) in any 1 year.”  139 The current threshold after adjustment for inflation is $176 million, using the 2022 GDP price deflator. The proposed rule would result in an expenditure in at least one year that meets or exceeds this amount.

FinCEN further estimates total aggregate industry costs of: $404,045,339.05 in internal time costs and $48,446,969.76 in annual external time costs. 140 The proposed rule does not foreseeably impose costs or other compliance burden that would impact any State, local, or Tribal government. FinCEN believes that the cost benefit analysis in section IV. Analysis of the Costs and Benefits Associated with the Proposed Rule, provides the analysis required by UMRA.

For the reasons set forth in the preamble, FinCEN and the SEC propose to add part 1032 to chapter X in title 31 of the Code of Federal Regulations to read as follows:

To be considered as part of this rulemaking, comments and requests for a hearing must be submitted in response to this proposed rule within the timeframe specified above, regardless of whether the comment, hearing request, or other information was previously submitted to the Drug Enforcement Administration (“DEA”) in connection with any prior matter relating to the scheduling of marijuana.

Please note that all comments received in response to this docket are considered part of the public record. DOJ will make comments available for public inspection online at https://www.regulations.gov. Such information includes personal or business identifiers (such as name, address, State or Federal identifiers, etc.) voluntarily submitted by the commenter. Generally, all information voluntarily submitted by the commenter, unless clearly marked as “Confidential Information” in the method described below, will be publicly posted. Comments may be submitted anonymously. The Freedom of Information Act, 5 U.S.C. 552, applies to all comments received.

Commenters submitting comments that include personal identifying information (“PII”) or confidential or proprietary business information that the commenter does not want made publicly available should submit two copies of the comment. One copy must be marked “CONTAINS CONFIDENTIAL INFORMATION” and should clearly identify all PII or business information the commenter does not want to be made publicly available, including any supplemental materials. DOJ will review this copy, including the claimed PII and confidential business information, in its consideration of comments. The second copy should be marked “TO BE PUBLICLY POSTED” and must have all claimed confidential PII and business information already redacted. DOJ will post only the version of the comment with redactions on https://www.regulations.gov for public inspection.

An electronic copy of this document and supplemental information to this proposed rule are available at https://www.regulations.gov for easy reference. DOJ specifically solicits written comments regarding the economic analysis of the impact of these proposed changes. DOJ requests that commenters provide detailed descriptions in their comments of any expected economic impacts, especially to small entities. Commenters should provide empirical data to illustrate the nature and scope of such impact.

Pursuant to 21 U.S.C. 811(a), this scheduling action is a formal rulemaking “on the record after opportunity for a hearing.” Such proceedings are conducted pursuant to the Administrative Procedure Act (“APA”), 5 U.S.C. 551-559. See 21 CFR 1308.41 through 1308.45; id part 1316, subpart D. Interested persons, as defined in 21 CFR 1300.01(b), may file requests for a hearing in conformity with the requirements of 21 CFR 1308.44(a) and 1316.47(a), and such requests must:

(1) state with particularity the interest of the person in the proceeding;

(2) state with particularity the objections or issues concerning which the person desires to be heard; and

(3) state briefly the position of the person regarding the objections or issues.

All requests for a hearing and waivers of an opportunity for a hearing or participation, together with a written statement of position on the matters of fact and law involved in such hearing, must be sent to DEA using the address information provided above.

The decision whether an in-person hearing will be needed to address such matters of fact and law in the rulemaking will be made by the Administrator of DEA. Upon the Administrator's determination to grant an in-person hearing, DEA will publish a notice of hearing on the proposed rulemaking in the Federal Register . See 21 CFR 1308.44(b), 1316.53.

If the Administrator determines to grant an in-person hearing to address such matters of fact and law in this rulemaking, the Administrator will then designate an Administrative Law Judge (“ALJ”) to preside over the hearing. The ALJ's functions shall commence upon designation, as provided in 21 CFR 1316.52. The ALJ will have all powers necessary to conduct a fair hearing, to take all necessary action to avoid delay, and to maintain order. Id. The ALJ's authorities include the power to hold conferences to simplify or determine the issues in the hearing or to consider other matters that may aid in the expeditious disposition of the hearing; require parties to state their position in writing; sign and issue subpoenas to compel the production of documents and materials to the extent necessary to conduct the hearing; examine witnesses and direct witnesses to testify; receive, rule on, exclude, or limit evidence; rule on procedural items; and take any action permitted by the presiding officer under DEA's hearing procedures and the APA. Id.

Comments on or objections to the proposed rule submitted under 21 CFR 1308.43(g) will be offered as evidence at the hearing, but the presiding officer shall admit only evidence that is competent, relevant, material, and not unduly repetitive. 21 CFR 1316.59(a).

Any interested person may file a waiver of opportunity for a hearing or to participate in a hearing in conformity with the requirements of 21 CFR 1308.44(c), together with a written statement of position on the matters of fact and law involved in any hearing. 21 CFR 1316.49. Such statement, if admissible, will be included in the record and considered as described in 21 CFR 1308.44(c).

In accordance with 21 U.S.C. 811 and 812, the purpose of a hearing would be to “receiv[e] factual evidence and expert opinion regarding” whether marijuana should be transferred to schedule III of the list of controlled substances. 21 CFR 1308.42. Concurrent with this rulemaking, DEA will consider the marijuana-specific controls that would be necessary to comply with relevant treaty obligations in the event that, after the hearing, a final order reschedules marijuana, and, to the extent such controls are needed if marijuana is rescheduled, will seek to finalize any such regulations as soon as possible.

All requests for hearing and waivers of an opportunity for a hearing or participation must be sent to DEA using the address information above, on or before the date specified above.

Under the CSA, 21 U.S.C. 801 et seq., the Attorney General shall, before initiating proceedings to control, decontrol, or transfer between schedules a drug or other substance, request from the Secretary of HHS a scientific and medical evaluation, and the Secretary's recommendations, as to whether such drug or other substance should be so controlled or removed as a controlled substance. 21 U.S.C. 811(b). The recommendations of the Secretary shall include recommendations with respect to the appropriate schedule, if any, under which such drug or other substance should be listed. Id.

HHS recommended in August 2023 that marijuana be rescheduled to schedule III. See Letter for Anne Milgram, Administrator, DEA, from Rachel L. Levine, M.D., Assistant Secretary for Health, HHS (Aug. 29, 2023) (“August 2023 Letter”). The Attorney General then sought the legal advice of the Office of Legal Counsel (“OLC”) at DOJ on questions relevant to this rulemaking proceeding. Among other conclusions, OLC concluded that “HHS's scientific and medical determinations must be binding until issuance of a notice of proposed rulemaking [(`NPRM')].” Questions Related to the Potential Rescheduling of Marijuana, 45 Op. O.L.C. __, at *25 (Apr. 11, 2024) (“OLC Op.”). 1 After the issuance of a notice of rulemaking proceedings, HHS's scientific and medical determinations are accorded “significant deference” through the rest of the rulemaking process. 2 OLC Op. at *26.

Under the CSA, when recommending or determining that a drug should be controlled (and if so, under which schedule), the Secretary and the Attorney General must consider eight factors set forth in 21 U.S.C. 811(c). The eight factors are:

1. The drug's actual or relative potential for abuse;

2. Scientific evidence of its pharmacological effect, if known;

3. The state of current scientific knowledge regarding the drug or other substance;

4. Its history and current pattern of abuse;

5. The scope, duration, and significance of abuse;

6. What, if any, risk there is to the public health;

7. Its psychic or physiological dependence liability; and

8. Whether the substance is an immediate precursor of a substance already controlled.

The United States is a party to the 1961 United Nations Single Convention on Narcotic Drugs, March 30, 1961, 18 U.S.T. 1407, 520 U.N.T.S. 151 (“Single Convention”), as amended by the 1972 Protocol, March 25, 1972, 26 U.S.T. 1439, 976 U.N.T.S. 3. Under 21 U.S.C. 811(d)(1), if control of a substance is required “by United States obligations under international treaties, conventions, or protocols in effect on October 27, 1970,” the Attorney General must issue an order controlling such drug “under the schedule he deems most appropriate to carry out such obligations, without regard to the findings” required by 21 U.S.C. 811(a) or 812(b), “and without regard to the procedures” prescribed by 21 U.S.C. 811(a) and (b). Marijuana is a drug covered by the Single Convention. See Single Convention art. 1(1)(b); OLC Op. at *26 & n.7.

OLC and the United States Court of Appeals for the D.C. Circuit have explained that section 811(d)(1) does not supersede the scheduling procedures set forth in sections 811(a) through (b) and 812(b), including the requirement to consider the eight factors set forth in section 811(c). Instead, section 811(d)(1) allows the Attorney General to “identify which schedules would satisfy the United States' international obligations with respect to a particular drug, and then—if more than one schedule would do so—select which schedule to use through the section 811(a) through (b) and 812(b) procedures.” OLC Op. at *29 n.8; accord Nat'l Org. for Reform of Marijuana Laws (NORML II) v. DEA, 559 F.2d 735, 747 (D.C. Cir. 1977). HHS performed the eight-factor analysis. See Memorandum for DEA, from HHS, Re: Basis for the Recommendation to Reschedule Marijuana to Schedule III of the Controlled Substances Act (“HHS Basis for Rec.”). As noted above, HHS's scientific and medical determinations are binding on DOJ until an NPRM is published, and, in addition, DOJ must accord “significant deference” to HHS's scientific and medical determinations throughout the rulemaking process. OLC Op. at *25-26.

Once the determination is made that a particular drug or substance must be controlled under the CSA, the Attorney General must determine the level of control over the drug or substance under the CSA. See 21 U.S.C. 811(a), (b). The CSA divides controlled substances into five levels of control, or “schedules,” based on (1) a drug's potential for abuse, (2) whether the drug has a currently accepted medical use in treatment in the United States (“CAMU”), and (3) whether there is a lack of accepted safety for use of the drug under medical supervision or the level of psychological or physical dependence that could result from abuse of the drug. See id. 812(b). Schedule I drugs have a high potential for abuse, no CAMU, and a lack of accepted safety for use under medical supervision. Id. 812(b)(1). Schedule II drugs also have a high potential for abuse but have a CAMU (or a CAMU with “severe restrictions”), and abuse of the drug may lead to severe psychological or physical dependence. Id. 812(b)(2). Schedule III drugs, meanwhile, have a lower potential for abuse when compared to drugs in schedules I and II, have a CAMU, and their abuse may lead to moderate or low physical dependence or high psychological dependence. 3 21 U.S.C. 812(b)(3). The initial schedules of controlled substances established by Congress are found at 21 U.S.C. 812(c), and the current list of all scheduled substances is published at 21 CFR part 1308.

Pursuant to 21 U.S.C. 811(a)(1), the Attorney General may, by rule, add to such a schedule or transfer between such schedules any drug or other substance if he (A) finds that such drug or other substance has a potential for abuse, and (B) makes with respect to such drug or other substance the findings prescribed by 21 U.S.C. 812(b) for the schedule in which such drug is to be placed. The CSA provides that proceedings for the issuance, amendment, or repeal of the scheduling of any drug or other substance may be initiated by the Attorney General on his own motion, at the request of the Secretary, or on the petition of any interested party. Id.

When Congress enacted the CSA in 1970, it placed marijuana in schedule I. Comprehensive Drug Abuse Prevention and Control Act of 1970, Public Law 91-513, tit. II, sec. 202(c), 84 Stat. 1236, 1249 (1970); 21 U.S.C. 812(c). 4 The Attorney General is authorized to amend this initial placement. 21 U.S.C. 812(c); see also id. 811, 812(b). Other schedule I substances include heroin, lysergic acid diethylamide (commonly known as LSD), and 3,4-methylenedioxymethamphetamine (ecstasy). See 21 CFR 1308.11. Drugs controlled under schedule II include cocaine, methamphetamine, methadone, oxycodone, and fentanyl. Id. § 1308.12. Drugs controlled under schedule III include products containing less than 90 milligrams of codeine per dosage unit, ketamine, and anabolic steroids. Id. § 1308.13. Petitioners have requested that marijuana be rescheduled several times over the years. See, e.g., Schedule of Controlled Substances: Petition To Remove Marihuana From Control or in the Alternative To Control Marihuana in Schedule V of the Controlled Substances Act, 37 FR 18097 (Sept. 7, 1972); Notice of Denial of Petition, 66 FR 20038 (Apr. 18, 2001); Denial of Petition To Initiate Proceedings To Reschedule Marijuana, 76 FR 40552 (July 8, 2011).

DEA and HHS last examined the issue of whether to reschedule marijuana eight years ago, in 2016, when DEA denied two petitions to reschedule marijuana. 5 At the time, HHS concurred that marijuana should remain a schedule I drug because it met the three criteria for placement in schedule I. 81 FR 53706-07. In accordance with the requirements for placement in schedule I, HHS found that: (1) marijuana had a high potential for abuse; (2) it did not have a CAMU; and (3) there was a lack of accepted safety for use of marijuana under medical supervision. Id. As discussed in detail below, in 2023, HHS conducted a scientific and medical evaluation of marijuana based on a comprehensive review of available data at that time and recommended that marijuana be transferred to schedule III.

Since 1996, 38 States, the District of Columbia, and 4 Federal Territories have legalized the use of medical marijuana. HHS Basis for Rec. at 30; OLC Op at *9. These laws typically allow the cultivation, sale, and use of marijuana by patients (or their caregivers) whose health care practitioners have recommended that they use marijuana to treat certain health conditions. See, e.g., Ohio Rev. Code secs. 3796.01(A)(6)(a)-(v), 3796.01(A); N.Y. Cannabis Law secs. 3(18), 30, 31; N.M. Stat. secs. 26-2B-3(F)(1)-(23), 26-2B-3(N), 26-2B-4(A). Further, beginning in Fiscal Year 2015, Congress has adopted an appropriations rider every year that prohibits DOJ from using funds to prevent certain States, Territories, and the District of Columbia from implementing their own laws with respect to medical marijuana. E.g., Consolidated Appropriations Act, 2024, Public Law 118-42, sec. 531, 138 Stat. 25; Consolidated Appropriations Act, 2023, Public Law 117-328, sec. 531, 136 Stat. 4459, 4561 (2022); see also Cong. Research Serv., R44782, The Evolution of Marijuana as a Controlled Substance and the Federal-State Policy Gap 26 & n.159 (updated Apr. 7, 2022) (collecting additional appropriations riders).

Marijuana is generally defined by statute to mean “the plant Cannabis sativa L., whether growing or not; the seeds thereof; the resin extracted from any part of such plant; and every compound, manufacture, salt, derivative, mixture, or preparation of such plant, its seeds or resin.” 21 U.S.C. 802(16)(A). In 2018, Congress amended the CSA to remove “(i) hemp, as defined in section [1639 o of title 7 of the U.S. Code]” from the definition of marijuana. 6 Agricultural Improvement Act of 2018, Public Law 115-334, sec. 12619, 132 Stat. 4490, 5018. Section 1639 o (1) of title 7 in turn defines hemp as “the plant Cannabis sativa L. and any part of that plant, including the seeds thereof and all derivatives, extracts, cannabinoids, isomers, acids, salts, and salts of isomers, whether growing or not, with a delta-9-tetrahydrocannabinol concentration of not more than 0.3 percent on a dry weight basis.” Delta-9-tetrahydrocannabinol (“Δ9-THC”) is the major psychoactive intoxicating cannabinoid in marijuana. See HHS Basis for Rec. at 10. This exclusion of hemp from the definition of marijuana had the effect of removing many products containing predominantly cannabidiol (“CBD”) derived from hemp and containing no more than 0.3 percent Δ9-THC on a dry weight basis from control as marijuana.

On October 6, 2022, President Biden requested that the Attorney General and the Secretary of HHS “initiate the administrative process to review expeditiously how marijuana is scheduled under federal law.”  7 HHS thereafter undertook a scientific and medical evaluation of marijuana as defined under the CSA in accordance with the President's request.

In a letter dated August 29, 2023, Admiral Rachel L. Levine, M.D., HHS's Assistant Secretary for Health, recommended to the Administrator of DEA that marijuana be controlled in schedule III of the CSA. August 2023 Letter. HHS found that marijuana has a potential for abuse less than the drugs or other substances in schedules I and II; that marijuana has a CAMU; and that the abuse of marijuana may lead to moderate or low physical dependence or high psychological dependence. HHS Basis for Rec. at 62-65. These findings correspond to the criteria for placement of a substance in schedule III. See 21 U.S.C. 812(b)(3). DEA has not yet made a determination as to its views of the appropriate schedule for marijuana.

The CSA vests the Attorney General with the authority to schedule, reschedule, or decontrol drugs. 21 U.S.C. 811(a). The Attorney General has delegated that authority to the DEA Administrator, see 28 CFR 0.100, but also retains the authority to schedule drugs under the CSA in the first instance, see 28 U.S.C. 509, 510. The HHS Assistant Secretary for Health has provided a recommendation for transferring marijuana to schedule III. In light of that recommendation, the Attorney General is exercising the Attorney General's authority under 21 U.S.C. 811(a) to initiate a rulemaking that proposes the placement of marijuana in schedule III.

DEA believes that additional information arising from this rulemaking will further inform the findings regarding the appropriate schedule for marijuana. DEA has maintained an active review of the scientific literature addressing marijuana with a focus on how it relates to the scientific and medical evaluation and informs any updates to the eight-factor analysis. In addition to HHS's scientific and medical determinations, which are binding until the issuance of this NPRM and which must be accorded significant deference throughout the rulemaking, DEA believes that factual evidence (including scientific data) and expert opinions, including additional data regarding different forms, formulations, and delivery methods for marijuana, as well as evidence regarding the effects of marijuana at various dosages or concentrations, may be relevant.

The HHS Basis for Recommendation, DEA's analyses explaining its decisions to deny the petitions to reschedule marijuana in 2016, and the 2024 OLC opinion (cited throughout) are available in their entirety under “Supporting and Related Material” of the public docket for this proposed rule at https://www.regulations.gov under docket number DEA-1362.

DOJ has reviewed the scientific and medical evaluation and scheduling recommendation provided by HHS and has conducted a separate review of the eight factors identified in 21 U.S.C. 811(c). At this point in the proceedings, DOJ must treat HHS's scientific and medical determinations as binding. See OLC Op. at *4, *25. HHS's scientific and medical determinations are included below, as well as certain information from DEA.

The first factor that DOJ and HHS must consider under 21 U.S.C. 811(c) is the actual or relative potential for abuse of marijuana. The term “abuse” is not defined in the CSA. However, consistent with the legislative history of the CSA, DEA and HHS have typically weighed the following factors in determining whether a particular drug or substance has a potential for abuse:  8

A. Whether there is evidence that individuals are taking the drug or drugs containing such a substance in amounts sufficient to create a hazard to their health or to the safety of other individuals or to the community.

As part of its analysis, HHS concluded that evidence shows that, although some individuals are taking marijuana in amounts sufficient to create a hazard to their health and to the safety of other individuals and the community, the vast majority of individuals who use marijuana are doing so in a manner that does not lead to dangerous outcomes to themselves or others. HHS Basis for Rec. at 6-7. The data supportive of this conclusion are discussed in detail in HHS's analysis of Factors 4, 5, and 6. See HHS Basis for Rec. at 28-57.

In particular, HHS emphasized that an evaluation of various epidemiological databases of adverse outcomes from 2015 to 2021 involving marijuana or comparator drugs that are used nonmedically showed that the utilization-adjusted rate of adverse outcomes involving marijuana was consistently lower than the utilization-adjusted rates of adverse outcomes involving heroin, cocaine, and, for certain outcomes, other comparators, including alcohol. Also, alcohol or heroin typically ranked first or in immediately subsequent positions among the comparators in terms of incidence of adverse outcomes, with marijuana in a lower place in that ranking. This pattern also was observed for serious medical outcomes, including death, observed in Poison Center data, where marijuana was in the lowest ranking group. This suggests consistency across databases, across drugs, and over time. HHS thus concluded that although abuse of marijuana produces clear evidence of harmful consequences, these appear to be relatively less common and less severe than the consequences of some other comparator substances. HHS Basis for Rec. at 7-8.

Importantly, these comparisons of the prevalence of adverse outcomes were from descriptive analyses only, following the established practice in previous eight-factor analyses. Thus, differences in outcome frequency and severity, and the ranked order across comparators, may be attributable in part to underlying differences in the populations being compared ( e.g., age or pre-existing medical conditions), among other things. Despite these limitations, qualitative synthesis of descriptive analyses is the established practice in previous eight-factor analyses, and HHS determined that it is the most appropriate approach here. HHS Basis for Rec. at 7-8.

HHS also concluded that the public-health risks posed by marijuana are lower compared to those posed by other drugs of abuse ( e.g., heroin, oxycodone, cocaine), based on HHS's evaluation of various epidemiological databases for emergency department (“ED”) visits, hospitalizations, unintentional exposures, and most importantly, overdose deaths. The rank order of the comparators in terms of greatest adverse consequences typically ranked heroin, benzodiazepines, and cocaine first or in immediately subsequent positions, with marijuana in a lower place in the ranking, especially when HHS adjusted for utilization. For overdose deaths, marijuana is always in the lowest ranking among comparator drugs. These evaluations demonstrate that there is consistency across databases, across substances, and over time. HHS thus concluded that although abuse of marijuana produces clear evidence of a risk to public health, that risk is relatively lower than that posed by most other comparator drugs. HHS Basis for Rec. at 7-8.

DEA notes that data provided by HHS in its recommendation included a 2023 national survey that tracks drug use trends among 8th-, 10th-, and 12th-grade students, and showed that by 12th grade, 20.2 percent of students reported using marijuana in the past month. 9 DEA also notes that the same study showed that the prevalence of ingesting marijuana by vaping is evidenced by students reporting vaping in the 30 days prior at the following rates: 4.2 percent for 8th graders, 10.3 percent for 10th graders, and 14.8 percent for 12th graders. 10 In 2022, the Substance Abuse and Mental Health Services Administration's (“SAMHSA”) Drug Abuse Warning Network (“DAWN”) reported that 11.9 percent of drug-related ED visits nationwide involved cannabis. 11 The rate of cannabis-related ED visits was highest in these demographic groups: 18 to 25 years old, male, Black or African American, and Not Hispanic or Latino. 12

In addition to the data considered by the HHS Basis for Recommendation, the data considered by HHS and DEA in their 2015 eight-factor analysis, and the additional data discussed above, DEA anticipates that additional data on seizures of marijuana by law enforcement, cannabis-related ED visits, as well as updated epidemiological survey data since 2022, may be appropriate for consideration.

B. Whether there is significant diversion of the drug or drugs containing such a substance from legitimate drug channels.

HHS found that there is a lack of evidence of significant diversion of marijuana from legitimate drug channels. HHS Basis for Rec. at 8. It noted that marijuana is used by researchers for clinical research under investigational new drug (“IND”) applications, and that there are multiple DEA registrants that are approved to produce marijuana and derived formulations for use in DEA-authorized nonclinical and clinical research. HHS observed that these authorizations represent the only federally sanctioned drug channels in the United States, and there is a lack of data indicating diversion occurring from these entities or activities. However, there are significant additional sources of marijuana in the United States, including from illicit cultivation and production, illicit importation from other countries, and from State programs that permit dispensing of marijuana for medical use and, in some States, recreational adult use. HHS Basis for Rec. at 8.

Given this unique landscape, DEA believes that the lack of data indicating diversion of marijuana from federally sanctioned drug channels to the illicit market is not indicative of a lack of potential for abuse of the drug. DEA anticipates that additional data on diversion from State programs and DEA-registered manufacturers may aid in a determination of whether diversion is taking place.

C. Whether individuals are taking the drug or drugs containing such a substance on their own initiative rather than on the basis of medical advice from a practitioner licensed by law to administer such drugs in the course of their professional practice.

As HHS notes, the Food and Drug Administration (“FDA”) has not approved a New Drug Application (“NDA”) for a drug product containing botanical marijuana for any therapeutic indication. Thus, the only way an individual can use marijuana on the basis of medical advice through legitimate channels under Federal law is by participating in research under an IND. However, 38 States and the District of Columbia have enacted laws allowing individuals to use marijuana under certain circumstances for medical purposes. Outside of the Federal- and State-sanctioned medical use of marijuana, individuals are using marijuana on their own initiative for medical, as well as nonmedical, purposes. Epidemiological data related to nonmedical use of marijuana is detailed in HHS's analysis of Factor 4. HHS Basis for Rec. at 8.

DEA notes that data is not available to determine the number of individuals using marijuana under State law. According to 2022 National Survey on Drug Use and Health (“NSDUH”) data on people who are 12 and older in the United States, 61.9 million people reported using marijuana in the past year, and marijuana was the illicit drug used with the greatest frequency. 13 Specifically, 42.3 million people reported use in the past month, including 14.7 million people who vaped marijuana in that same period, representing 5.2 percent of the study's target population. 14 Furthermore, as reported by NSDUH in 2022, 3.7 million people initiated marijuana use in the past year, with more than half (53 percent or 2.0 million people) initiating marijuana use before the age of 21. 15 DEA also notes that HHS concluded that, outside of the Federal- and State-sanctioned medical use of marijuana, individuals are using marijuana on their own initiative for medical as well as nonmedical purposes. HHS Basis for Rec. at 8. In 2016, DEA reached a similar conclusion. 16 In addition to the data considered in the HHS Basis for Recommendation, and by HHS and DEA in their earlier eight-factor analyses, DEA anticipates that updated epidemiological survey data since 2022 may be appropriate for consideration.

D. Whether the drug or drugs containing such a substance are new drugs so related in their action to a drug or drugs already listed as having a potential for abuse to make it likely that it will have the same potentiality for abuse as such drugs, thus making it reasonable to assume that there may be significant diversions from legitimate channels, significant use contrary to or without medical advice, or that they have a substantial capability of creating hazards to the health of the user or to the safety of the community.

Marijuana has been a schedule I substance since the CSA was enacted in 1970. See Public Law 91-513, tit. II, sec. 202(c), 84 Stat. 1236, 1249 (1970); 21 U.S.C. 812(c); see also 21 CFR 1308.11(d)(23). The primary compound in marijuana that is responsible for its abuse potential is Δ9-THC (also known as dronabinol, when specifically referring to the (-)-trans- Δ9-THC stereoisomer), which has agonist activity at cannabinoid CB 1 receptors. HHS found that there are extensive nonclinical and clinical studies establishing that marijuana, due to the CB 1 agonist activity of its main cannabinoid constituent Δ9-THC, produces rewarding effects that would be consistent with observed long-term patterns of nonmedical use and abuse, both before and in the years since enactment of the CSA. HHS Basis for Rec. at 9. For further discussion of these effects, see HHS Basis for Rec. at 9-18 (Factor 2), 28-37 (Factor 4).

Additionally, FDA has approved two drug products containing dronabinol: Marinol (in 1985; schedule III) and Syndros (in 2016; schedule II). HHS Basis for Rec. at 9. Marinol was approved by FDA in 1985 for the treatment of nausea and vomiting associated with cancer chemotherapy in patients who did not respond to conventional anti-emetic treatments. FDA approved Marinol in 1992 for the treatment of anorexia associated with weight loss in patients with acquired immunodeficiency syndrome (“AIDS”). After the first FDA approval, Marinol was transferred from schedule I to schedule II and was later rescheduled to schedule III. Syndros, a drug product also containing dronabinol but formulated in an oral solution, was approved by FDA in 2016 for the treatment of anorexia associated with weight loss in patients with AIDS, as well as nausea and vomiting associated with cancer chemotherapy in patients who failed to respond adequately to conventional anti-emetic treatments. In 2017, DEA rescheduled “FDA-approved products containing dronabinol in an oral solution” from schedule I into schedule II. HHS Basis for Rec. at 4.

When Marinol and Syndros were being developed, they underwent a systematic evaluation of their abuse potential based on animal and human behavioral studies, which showed that dronabinol has abuse potential. The abuse-related studies confirmed the abuse potential of Δ9-THC. HHS has concluded that these findings suggest that marijuana will continue to be used nonmedically, diverted from legitimate channels, and trafficked in illicit channels as a potential source for continued nonmedical use in the United States. HHS Basis for Rec. at 9; see also HHS Basis for Rec. at 37-45 (Factor 5).

HHS determined that epidemiological data indicate that marijuana has the potential for creating hazards to the health of the user and to the safety of the community. However, as a relative finding on abuse liability, when comparing marijuana to heroin, oxycodone, hydrocodone, fentanyl, cocaine, ketamine, benzodiazepines, zolpidem, tramadol, and alcohol in various epidemiological databases that allow for some or all of these comparisons, marijuana is not typically among the substances producing the most frequent incidence of adverse outcomes or severity of substance use disorder. HHS Basis for Rec. at 9; see also HHS Basis for Rec. at 28-57 (Factors 4, 5, and 6). But as noted above, there are limitations in comparing descriptive data on adverse outcomes across drugs, although descriptive analyses of epidemiologic data are an established practice in previous eight-factor analyses. HHS Basis for Rec. at 9.

In 2016, DEA found that “[m]arijuana has a high potential for abuse. Preclinical and clinical data show that it has reinforcing effects characteristic of drugs of abuse. . . . Data on marijuana seizures show widespread availability and trafficking.” 81 FR 53739. DEA believes that additional data in this area may be appropriate for consideration in assessing marijuana's actual or relative potential for abuse.

The second factor that DOJ and HHS must consider under 21 U.S.C. 811(c) is the scientific evidence of marijuana's pharmacological effects, if known. In making its recommendation, HHS considered the scientific evidence of the pharmacological effects of marijuana based on the effects of Δ9-THC. HHS conducted a scientific evaluation of the neurochemistry, receptor pharmacology, animal abuse-related behavioral effects, and human behavioral and physiological effects of marijuana. HHS Basis for Rec. at 9.

Cannabis is the genus of a plant that contains numerous natural constituents, including cannabinoids. See HHS Basis for Rec. at 18-21 (discussing Factor 3). Because cultivated chemovars may vary in their composition and concentration of various chemical constituents, including with respect to whether they contain significant amounts of Δ9-THC or other cannabinoids, marijuana products from different strains will have differing biological and pharmacological profiles. HHS Basis for Rec. at 10.

Marijuana contains at least 560 identified natural constituents, including 125 compounds classified as cannabinoids. Most major cannabinoid compounds occurring naturally in cannabis have been identified chemically, but new and minor compounds are continuously being characterized. HHS Basis for Rec. at 10.

The two most abundant cannabinoids present in marijuana are Δ9-THC and CBD. Δ9-THC is the major psychoactive intoxicating cannabinoid in marijuana and is the component of marijuana that is primarily responsible for its abuse potential. In contrast, CBD has negligible abuse potential, as assessed by FDA during the NDA review for Epidiolex, an FDA-approved drug product containing plant derived, highly purified CBD. HHS Basis for Rec. at 10.

There are two cannabinoid receptors: CB 1 and CB 2 . CB 1 and CB 2 receptors belong to the family of G-protein-coupled receptors and present a typical seven transmembrane-spanning domain structure. Cannabinoid receptors primarily link to an inhibitory G protein (G i/o ), such that adenylate cyclase activity is inhibited when a cannabinoid ligand binds to the receptor. This, in turn, prevents the conversion of adenosine triphosphate to the second messenger, cyclic AMP (“cAMP”), which decreases cAMP levels. As HHS's analysis described, G proteins also contain beta/gamma G protein units that are also liberated following ligand binding, which then bind to and alter ion channel function, including inhibition of voltage-gated ion channels and activation of potassium channels. Ligand binding can also activate some subforms of phospholipase C as well as beta-arrestin protein. All of these second messenger routes amplify the neural signal following cannabinoid binding at the CB 1 and CB 2 receptors. HHS Basis for Rec. at 10.

CB 1 receptors are found primarily in the central nervous system (“CNS”), but are also present in peripheral tissues, such as the liver, heart, and lungs. In the brain, CB 1 receptors are expressed with highest density in the cortical regions, hippocampus, basal ganglia, and cerebellum and with lowest density in brainstem and hypothalamic areas. The localization of these receptors may explain cannabinoid effects on movement coordination, memory, and cognition. Additionally, CB 1 receptors are found in glial cells as well as in the immune system. However, the concentration of CB 1 receptors is considerably lower in peripheral tissues than in the CNS. CB 2 receptors are found primarily in the immune system, including in numerous leukocyte cell types, as well as in activated CNS microglia. Additionally, there is some evidence that CB 2 receptors are localized in the brain, primarily in the cerebellum and hippocampus. The distribution of CB 2 receptors throughout the body is less extensive than the distribution of CB 1 receptors. HHS Basis for Rec. at 10-11.

There are two endogenous cannabinoid receptor agonists: anandamide and arachidonyl glycerol (“2-AG”). At CB 1 receptors, anandamide is a partial agonist with low intrinsic efficacy while 2-AG is a full agonist with high intrinsic efficacy. These endogenous cannabinoid ligands are present in central as well as peripheral tissues. A combination of uptake and hydrolysis terminates the action of anandamide and 2-AG. The endogenous cannabinoid system is a locally active signaling system activated on demand in response to changes to the local conditions to help restore homeostasis. The endogenous cannabinoid system, including the endogenous cannabinoids and the cannabinoid receptors, demonstrates substantial plasticity in response to several physiological and pathological stimuli. This plasticity is particularly evident in the CNS. HHS Basis for Rec. at 11.

Δ9-THC and CBD have varying affinity and effects at the cannabinoid receptors. HHS determined that Δ9-THC is a partial agonist at both CB 1 (Ki = 18-218 nM) and CB 2 receptors (Ki = 36-309 nM). However, CB 1 receptors are the main pharmacological site of action for Δ9-THC, making CB 1 receptors the site that is responsible for the abuse potential of marijuana. The other CNS site where Δ9-THC may have activity is the 5HT 3 receptor, where it functions as an antagonist. In contrast, CBD has low affinity for both CB 1 and CB 2 receptors and may act as a negative allosteric modulator or weak antagonist at these sites. CBD has additional CNS effects as a serotonin 5HT 1A agonist and a serotonin 5HT 2A weak partial agonist, as well as a serotonin 5HT 3A antagonist. HHS Basis for Rec. at 11.

In the past 30 years, the potency of marijuana with regard to Δ9-THC has increased dramatically. HHS described one study finding that the concentration of Δ9-THC in marijuana samples in the United States increased from 3 percent in 1991 to 17.1 percent in 2017. These increases are likely due to an increase in the number of high potency samples ( i.e., sinsemilla) in the overall samples tested. Based on an evaluation of marijuana seized by DEA, the majority of samples contained high concentrations of Δ9-THC and low concentrations of CBD. HHS Basis for Rec. at 11-12.

Self-administration is a method that assesses the ability of a drug to produce rewarding effects. The presence of rewarding effects increases the likelihood that individuals will try to obtain additional quantities of a drug. Animal self-administration of a drug is often useful in suggesting whether humans will experience a particular substance as having rewarding effects, which is indicative of abuse potential. For example, the tendency of rhesus monkeys to self-administer a drug is correlated with humans' propensity to abuse it. HHS Basis for Rec. at 12.

Since self-administration is a methodology in which the test drug is typically administered intravenously to rats, it is not possible to evaluate botanical marijuana through self- administration. However, given that Δ9-THC is the primary substance that confers abuse potential to marijuana, its ability to induce self-administration can serve as an indicator of the abuse potential of marijuana. HHS Basis for Rec. at 12.

HHS concluded, after weighing the relevant scientific evidence, that Δ9-THC produces rewarding effects that lead an animal to repeatedly seek out the substance. HHS Basis for Rec. at 12. Specifically, some studies have demonstrated successful animal self-administration of Δ9-THC following intravenous administration, administration of inhaled vapor, oral administration, and intracerebroventricular administration. Other recent animal studies have not been able to produce Δ9-THC self-administration following intravenous administration and oral administration, but HHS concluded that these results were due to the specific methodology of those respective studies, rather than valid evidence of the rewarding effects of Δ9-THC, and thus do not negate HHS's reliance on studies in which Δ9-THC was actively self-administered by animals. HHS Basis for Rec. at 12-13.

Furthermore, a comprehensive deconstruction of which animal methodology is optimal for producing preclinical self-administration of Δ9-THC is not necessary for an evaluation of the abuse potential of marijuana in humans because it is already clear that humans utilize marijuana for its rewarding properties. HHS Basis for Rec. at 13. Animal self-administration is used primarily to predict whether a novel substance is likely to be used by humans for its rewarding properties as an indication of its abuse potential. However, epidemiological data already amply demonstrates that humans self-administer substances that contain Δ9-THC, including botanical marijuana, for their ability to produce positive subjective responses, including euphoria. HHS Basis for Rec. at 13; see also sections VI.4-6 of this preamble (discussing Factors 4-6).

A conditioned place preference (“CPP”) study is another method for determining whether drugs have rewarding properties; a CPP study relies on an animal's decision to spend time in a location associated with receiving a drug. The studies in which Δ9-THC successfully produced CPP occurred under very specific experimental conditions, similar to the Δ9-THC self-administration studies in animals. Experimental manipulations in CPP studies with Δ9-THC have included varying the animal species, sex, dose, or route of administration; introducing flavors to obscure unpleasant taste; and varying the drug history of the animals tested. However, as with animal self-administration, the purpose of CPP studies is typically to determine if a new drug produces rewarding sensations, which would suggest that a drug has abuse potential in humans. Since it is clear that humans self-administer substances that contain Δ9-THC, including botanical marijuana, HHS determined that it was not necessary to determine which CPP methods are optimal for demonstrating that Δ9-THC has rewarding properties in animals. HHS Basis for Rec. at 13.

Drug discrimination is a method in which animals indicate whether a test drug produces sensations similar to those produced by a training drug with a known pharmacological mechanism of action. Drug discrimination is considered to be an abuse-related study only when the training drug is a known drug of abuse that is scheduled under the CSA and the test drug may have abusable effects similar to the training drug based on having a similar mechanism of action to the training drug. Because animal drug discrimination studies often use Δ9-THC as the standard for establishing if new drugs have classic marijuana-like pharmacological activity, HHS did not examine whether this method should be applied when evaluating the abuse potential of Δ9-THC. HHS Basis for Rec. at 14.

The psychological, behavioral, and subjective responses to marijuana in humans have been known and characterized since antiquity. In the modern period, data on the psychological, behavioral, and subjective responses to marijuana are available from the drug labels of FDA-approved drug products, from prospective human abuse potential (“HAP”) studies, from accounts published in the scientific and medical literature, and from an evaluation published in 2017 by the National Academies of Science, Engineering, and Medicine (“NASEM”). HHS Basis for Rec. at 14.

Clinical scientific studies investigated the effects of Δ9-THC on humans during the development of the FDA-approved drug product Marinol, which contains 2.5, 5, and 10 mg dronabinol ( (−) - trans -Δ9-THC of synthetic origin in sesame seed oil). During controlled clinical trials (as reported in section 6.1 of the drug labels for Marinol and Syndros (which relied on the safety data from Marinol during drug development)), various adverse events (“AEs”) were observed, including amnesia, anxiety/nervousness, ataxia, confusion, depersonalization, hallucination, asthenia, palpitations, tachycardia, vasodilation/facial flush, euphoria, paranoid reaction, somnolence, abnormal thinking, dizziness, abdominal pain, nausea, and vomiting. HHS Basis for Rec. at 14-15.

HAP studies evaluate whether a test drug produces positive subjective responses compared to a placebo and a known drug of abuse that is scheduled under the CSA and serves as the positive control. If the test drug produces rewarding effects that are statistically significantly greater than the placebo, and beyond the acceptable placebo range of response, it is an indication that the drug may have abuse potential. The relative abuse potential is suggested by the responses from the positive control on these measures in comparison to the test drug. HHS Basis for Rec. at 15.

After analyzing a number of HAP studies of marijuana and Δ9-THC, which varied in the in dose of Δ9-THC, the route of administration, or whether the Δ9-THC was provided in the form of marijuana or isolated compound, HHS identified a number of commonalities. First, following administration of the study drug ( i.e., marijuana or Δ9-THC), there were increases in positive subjective responses. Second, the studies demonstrated increases on the Addiction Research Center Inventory scales for the morphine benzedrine group (euphoria), marijuana, and amphetamine. HHS concluded that these data consistently demonstrated that Δ9-THC, in the form of marijuana or as an isolated compound, produces rewarding effects that are indicative of abuse potential when it is administered under controlled experimental conditions. Third, and in contrast to the prior findings, the data also demonstrated that the administration of marijuana or Δ9-THC may result in negative subjective responses reflecting negative drug effects and sedation; these are often delayed in onset from when the positive subjective effects begin. HHS noted that the positive and negative subjective responses following administration of marijuana or Δ9-THC were often dose-dependent. It also noted that there were typically few differences between responses to marijuana and Δ9-THC, or between responses based on route of administration of the study drug. HHS Basis for Rec. at 15.

HHS concluded that the responses to dronabinol reported during development of Marinol and the responses to marijuana and Δ9-THC reported in HAP studies paralleled the common responses to marijuana that have been described by other medical scientists. These responses include positive subjective responses (such as euphoria or happiness), sedative responses (such as drowsiness or changes in sleep), anxiety and negative responses (such as panic attacks, agitation, and paranoia), perceptual changes (such as hallucinations and changes in perception), psychiatric, social, and cognitive changes (such as drug abuse, delusions, memory and concentration impairment, and impaired judgment), and physiological responses (such as nausea, tachycardia, facial flushing, dry mouth, tremor, dizziness, ataxia, and hyperemesis). The literature reviewed by HHS also concluded that the positive changes that occur following use of marijuana are pleasurable to many humans and are associated with drug-seeking and drug-taking; and that these effects are typically dose-dependent, with higher doses and routes of administration that produce faster onset producing more intense responses and the likelihood of more negative subjective effects. HHS Basis for Rec. at 16-17.

HHS also reviewed a book-length evaluation of marijuana by NASEM entitled The Health Effects of Cannabis and Cannabinoids: The Current State of Evidence and Recommendations for Research. 17 According to HHS, in this evaluation, NASEM provided a brief summary of the clinical features of marijuana intoxication and found that (1) during acute cannabis intoxication, the user's sociability and sensitivity to certain stimuli ( e.g., colors, music) may be enhanced, the perception of time is altered, and the appetite for sweet and fatty foods is heightened; (2) some users report feeling relaxed or experiencing a pleasurable rush or buzz after smoking cannabis; (3) these subjective effects were often associated with decreased short-term memory, dry mouth, and impaired perception and motor skills; and (4) when very high blood levels of Δ9-THC were attained, persons might experience panic attacks, paranoid thoughts, and hallucinations. HHS Basis for Rec. at 17-18.

Based on its analysis of the studies discussed above, HHS concluded that Δ9-THC, the substance largely responsible for the abuse potential of marijuana, is a partial agonist at the cannabinoid CB 1 receptor. When Δ9-THC is administered to animals, it produces rewarding responses, as evidenced by its ability to induce self-administration and CPP. This is consistent with the data from human studies and from clinical observations, where administration of Δ9-THC or use of marijuana produces euphoria and other pleasurable responses, as well as sedation and anxiety responses. Psychiatric, social, and cognitive responses, which are often experienced as negative, are also reported, as are physiological responses such as dry mouth, ataxia, and increased hunger. As described in HHS's analysis of Factor 4, see HHS Basis for Rec. at 32-37, the rewarding responses observed in humans are consistent with the prevalence of nonmedical use of marijuana, which includes abuse of the substance. Abuse of marijuana by individuals can lead to other negative consequences, including addiction and the need to seek medical attention through calls to poison centers or visits to an ED, as described in Factor 5, see HHS Basis for Rec. at 38-39, 42. HHS Basis for Rec. at 18.

DEA believes that additional data on marijuana's pharmacological effects may be appropriate for consideration in assessing this factor.

The third factor that DOJ and HHS must consider under 21 U.S.C. 811(c) is the state of current scientific knowledge regarding marijuana. In considering this factor and making its recommendation, HHS examined the chemistry of marijuana and the human pharmacokinetics of marijuana. HHS Basis for Rec. at 18-24.

Cannabis is a genus of annual flowering plant with digitate leaves in the family Cannabaceae Martinov that likely originated in Central or Southeast Asia over 10,000 years ago and was first cultivated in China for fiber and seed production. Cultivation eventually spread across Asia, Africa, and Europe and then to the Americas. A longstanding and significant historical debate by botanists and taxonomists continues today regarding the number of species in the Cannabis genus, but it is generally treated as a single, highly polymorphic species known as Cannabis sativa L., with the other two previously reported species listed as Cannabis indica Lam. and Cannabis ruderalis Janisch. Plants previously believed to be part of the latter two species are generally recognized as varieties (or subspecies) of Cannabis sativa L., which are commonly referred to as sativa var. indica and sativa var. ruderalis. Cannabis sativa and sativa var. indica plants are widely cultivated for their size, branching, and cannabinoid content, while sativa var. ruderalis is rarely cultivated alone because it is shorter, is often unbranched, and has very low cannabinoid content. Worldwide Cannabis varieties are separated into hundreds of different cultivars and strains. Plants selected for cultivation are known as cultivated varieties or cultivars, whereas plants reproduced asexually from a cultivar through clonal propagation are known as strains. These practices have resulted in significantly different chemical profiles for Cannabis cultivars, and the classification term to account for these chemical profile differences has evolved. The term “chemovar” accounts for the plant's chemical profile and is a more meaningful classification for clinical researchers studying the plant's potential drug effects. Marijuana products developed from diverse chemovars will have different safety, biological, pharmacological, and toxicological profiles. HHS Basis for Rec. at 18-19.

Cannabis is a dioecious plant, meaning female and male flowers occur on separate plants, and rarely occurs as a monoecious plant ( i.e., single plant containing male and female flowers). The glandular trichomes found on the female plant's unfertilized flower heads and bracts contain the highest concentrations of cannabinoids. For this reason, unfertilized female chemovars are favored to harvest large inflorescences ( i.e., complete flower head) for their rich cannabinoid and terpene content. HHS Basis for Rec. at 19.

The Cannabis sativa L. plant naturally contains many different compounds, and more than 550 have been identified, such as cannabinoids, terpenoids, flavonoids, stilbenoids, steroids, polysaccharides, benzoquinone, phenanthrenes, spiroindans, lignans, fatty acids, sugars, hydrocarbons, amino acids, and proteins. Cannabinoids are mainly found in living Cannabis sativa L. plants in their non-psychoactive carboxylated forms ( i.e., acid form), which require drying, heating, combustion, or aging to decarboxylate to their neutral forms, and are primarily composed of C 21 terpenophenolic compounds. The most abundant neutral form cannabinoids are Δ9-THC and CBD, but nearly 200 have been identified in the plant and are divided into subclasses: cannabigerols, cannabichromenes, CBDs, Δ9-THCs, (-)- Δ8- trans -tetrahydrocannabinols (“Δ8-THCs”), cannabicyclols, cannabielsoins, cannabinols, cannabinodiols, cannabitriols, and the miscellaneous cannabinoids. HHS Basis for Rec. at 19.

Like any other botanical substance, marijuana plants are heterogeneous in nature and contain a complex chemical profile. Moreover, variable organic plant material, as well as manufactured preparations, result in a variety of product forms that dictate different routes of administration, associated risks, and differences in quality of the product used, which may also influence risk for users. Among other things, these differences can result from differences in harvest location, growing conditions, the season in which the marijuana is harvested, and the manner in which the marijuana is processed, handled, transported, and tested. The potential for high variability of marijuana and marijuana-derived products, both in product composition and impurity profile, is a major consideration for the potential variability of drug effects and safety. HHS Basis for Rec. at 19-20.

Processing of marijuana and its use in further manufacturing can lead to a range of forms that individuals may use or consume, including crude mixtures and highly purified substances of botanical origin, many of which may be cannabinoid compounds. Among known cannabinoids in the Cannabis plant, both Δ9-THC and Δ8-THC produce marijuana's psychoactive effects. Because Δ9-THC is significantly more abundant than Δ8-THC, marijuana's intoxicating effects are largely attributed to the former. Only small quantities of Δ8-THC acid and Δ8-THC have been identified in plants. HHS Basis for Rec. at 20.

As noted above, the 2018 amendments to the CSA removed hemp from the definition of marijuana. However, the term “cannabis” is still often broadly used to refer to a wide variety of products manufactured from the Cannabis sativa L. plant, regardless of their control status. As a result of the 2018 amendments to the CSA, a large hemp marketplace exists, containing a wide variety of products. In addition, the public has access to cannabis products within the CSA definition of marijuana through State-authorized adult-use ( i.e., nonmedical use) and medical-use programs, as well as via the illicit marketplace. See HHS Basis for Rec. at 28-37 (Factor 4). Because of these diverse sources of marijuana, there is a lack of unified controls on cultivation and manufacturing, which raises concerns related to the safety, quality, and consistency of botanical substances ( e.g., botanical raw materials, extracts, and intermediates) and final product formulations that are currently accessed for medical and nonmedical use. Products sourced from State-authorized adult-use and medical-use programs are subject to a patchwork of inconsistent product standards and safety requirements. Although some State programs have a set of standards (for example, on manufacturing, testing, labeling, and packaging), each program's controls are different, leading to a wide variation of products across State-authorized programs. And the illicit marketplace is not subject to any standards or oversight. As a result, the range of products within the CSA's definition of marijuana encompasses a large degree of variation in forms for consumption, composition of biologically relevant constituents, potency, and contaminants. HHS Basis for Rec. at 21.

In short, marijuana has hundreds of chemovars containing variable concentrations of Δ9-THC, cannabinoids, and other compounds. As a result, in evaluating whether to recommend that marijuana be rescheduled, HHS focused to the greatest extent possible on wide-ranging substances derived from cannabis plants that are vehicles for the self-administration of Δ9-THC as the key biologically active substance on which the CSA's current definition of marijuana is based. HHS Basis for Rec. at 21.

HHS reported that the pharmacokinetics of Δ9-THC in humans— i.e., the study of how the body interacts with Δ9-THC—have been evaluated following inhaled administration of marijuana and oral administration of marijuana. These are the most frequently used routes of administration for marijuana or isolated Δ9-THC. HHS Basis for Rec. at 21.

Marijuana is commonly administered by humans via inhalation through smoking and, more recently, through vaping ( e.g., heating and inhalation of botanical matter or other volatile substances containing Δ9-THC). Generally, inhalation of a drug is the route that produces the fastest rate of drug absorption. Once marijuana is inhaled, Δ9-THC is absorbed through the lungs in the form of an aerosol within seconds. Peak plasma levels of Δ9-THC following inhalation occur very quickly, within 6 to 10 minutes. Psychoactive effects begin immediately following absorption, although peak subjective effects do not coincide with peak plasma Δ9-THC levels and are often delayed. Following administration of marijuana through inhalation, the bioavailability of Δ9-THC is 10 percent to 35 percent. That bioavailability is relatively low and varies widely due to several factors. An individual's experience and technique with smoking marijuana also determines the dose absorbed. HHS Basis for Rec. at 22.

When marijuana or Δ9-THC is administered orally (such as by eating marijuana-infused foods), the effects start within 30 to 90 minutes, reach their peak at 1.5 to 3 hours, and remain measurable for 4 to 12 hours. Oral bioavailability of Δ9-THC, following ingestion of an edible containing marijuana or isolated Δ9-THC, ranges from 5 to 20 percent. The low and variable bioavailability of Δ9-THC from oral ingestion is a consequence of its first-pass hepatic elimination from blood and erratic absorption from stomach and bowel. Ingestion of brownies containing marijuana also results in lower Δ9-THC plasma levels relative to inhalation of marijuana. HHS Basis for Rec. at 22-23.

Although there are differences in absorption of Δ9-THC depending on route of administration, the distribution, metabolism, and excretion of Δ9-THC is similar regardless of how the drug is administered. Plasma concentrations of Δ9-THC decrease quickly after absorption through rapid distribution into tissues and through liver metabolism. Because Δ9-THC has high lipophilicity, the apparent volume of distribution of Δ9-THC is high (10 L/kg) as it is distributed initially into organs such as lung, heart, brain, and liver that are highly perfused. Over time with regular exposure to marijuana, Δ9-THC will concentrate and be retained in fat. HHS Basis for Rec. at 23.

Metabolism of Δ9-THC occurs primarily via cytochrome P450 isozymes (CYP2C9, CYP2C19, and CYP3A4) via microsomal hydroxylation to both active and inactive metabolites. The primary active metabolite of Δ9-THC is 11-hydroxy-Δ9-THC. Δ9-THC clears from the blood relatively rapidly, largely because it is redistributed to other tissues in the body. Metabolism of Δ9-THC in most tissues is relatively slow or absent. The majority of the absorbed Δ9-THC dose is eliminated in feces, and about 33 percent in urine. HHS Basis for Rec. at 23.

In conclusion, HHS found that the pharmacokinetic profile of marijuana varies greatly depending on route of administration. Inhalation of marijuana produces a rapid increase in plasma levels of Δ9-THC and an immediate onset of psychological effects. In comparison, oral administration of marijuana produces a much slower increase in plasma levels of Δ9-THC and onset of psychological effects. Once Δ9-THC has been absorbed, however, the metabolism and excretion of Δ9-THC follows a standard path. HHS Basis for Rec. at 24.

DEA likewise notes that there is considerable variability in the cannabinoid concentrations and chemical constituency among marijuana samples and that the interpretation of clinical data related to marijuana is complicated. A primary issue is the lack of consistent concentrations of Δ9-THC and other substances in marijuana, which complicates the interpretation of the effects of different marijuana constituents. Additionally, the non-cannabinoid components in marijuana may potentially modify the overall pharmacological and toxicological properties of various marijuana strains and products. DEA anticipates that additional data on other marijuana constituents, routes of administration of marijuana, and the impact on Δ9-THC potency may be appropriate for consideration.

The fourth factor that DOJ and HHS must consider under 21 U.S.C. 811(c) is marijuana's history and current pattern of abuse, which can include its abuse relative to relevant comparator substances that are abused. See HHS Basis for Rec. at 28-37. HHS concluded that it is appropriate to consider the Federal- and State-level history of marijuana control, marijuana sources for nonmedical and medical use, marijuana use in the United States since passage of the CSA, and current patterns of use and abuse of marijuana. HHS Basis for Rec. at 28.

According to HHS, marijuana was described in the United States Pharmacopoeia  18 as early as 1850. Around the time that Congress passed the Pure Food and Drug Act of 1906, Public Law 59-384, 34 Stat. 768, drugs such as marijuana, alcohol, heroin, morphine, and cocaine began to be characterized by the Federal Government as addictive and dangerous. These drugs were frequently included in patent medicines, often without the consumer's knowledge. The 1906 law required accurate drug labeling with respect to ingredients and dosage. But it did not prohibit the sale or possession of drugs characterized as addictive and dangerous drugs, including marijuana. As nonmedical use of marijuana and opioids became more popular in the United States, Congress provided funding in 1929 for two “narcotic farms” in Lexington, Kentucky, and Fort Worth, Texas, which were medical treatment centers run by the Public Health Service for federal prisoners who were “habitual users of narcotics,” including marijuana-derived products. HHS Basis for Rec. at 28-29.

In the first half of the twentieth century, marijuana use was curbed by several Federal laws. In 1931, the importation of marijuana into the United States began to be restricted under regulations under the Pure Food and Drug Act, except for medicinal purposes. The Marihuana Tax Act of 1937, Public Law 75-238, 50 Stat. 551, imposed taxes that effectively prohibited marijuana use for medical, nonmedical, scientific, or industrial purposes. Five years later, in 1942, marijuana was removed from the United States Pharmacopoeia. Through the imposition of mandatory minimums, the Boggs Act of 1951, Public Law 82-255, 65 Stat. 767, lengthened the average sentence for first time marijuana offenders to 2 to 5 years, similar to that for opioid offenses, regardless of whether the individual was a nonmedical user or a trafficker. The Narcotic Control Act of 1956, Public Law 84-728, 70 Stat. 567, increased the minimum sentence for a first offender for marijuana to 2 to 10 years. HHS Basis for Rec. at 29.

Despite the legal consequences, nonmedical marijuana use increased dramatically in the 1960s, especially among youth. Congress passed the CSA in 1970. The CSA effectively repealed all previous Federal drug laws, including the Marihuana Tax Act, and provided a unified framework for control of drugs with abuse potential. When the CSA was enacted, marijuana was placed into schedule I, which prohibited use of marijuana for medicinal or nonmedical purposes other than legitimate scientific research and analysis. This placement was consistent with the criteria established by the CSA under 21 U.S.C. 812. HHS Basis for Rec. at 29-30.

According to HHS, changes in State-level marijuana laws in the United States in the modern era began in 1996 with the approval of Proposition 215, the Compassionate Use Act, by voters in California. This law legalized the use, possession, and cultivation of marijuana for treatment of patients with cancer, anorexia, AIDS, chronic pain, spasticity, glaucoma, arthritis, migraine, or any other illness for which marijuana provides relief, when recommended by a physician. Under the law, marijuana could also be cultivated by patient caregivers. HHS Basis for Rec. at 30.

As of August 2023, when HHS submitted its Basis for Recommendation to DEA, State-level laws allowing medicinal use of marijuana had been passed in a total of 38 States, plus the District of Columbia: Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Delaware, District of Columbia, Florida, Hawaii, Illinois, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Michigan, Minnesota, Mississippi, Missouri, Montana, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Dakota, Utah, Vermont, Virginia, Washington, and West Virginia. Medical use of marijuana was legalized through the action of 20 State legislatures and by 18 ballot measures. 19 HHS Basis for Rec. at 30.

In 2012, Colorado and Washington became the first States to legalize the nonmedical use of marijuana. As of August 2023, State-level legalization of the nonmedical use of marijuana has occurred in a total of 23 States and the District of Columbia: Alaska, Arizona, California, Colorado, Connecticut, Delaware, District of Columbia, Illinois, Maine, Maryland, Massachusetts, Michigan, Minnesota, Missouri, Montana, Nevada, New Jersey, New Mexico, New York, Oregon, Rhode Island, Vermont, Virginia, and Washington. Nonmedical use of marijuana was legalized by ballot initiatives in 13 States and by State legislatures in 9 States. HHS Basis for Rec. at 30.

Marijuana use has varied since the CSA was passed in 1970. Gallup Poll data from 1969 to 2013 show a steady increase over time in affirmative responses to whether the respondent had personally tried marijuana, with only 4 percent of people saying they had tried marijuana in 1969 compared to 38 percent in 2013. As HHS observed, the 2017 NASEM report stated that the prevalence of marijuana use peaked in the late 1970s, declined through the 1980s, and then increased again in the mid-1990s. From 2007 to 2017, there were steady year-over-year increases in the share of the general population that used marijuana in the past month, although there is no clear explanation for the post-2007 increase in use rates. HHS Basis for Rec. at 31-32.

In considering current patterns of use and abuse of marijuana and marijuana-derived products, HHS analyzed epidemiological databases from 2015 to the most recent years of available data (which vary among data sources). A wide variety of epidemiological databases provide necessary data for HHS's analyses. These include the NSDUH; Behavioral Risk Factor Surveillance System (“BRFSS”); Research Abuse, Diversion and Addiction-Related Surveillance (“RADARS”); Nonmedical Use of Prescription Drugs (“NMURx”); Monitoring the Future (“MTF”); Youth Risk Behavioral Surveillance System (“YRBSS”); and International Cannabis Policy Study (“ICPS”). HHS Basis for Rec. at 32.

Based on NSDUH data, HHS concluded that from 2015 to 2019 the past-year use of marijuana for any reason (nonmedical and medical) among people ages 12 years and older increased from 14 percent to 18 percent. By contrast, past-year (nonmedical and medical) use of comparator drugs that have FDA-approved therapeutic indications declined or remained relatively stable over the same timeframe, including hydrocodone (22 percent to 16 percent), benzodiazepines (12 percent to 11 percent, 2017 to 2019 only), oxycodone (11 percent to 9 percent), tramadol (7 percent to 6 percent), zolpidem (4 percent to 3 percent), and ketamine (less than 1 percent). Although there were trend breaks for the years 2020 and 2021, 20 marijuana past-year use continued to increase during these two years. HHS Basis for Rec. at 32-33.

Based on NSDUH data, HHS concluded that from 2015 to 2019, the prevalence of past-year nonmedical use of marijuana ( i.e., use without a health care provider (“HCP”) recommendation) among people ages 12 years and older also increased. HHS's finding was based on an increase in the prevalence of overall nonmedical use of marijuana from 12 percent to 15 percent and on an increase in nonmedical use of marijuana only, without nonmedical use of other drugs that are abused, from 8 percent to 11 percent during this period. There was a slight decrease in both categories in 2020, but the prevalence of both kinds of uses increased again in 2021 (to 16 percent and 11 percent, respectively) to levels that were higher than those reported in 2019. In contrast, the prevalence of past-year nonmedical use of comparator drugs was less than 3 percent for heroin, cocaine, oxycodone, hydrocodone, tramadol, benzodiazepines, and zolpidem, which is much less than that for marijuana, either alone or with other drugs. Over the 2015 to 2021 reporting period, the overall use of these comparator drugs declined slightly or remained fairly stable. Notably, the majority of individuals who reported nonmedical use of marijuana did not report nonmedical use of the comparator drugs. And over the same reporting period of 2015 to 2021, the prevalence of past-year use of alcohol ranged from 62 percent to 65 percent for individuals ages 12 years and older, far exceeding the prevalence for marijuana or other comparator drugs. These data demonstrate that alcohol has the highest prevalence of past-year-only use, followed by nonmedical use of marijuana. The prevalence of the other comparators is far below that of alcohol and marijuana. HHS Basis for Rec. at 33.

HHS also concluded that the NSDUH data show that most individuals who used marijuana in the past year did not do so based on a recommendation from an HCP, but marijuana use was more frequent among users with an HCP recommendation. The yearly percentage of individuals who used marijuana but did not have an HCP recommendation ranged between 84 and 89 percent between 2015 and 2021; by comparison, exclusive medical use of marijuana that was recommended by an HCP ranged between 7 and 10 percent of marijuana users in the same period. According to HHS, approximately 50 percent of those individuals without an HCP recommendation used marijuana for 60 or fewer days in the year, while 29 percent used marijuana for more than 241 days in the year. In contrast, for those individuals whose use of marijuana was sometimes or always recommended by an HCP, 51 percent and 55 percent (respectively) used marijuana at least 241 days in the year. HHS Basis for Rec. at 33-34.

The NSDUH data from 2021 showed that among individuals who used any marijuana in the past year, 69 percent used marijuana in the prior month. For comparator drugs, the percentage of individuals with past-year use who used each substance nonmedically in the past month was 76 percent for alcohol, 49 percent for heroin, 38 percent for cocaine, and 28 percent for ketamine. HHS Basis for Rec. at 34.

BRFSS is a national, State-based, cross-sectional telephone survey conducted by the Centers for Disease Control and Prevention (“CDC”). The participants in the 2021 BRFSS module for marijuana included approximately 68 million individuals 18 years and older, residing in 24 States and Territories: Alaska, Connecticut, Delaware, Hawaii, Idaho, Illinois, Indiana, Kentucky, Maine, Maryland, Minnesota, Montana, Nebraska, Nevada, New Hampshire, New York, North Dakota, Ohio, Oklahoma, Rhode Island, Utah, Vermont, Wyoming, and Guam. HHS Basis for Rec. at 34.

For the 2021 survey year, the estimated prevalence of past-month marijuana use for any reason in the BRFSS survey was 12 percent, with 88 percent reporting no marijuana use. Among those with past-month marijuana use, the mean frequency of use was 17 days per month, with half of respondents reporting that they used marijuana 20 to 30 days per month. This pattern was consistent across all age and sex categories. HHS Basis for Rec. at 34.

When the reason for use was evaluated, the percentage of individuals who reported use for both medical and nonmedical reasons was 39 percent, compared to 36 percent for those who reported use for nonmedical reasons only, and 25 percent for those who reported use for medical reasons only. Those individuals who reported past-month use of marijuana for medical reasons were more likely to be adults 55 years and older, while individuals who reported past-month marijuana use for nonmedical reasons only were more likely to be younger adults aged 18 to 24 years. HHS Basis for Rec. at 34.

Individuals who reported using marijuana in the past 30 days for both nonmedical and medical reasons were more likely (62 percent) to report marijuana use near daily (20 to 30 days per month) than individuals who reported marijuana use for nonmedical reasons only (34 percent). Similarly, individuals who used marijuana for medical reasons only were also more likely (57 percent) to report near daily use than those who used it for nonmedical reasons only. HHS Basis for Rec. at 34.

The RADARS System conducts the NMURx Program, a serial, cross-sectional, online survey of the general adult population (18 years and older) to elicit information on the nonmedical use of drugs (prescription, nonprescription, unapproved, and illicit). The NMURx Program estimates represent measures of past-year drug use in an enriched sample of United States adults with higher-than-average nonmedical use of prescription pain relievers and illicit drugs. NMURx program data demonstrated that past-year use of marijuana was reported by 21 percent of individuals, while past-year use of comparator substances was substantially lower: benzodiazepines (4 percent), hydrocodone, oxycodone, tramadol (2 percent), cocaine or crack (less than 2 percent), and illicit fentanyl, heroin, and ketamine (less than 1 percent). This pattern of much greater marijuana use compared to other drugs is consistent with the patterns reported in NSDUH and BRFSS. HHS Basis for Rec. at 35.

MTF collects information on the use of selected prescription and illicit drugs and alcohol by conducting an annual, nationally representative, cross-sectional survey of 8th, 10th, and 12th graders in public and private schools. 21 MTF data showed that during the years 2012 to 2022, the illicit drug most frequently used by 12th-grade students who reported past-year drug use was marijuana/hashish (approximately 35 percent per year from 2012 to 2020, with a reduction to 30 percent per year in 2021 and 2022). In contrast, in 2022, alcohol was used by 52 percent of 12th-grade students within the last 12 months, similar to percentages in 2019 and 2020 (52 percent and 55 percent, respectively), but higher than the 2021 level of 47 percent. All other comparator drugs (hydrocodone, heroin, tramadol, cocaine, ketamine, and zolpidem) were each used in the past year by fewer than 5 percent of 12th graders from 2012 to 2022. HHS Basis for Rec. at 35.

MTF data for past-month use showed a similar pattern. During the years 2012 to 2022, the illicit drug most frequently used by 12th-grade students who reported past-month drug use was marijuana/hashish (approximately 20 to 22 percent per year) compared to past-month use of cocaine (approximately 1 percent per year) or heroin (less than 0.5 percent per year). However, past-month alcohol use by 12th-grade students (28 percent) exceeded that of marijuana in 2022. For those who used marijuana in the past month, 6 to 7 percent used it daily. By comparison, for those who used cocaine and heroin in the last month, less than one percent used it daily. MTF does not provide past-month use data for hydrocodone, heroin, tramadol, ketamine, or zolpidem. HHS Basis for Rec. at 35.

YRBSS was established by the CDC and conducts school-based surveys every 2 years, in partnership with State, local, Territorial, and Tribal governments, with a focus on youth health behavior in the United States. The YRBSS high school component, the Youth Risk Behavior Survey, includes a nationally representative survey of 9th- through 12th-grade students. YRBSS data showed that from 2009 to 2019, approximately 20 percent of students in 9th through 12th grade reported using marijuana at least once in the past month during each year evaluated. When students 17 years and older were asked how old they were when they first used marijuana, 43 percent reported they initiated use between the ages of 15 to 16 years, 25 percent initiated use between the ages of 13 to 14 years, and 13 percent initiated use at 12 years of age and younger. YRBSS data also showed, however, that past-month alcohol use by high school students (29 percent) in 2019 was greater than that of marijuana use, while past month prescription opioid misuse (including codeine, hydrocodone, or oxycodone) (7 percent) in 2019 was much lower than that of both alcohol and marijuana use. HHS Basis for Rec. at 36.

ICPS conducted serial, cross-sectional surveys from 2019 to 2021 of individuals ages 16 to 65 years living in the United States to understand the public health impact of marijuana legalization. HHS's evaluation of that survey data focused on respondents who reported at least some past-year marijuana nonmedical use (by indicating that they were not a medical marijuana user, defined as someone who uses marijuana only to treat a medical condition). HHS Basis for Rec. at 36.

According to HHS, ICPS data showed that the prevalence of past-year nonmedical use of marijuana ranged from 18 percent to 22 percent of individuals surveyed from 2019 to 2021, while the prevalence of past-month nonmedical use was lower, ranging from 12 percent to 14 percent of individuals surveyed. Individuals aged 26 to 34 years had the highest relative prevalence of nonmedical marijuana use, with 26 percent reporting past-year use and 18 percent reporting past-month use. When those individuals who reported past-year marijuana use in 2021 were asked why they used the drug, 33 percent reported use for medical reasons, while 61 percent were classified as using marijuana for nonmedical reasons only. (The other 6 percent did not respond.) HHS Basis for Rec. at 36.

When frequency of nonmedical use of marijuana was evaluated in ICPS for those individuals who used marijuana nonmedically at least once a year, individuals aged 16 to 17 years had the highest percentage of use less than once a month (approximately 40 percent, compared to approximately 25 to 31 percent for other age cohorts); while individuals aged 26 to 34 years had the highest percentage of daily use (approximately 43 percent, compared to approximately 34 to 37 percent for individuals in other adult cohorts and approximately 24 percent among individuals 16 and 17 years). Among individuals who used marijuana for nonmedical reasons in the past year, 49 percent reported never using alcohol and marijuana at the same time, while 35 percent sometimes used the two substances together, 9 percent often used them together, and 5 percent used alcohol every time they used marijuana. HHS Basis for Rec. at 36-37.

In light of the evidence cited above, HHS determined that certain conclusions could be drawn about marijuana's current pattern of abuse. HHS concluded that the use of marijuana for medical and nonmedical purposes is extensive in the United States. HHS also concluded that the prevalence of marijuana use is less than that of alcohol and significantly more than that of other drugs of abuse that are scheduled under the CSA. Specifically, HHS noted that NSDUH data from 2015 to 2019 showed that the prevalence of past-year use of alcohol was five to six times greater than that of nonmedical use of marijuana. In contrast, the prevalence of past-year nonmedical use of heroin, cocaine, oxycodone, hydrocodone, tramadol, benzodiazepines, and zolpidem was four to five times less than that of marijuana nonmedical use. Similar past-year comparative drug use data were reported in RADARS-NMURx, MTF, and ICPS. HHS Basis for Rec. at 37. In 2016, DEA found that marijuana continues to be the most widely used illicit drug. It noted that in 2014, there were 22.2 million current users; that there were also 2.6 million new users, most of whom were less than 18 years of age; and that marijuana was the most frequently identified drug identified in Federal, State, and local forensic laboratories. 81 FR 53739. In addition to the data provided in the HHS Basis for Recommendation and the data considered by HHS and DEA in their 2015 eight-factor analyses, DEA anticipates that additional information arising from this rulemaking will further inform the findings that must be made to reschedule marijuana, including with respect to this factor. DEA also notes that, according to the World Health Organization, cannabis is globally the most commonly used psychoactive substance under international control. 22 Accounting for half of all drug seizures worldwide, the global annual prevalence of cannabis consumption is 2.5 percent or about 147 million people. 23 In 2016, an estimated 28.6 million individuals age 12 or older were current (in the past month) illicit drug users. 24 By 2020, approximately 59.3 million individuals age 12 or older reported using an illicit drug within the past year; 83.6 percent (49.6 million) of those past-year illicit drug users reported using marijuana. 25 In 2022, the Domestic Cannabis Eradication and Suppression Program was responsible for the eradication of 4,435,859 illegally cultivated outdoor cannabis plants and 1,245,980 illegally cultivated indoor plants for a total of 5,681,839 illegally cultivated marijuana plants. 26 DEA believes that additional data on marijuana's pattern of abuse may be appropriate for consideration in assessing this factor.

The fifth factor that DOJ and HHS must consider under 21 U.S.C. 811(c) is the scope, duration, and significance of marijuana abuse. In conducting its analysis, HHS analyzed the consequences over time of marijuana abuse compared to the abuse of other substances based on data from the United States Poison Centers National Poison Data System (“NPDS”), NSDUH, the Treatment Episode Data Set (“TEDS”), the National Addictions Vigilance Intervention and Prevention Program (“NAVIPPRO”), the National Emergency Department Sample (“NEDS”), the National Inpatient Sample (“NIS”), and the National Forensic Laboratory Information System (“NFLIS”). HHS Basis for Rec. at 37-45.

Data from NPDS provide information on the scope of contacts with a poison center (“PC”) following marijuana abuse relative to abuse of selected comparators. HHS Basis for Rec. at 38.

The number of PC abuse cases for a substance (either alone or in combination with another substance) for the period of 2015 to 2021 showed that the highest number of PC abuse cases was for alcohol, followed by heroin and then benzodiazepines. The fourth highest number of PC abuse cases was for marijuana, with all other comparators showing fewer PC abuse cases. When the PC abuse cases for 2015 to 2021 were analyzed for cases involving a single substance only, the rank order of PC abuse cases by number was the same as the order from all PC abuse cases for substances used alone or in combination with another substance, meaning that marijuana accounted for the fourth highest number of PC abuse cases for a single substance. HHS Basis for Rec. at 38.

HHS's analysis of the data from 2015 to 2021 showed cases resulting from abuse (as opposed to those resulting from other causes, such as accidental ingestion) made up the largest proportion of PC cases for illicit fentanyl (72 percent), heroin (65 percent), cocaine (41 percent) and ketamine (40 percent). The fifth highest percentage was for cases involving marijuana (36 percent), followed by alcohol (15 percent), oxycodone (13 percent), benzodiazepines (8 percent), hydrocodone (5 percent), tramadol (4 percent), and zolpidem (3 percent). A similar analysis for single-substance-only abuse for the same period showed that the three substances most likely to lead to a PC call following abuse were heroin (65 percent), oxycodone (47 percent), and tramadol (47 percent). The fourth highest percentage was for marijuana and ketamine (46 percent), followed by alcohol (43 percent), zolpidem (40 percent), hydrocodone (37 percent), illicit fentanyl (34 percent), benzodiazepines (32 percent), and cocaine (28 percent). HHS Basis for Rec. at 38.

Annual utilization-adjusted abuse case rates were calculated by dividing the number of PC abuse case counts by the prevalence of past-year use based on NSDUH estimates from people aged 12 years and older, for the period 2015 to 2019, for both (1) any past-year use of the substance and (2) past-year nonmedical use of the substance. These utilization-adjusted rates convey the likelihood that use of a drug will result in PC abuse cases when considering how many people use the drug for either (1) any reason or (2) nonmedical reasons. The utilization-adjusted abuse rates for any past-year use of a substance showed the highest rate for heroin (increasing from 4,038 to 7,201 cases per one million people). The next highest rates were for ketamine, cocaine, and benzodiazepines; all these rates were considerably lower than the rate for heroin. The rates for marijuana (relatively stable at 75 to 70 cases per one million people) and oxycodone were similar, as were the rates for alcohol, zolpidem, tramadol, and hydrocodone; all these rates were considerably lower than the rates for ketamine, cocaine, and benzodiazepines. A similar pattern of utilization-adjusted abuse rates was seen among cases involving a single substance only during the same time period. HHS Basis for Rec. at 39.

An analysis of medical outcomes related to exposure based on severity, timing, and assessment of clinical effects for all single-substance PC abuse cases involving marijuana or comparator drugs showed that serious medical outcomes (moderate effect, major effect, or death) were greatest for illicit fentanyl (81 percent) and heroin (79 percent), followed by oxycodone (70 percent), ketamine (64 percent), tramadol (62 percent), cocaine (59 percent), hydrocodone (44 percent), marijuana (41 percent), benzodiazepines (32 percent), alcohol (31 percent), and zolpidem (27 percent). HHS noted that death rates are underreported in NPDS, but HHS observed that the highest death rate was for fentanyl (25 percent); cocaine, heroin, and alcohol had comparatively very low death rates (3 percent, 2 percent, and 2 percent, respectively), with all other comparators reporting death rates of less than 1 percent. HHS Basis for Rec. at 39-40.

Data from NSDUH provide nationally representative information on the prevalence of substance use disorder (“SUD”) in 2021 among individuals aged 12 years or older who reported nonmedical use of marijuana in past year in comparison to heroin, cocaine, or alcohol use in the past year. A diagnosis of SUD is made when an individual endorses at least 2 of the 11 criteria for SUD according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (“DSM-V”). Individuals are classified with a mild SUD if they meet two to three of the criteria, a moderate SUD if they meet four to five of the criteria, and a severe SUD if they meet six or more of the criteria. HHS Basis for Rec. at 40.

NSDUH data showed that, among individuals with past-year heroin use in 2021, there was an 81 percent prevalence of meeting the criteria for a heroin SUD. In comparison, there was a 30 percent prevalence of meeting the criteria for marijuana SUD among individuals who used marijuana for nonmedical reasons only (17 percent mild, 8 percent moderate, and 5 percent severe). For individuals who used marijuana for nonmedical purposes and did not use other drugs illicitly, there was a slightly lower prevalence (24 percent) of meeting the criteria for SUD (15 percent mild, 6 percent moderate, and 3 percent severe). For cocaine, 30 percent of individuals who used cocaine in the past year met criteria for cocaine SUD (13 percent mild, 5 percent moderate, and 12 percent severe). For individuals who used alcohol in the past year, the prevalence of alcohol SUD was 17 percent (10 percent mild, 4 percent moderate, and 3 percent severe). HHS Basis for Rec. at 40.

Although the 2021 NSDUH data showed that the likelihood of meeting the criteria for a SUD was highest for heroin, followed by marijuana, cocaine, and alcohol, the absolute number of individuals who met the criteria had a different order. Alcohol had the highest number of such individuals (approximately 29,544,000), followed by marijuana (approximately 13,078,000 people with marijuana nonmedical-only use, and approximately 7,454,000 with nonmedical-only use and no nonmedical use of other drugs), cocaine (approximately 1,408,000), and heroin (approximately 894,000). HHS Basis for Rec. at 40.

TEDS is a database run by SAMHSA within HHS that presents information on the demographic and substance use characteristics of annual admissions for treatment for alcohol and drug abuse in State-approved facilities that are required by the States to provide TEDS client-level data. Because TEDS is based only on reports from these facilities, TEDS data do not represent the total national demand for substance abuse treatment or the prevalence of substance abuse in the general population. HHS Basis for Rec. at 40-41.

Out of 1.4 million admissions documented in the 2020 TEDS dataset, the most frequently reported primary drug of admission was alcohol (31 percent, or 442,014 admissions), followed by heroin (21 percent, or 292,126 admissions), marijuana (10 percent, or 139,481 admissions), and cocaine (5 percent, or 71,725 admissions). Other comparator drugs were each reported as the primary drug in less than 2 percent of admissions. Over the reporting period of 2015 to 2020, the proportion of admissions each year ranged from 30 to 33 percent for alcohol; from 21 to 26 percent for heroin; from 10 to 14 percent for marijuana; and from 5 to 6 percent for cocaine. The proportion of admissions with marijuana as the primary drug declined each year from 14 percent in 2015 to a low of 10 percent in 2020, while the proportion of admissions with cocaine as the primary drug increased slightly during this time from 5 percent in 2015 to 6 percent in 2019. During this reporting period, other comparator drugs were each reported as the primary drug in less than 2 percent of admissions each year. HHS Basis for Rec. at 41.

In 2020, marijuana and cocaine were most likely to be reported as the secondary drug at admission (25 percent and 24 percent, respectively), followed by alcohol (15 percent), heroin (8 percent), and benzodiazepines (6 percent), with all other comparators reported as less than 2 percent. For tertiary drugs at admission, marijuana (29 percent) was reported most frequently, followed by cocaine (18 percent), alcohol (16 percent), and heroin (5 percent), with all other comparators reported as less than 2 percent. HHS Basis for Rec. at 41.

NAVIPPRO is a surveillance system for substance use and nonmedical use of prescription medication in a convenience sample of adults seeking treatment or being assessed for SUD treatment at participating facilities across the United States. NAVIPPRO Addiction Severity Index-Multimedia Version (“ASI-MV”) is a clinical assessment tool that collects data on recent drug use behaviors for evaluation and treatment planning at intake. From 2020 through 2021, there were a total of 76,249 NAVIPPRO ASI-MV assessments in individuals entering or being assessed for SUD treatment at a center participating in the NAVIPPRO network. The drug most frequently endorsed for past-month use was marijuana (20,458 individuals, or 27 percent), followed by alcohol (5 or more alcoholic drinks per day, 16,388 individuals, or 22 percent), heroin (9,078 individuals, or 16 percent), fentanyl (6,186 individuals, or 8 percent), hydrocodone (3,448 individuals, or 5 percent), oxycodone (3,186 individuals, or 4 percent), cocaine or crack (5,417 individuals, or 7 percent), tramadol (543 individuals, or 1 percent), and ketamine (169 individuals, or less than 1 percent). HHS Basis for Rec. at 41.

NEDS is the largest all-payer ED database in the United States, as developed for HHS's Agency for Healthcare Research and Quality (“AHRQ”). NEDS is a sample of records from ED visits from the State Emergency Department Databases, which capture discharge information on all ED visits that do not result in hospital admission, and the State Inpatient Databases, which contain information on patients first seen in the ED and then admitted. The 2020 ED sample covered 995 hospital EDs and 41 States; the unweighted 2020 sample contained data from over 28 million ED visits, which resulted in a weighted estimate of 123 million ED visits. HHS compared ED visits that noted an alcohol, marijuana, or cocaine-related disorder; this comparison included ED visits not directly due to a specific substance-related disorder, but in which the patient was recorded as having had an alcohol, marijuana, or cocaine-related disorder in the administrative claim associated with the visit. HHS Basis for Rec. at 42.

Based on NEDS data, from 2016 to 2020, the highest estimated number of annual ED visits was for an alcohol-related disorder, with between 4 million and 4.1 million visits each year, 3.2 million of which involved alcohol as a single substance. Over the same timeframe, estimated annual ED visits involving a marijuana-related disorder ranged from approximately 1.3 million to over 1.7 million, with the estimated annual ED visits for single-substance marijuana disorder ranging from 757,731 to 1.08 million. For cocaine, the estimated annual ED visits involving a related disorder were between 559,165 and 774,737, with annual visits for single-substance cocaine-related disorder ranging from 204,257 to 266,614. HHS Basis for Rec. at 42.

HHS calculated a utilization-adjusted rate of estimated ED visits, and the highest rate was for cocaine-related disorder, which ranged from 11,765 to 14,014 visits per 100,000 individuals, of which 4,011 to 4,952 were single-substance visits. Marijuana had the second-highest utilization-adjusted rate of estimated ED visits, ranging from 3,472 to 3,940 per 100,000 individuals 2,017 to 2,413 of which were single-substance visits. The utilization-adjusted rate of visits involving an alcohol disorder, the lowest of the three substances, ranged from 2,225 to 2,327 per 100,000 individuals, of which 1,775 to 1,843 were single-substance ED visits. HHS Basis for Rec. at 42-43.

NIS is the largest publicly available all-payer inpatient administrative health care database in the United States, and it is sponsored by AHRQ. It is a sample of discharges from participating community hospitals from 46 to 48 States and the District of Columbia each year, with approximately 7 million unweighted inpatient stays annually, accounting for weighted annual estimates of 35 million hospitalizations. HHS used NIS data to compare alcohol, marijuana, and cocaine. HHS Basis for Rec. at 43.

From 2016 to 2020, alcohol-related disorder had the highest estimated annual number of hospitalizations, at approximately 1.8 million each year, of which approximately 1.2 to 1.25 million involved single-substance alcohol-related disorder. Marijuana-related disorder had the second-highest estimated annual number of hospitalizations, increasing from 795,140 in 2016 to 914,810 in 2020, of which 373,160 to 452,985 were for single-substance marijuana-related disorder. Cocaine had the lowest estimated annual number of hospitalizations, ranging from 387,385 to 453,955, of which 94,695 to 112,725 were for single-substance cocaine-related disorder. HHS Basis for Rec. at 43.

HHS then calculated a utilization-adjusted rate of estimated hospitalizations, and the highest rate was for cocaine-related disorder, which ranged from 7,185 to 8,211 hospitalizations per 100,000 individuals with any past-year use, of which 1,796 to 2,039 were single-substance hospitalizations. Marijuana-related disorder had the second-highest rate of estimated hospitalizations, ranging from 1,850 to 2,117 per 100,000 individuals, of which 906 to 1,026 were single-substance hospitalizations. Alcohol had the lowest rate, ranging from 987 to 1,039 per 100,000 individuals, of which 675 to 715 were single-substance hospitalizations. HHS Basis for Rec. at 43.

NFLIS is a program of the Diversion Control Division of DEA. The NFLIS-Drug system is a component of the NFLIS that contains data that serve as a surveillance resource to monitor drug encounters by law enforcement across the United States, including data on drugs seized by law enforcement and submitted to Federal, State, and local forensic laboratories for analysis. In NFLIS, a law enforcement investigation (“case”) may result in one or more “reports” or “exhibits” of drug evidence, and each report or exhibit may contain one drug or multiple drugs. However, NFLIS-Drug data has limitations because not all drugs encountered by law enforcement are sent for analysis and not all drugs sent to reporting forensic laboratories are tested. To account for nonreporting laboratories, among other things, DEA publishes NFLIS-Drug national report estimates annually and semiannually. Analyzing national estimates data allows for a comparison of the number of reports by year and reporting trends. In calculating national and regional estimates, DEA uses all NFLIS-Drug reporting laboratories. HHS Basis for Rec. at 43-44.

In 2021, there were 1,326,205 drug reports from State and local forensic laboratories in the United States, an increase of 3 percent from 2020. Nationally, 61 percent of all drug reports in NFLIS were identified as involving methamphetamine (406,200 reports or 31 percent), cannabis/THC (167,669 reports or 13 percent), cocaine (165,162 reports or 12 percent), or heroin (72,315 reports or 5 percent). HHS Basis for Rec. at 44-45.

In 2021, there were 1,027,219 drug-specific cases submitted to and analyzed by State and local laboratories, a 2 percent increase from 2020. Although the total NFLIS-Drug number of drug reports increased in 2021 from 2020, the total number of cases and drugs reported continues to be noticeably lower than the numbers reported for the years before the COVID-19 pandemic. Nationally, in 2021, 45 percent of all drug cases contained one or more reports of methamphetamine, followed by cocaine (18 percent), cannabis/THC (17 percent), and heroin (8 percent). Nationally, the number of cannabis/THC reports as well as the number of cases in which cannabis/THC was identified decreased from 2015 through 2021, including a decrease from 188,735 to 167,669 from 2020 to 2021. HHS noted that this could mean there was a decrease in the number of cannabis/THC encounters, but it could also mean that there was a decrease in the number of exhibits submitted by law enforcement for analysis or a decrease in the number of exhibits processed (analyzed) by forensic laboratories. HHS Basis for Rec. at 45.

In HHS's view, the most notable conclusion from its evaluation of epidemiological databases related to the medical outcomes from drug abuse is that, for all evaluated measures from 2015 to 2020, the rank order of comparators in terms of greatest adverse consequences typically placed alcohol (unscheduled), heroin (schedule I), and cocaine (schedule II) in the first or immediately subsequent position, with marijuana in a lower position. This pattern also held for PC data for serious medical outcomes, including death, where marijuana was in the lowest ranking group. HHS determined that this demonstrated that there is consistency across databases, across substances, and over time, and that although abuse of marijuana produces clear evidence of harmful consequences, including SUD, the consequences are relatively less common and less harmful than some other comparator drugs. Additionally, HHS concluded, the number of law enforcement encounters with marijuana decreased from 2020 to 2021, at a time when law enforcement encounters were increasing for other scheduled drugs of abuse. However, as it noted with respect to Factor 1.A, HHS emphasized that there are limitations in comparing descriptive data on adverse outcomes across drugs, although descriptive analyses of epidemiologic data are an established practice in previous eight-factor analyses. HHS Basis for Rec. at 45.

In 2016, DEA found that abuse of marijuana is widespread and significant. 81 FR 53739. In addition, DEA found in 2016 that a significant proportion of all admissions for substance abuse treatment are for marijuana/hashish as the primary drug of abuse. Id. DEA notes that national data demonstrate that marijuana is one of the most widely used federally illicit substances in the United States, consistent with findings from the HHS Basis for Recommendation. According to the NSDUH, in 2022, among people aged 12 or older in the United States, an estimated 61.9 million people (22 percent) had used marijuana in the past year, and 42.3 million (15.0 percent) had used it in the past month. DEA notes that, according to one National Institutes of Health-supported study, the prevalence of daily marijuana use reached its highest level reported in 2021, at 11 percent of Americans aged 12 or older, a 3 percent increase from 2017 and a 5 percent increase from 2012. 27 It also notes that the average percentage of Δ9-THC in seized marijuana has increased over time. 28 Also, TEDS data showed that, in 2020, marijuana was the primary drug of admission in approximately 10 percent of all admissions to substance abuse treatment among patients aged 12 and older. HHS Basis for Rec. at 41, 46. DEA also notes that TEDS data for 2021 reported that marijuana/hashish was the primary substance of abuse in 10.2 percent of all admissions to substance abuse treatment among patients aged 12 and older. 29 The 2021 TEDS data further reported that New York, California, Georgia, North Carolina, New Jersey, Texas, Minnesota, South Carolina, Florida, and Connecticut accounted for 55.9 percent of admissions to substance use treatments services where marijuana/hashish was listed as the primary substance. 30 DEA also believes that additional information regarding the scope, duration, and significance of marijuana abuse may be appropriate for consideration in assessing this factor.

The sixth factor that DOJ and HHS must consider under 21 U.S.C. 811(c) is the risk posed to the public health by marijuana. In analyzing this factor, HHS examined NSDUH data related to the demographics of U.S. individuals meeting criteria for marijuana use disorder, TEDS data related to the demographics of admission to treatment centers for marijuana use disorder, NEDS and NIS data on admissions to EDs and hospitals related to marijuana poisoning, ToxIC Core Registry data on intentional and unintentional exposure, and NPDS data describing the risks to youth of unintentional exposure to marijuana. HHS also assessed the risks to the public health through NSDUH data on driving under the influence of marijuana in adults and high school students. Finally, HHS reported data regarding the risk of serious AEs and death associated with nonmedical use/use of uncertain intent of marijuana as reported to the FDA Adverse Event Reporting System, Center for Food Safety and Applied Nutrition Adverse Event Reporting System, National Vital Statistics System-Mortality (“NVSS-M”), DAWN, FDA's Sentinel Distributed Database System, and Centers for Medicare and Medicaid Services, and as reflected in the Drug-Involved Mortality data linking NVSS-M to death certificates. HHS Basis for Rec. at 46.

HHS's detailed analysis of the risks posed by marijuana to the public health can be found at pages 46-57 of the HHS Basis for Recommendation. In summary, HHS found that the risks to the public health posed by marijuana are low compared to other drugs of abuse ( e.g., heroin (schedule I), cocaine (schedule II)), based on its evaluation of various epidemiological databases for ED visits, hospitalizations, unintentional exposures, and, most importantly, for overdose deaths. The rank order of comparator drugs in terms of greatest adverse consequences typically places heroin, benzodiazepines, or cocaine in the first or immediately subsequent positions, with marijuana in a lower place in the ranking, especially when comparing among individuals who reported using the respective drugs at least once in the prior year. For overdose deaths, marijuana is always ranked the lowest among comparator drugs. HHS interpreted these evaluations to demonstrate that there is consistent evidence across databases, across substances, and over time that, although the abuse of marijuana poses a risk to public health, the risk is relatively lower than that posed by most other comparator drugs. However, as HHS noted in its discussion of Factor 1, see HHS Basis for Rec. at 7-8, there are limitations in comparing descriptive data on adverse outcomes across drugs. HHS Basis for Rec. at 57.

In 2016, DEA found that, “[t]ogether with the health risks outlined in terms of pharmacological effects above, public health risks from acute use of marijuana include impaired psychomotor performance, impaired driving, and impaired performance on tests of learning and associative processes. Chronic use of marijuana poses a number of other risks to the public health including physical as well as psychological dependence.” 81 FR 53739-40. In addition to the data provided in the HHS Basis for Recommendation and the data considered by HHS and DEA in their prior eight-factor analyses, DEA anticipates that additional data on public safety risks, risks from acute and chronic marijuana use via oral and inhaled administration routes, and the impact of Δ9-THC potency may be appropriate for consideration.

As discussed in the HHS Basis for Recommendation, DEA notes that studies have examined the risk associated with marijuana use and driving. HHS Basis for Rec. at 50. The Rocky Mountain High Intensity Drug Trafficking Area reported in a publication that traffic deaths in Colorado in which drivers tested positive for marijuana more than doubled from 55 in 2013 to 131 in 2020, although other evidence in the same report suggests that driving under the influence citations involving marijuana have grown at a rate similar to the rate for citations involving other drugs. 31 DEA also identified some evidence suggesting that, among drivers who test positive for at least one drug in a traffic stop, a growing share test positive for cannabis. 32

The seventh factor that DOJ and HHS are required to consider under 21 U.S.C. 811(c) is the psychic or physiologic dependence liability of marijuana.

The term “psychic or psychological dependence” has been used to refer to a state similar to addiction. For diagnosis purposes, the DSM-V has combined the diagnoses “abuse” and “drug dependence” ( i.e., addiction), which the DSM's Fourth Edition specified separately, into a single “substance use disorder,” which may occur in a broad range of severity, from mild to severe. HHS Basis for Rec. at 57.

The abuse potential of a drug can be assessed, in part, by evaluating the rewarding effects produced by that drug in humans and animals. As HHS described in its analysis of Factor 2, see HHS Basis for Rec. at 12-13, rodent behavioral studies show that Δ9-THC produces both self-administration and CPP. HHS determined that these results demonstrate that Δ9-THC has rewarding properties that are indicative of abuse potential. Further, as HHS described in its analysis of Factor 4, see HHS Basis for Rec. at 32-37, there is ample epidemiological evidence that marijuana is self-administered by humans, which may result from its ability to produce rewarding psychological effects, such as euphoria, see HHS Basis for Rec. at 15. HHS Basis for Rec. at 58.

In some individuals, extensive use of marijuana can lead to SUD. HHS noted that, in general, SUDs listed in the DSM-V are defined by an inability to cease drug use despite harmful consequences; Cannabis Use Disorder (“CUD”) shares this and other diagnostic criteria common to SUDs for other drugs of abuse. Estimates of CUD in individuals who regularly use marijuana vary and range from about 10 to 20 percent. These estimates are similar to data from the United States National Comorbidity Study, which showed that 9 percent of lifetime cannabis users met the criteria for dependence outlined in the DSM's revised Third Edition at some time in their life, compared to 32 percent of tobacco users, 23 percent of opiate users, and 15 percent of alcohol users. The National Epidemiologic Survey on Alcohol and Related Conditions also reported a nine percent lifetime cumulative probability of transitioning from marijuana use to dependence, with a higher risk of dependence in individuals with a history of psychiatric or other substance dependence comorbidity. In the United States, data from the 2020 NSDUH show that approximately 14 million individuals aged 12 or older who use marijuana or other cannabinoid preparations met criteria for CUD, representing 5.1 percent of all individuals aged 12 or older meeting the NSDUH survey inclusion criteria. HHS Basis for Rec. at 58.

Individuals who develop a SUD, including CUD, may seek treatment. From 2015 to 2020, TEDS documented approximately 10.8 million treatment episode admissions reported by individuals treated at publicly funded substance use treatment programs. Out of 1.4 million treatment admissions documented by TEDS in 2020, marijuana was reported as the primary substance of abuse in approximately 10 percent of admissions, making it the third most frequently reported primary substance of abuse, after alcohol (31.2 percent) and heroin (20.6 percent). A similar pattern was seen from 2015 to 2019. HHS Basis for Rec. at 58.

HHS concluded that the animal behavioral data show that Δ9-THC produces rewarding properties that underlie the abuse potential of marijuana. Epidemiological data demonstrate that some individuals who use marijuana for its rewarding properties go on to develop CUD, which shows that marijuana can produce psychological dependence. Among those individuals who seek admission for treatment for SUD associated with a drug of abuse, marijuana was the third most frequently reported primary substance of abuse. Thus, marijuana can produce psychic dependence in some individuals who use the drug. HHS Basis for Rec. at 58-59.

Physical dependence is a state of adaptation manifested by a drug-class specific withdrawal syndrome produced by abrupt cessation, rapid dose reduction, decreasing blood level of the drug, or administration of an antagonist. Although physical dependence is often associated with addiction, it can be produced by repeated administration of drugs both with and without abuse potential. HHS Basis for Rec. at 59.

As HHS discussed in its analysis of Factor 2, see HHS Basis for Rec. at 11, Δ9-THC is a partial agonist at CB 1 receptors. When marijuana (or isolated Δ9-THC) is administered chronically, there is a down-regulation of CB 1 receptors, which leads to behavioral tolerance. The underlying mechanism for marijuana withdrawal appears to be the uncoupling or desensitization of CB 1 receptors that precedes receptor down-regulation. Abrupt discontinuation of marijuana after prolonged administration produces withdrawal symptoms in rats and in humans that are typically opposite to those that occur with activation of the CB 1 receptor. Precipitated withdrawal can also be induced with administration of CB 1 antagonists following chronic administration, while administration of CB 1 agonists can attenuate some withdrawal symptoms associated with marijuana discontinuation. These data confirm the importance of the CB 1 receptor in marijuana physical dependence. HHS Basis for Rec. at 59.

HHS noted that research has not yet documented the occurrence of withdrawal symptoms in individuals who use marijuana only occasionally. However, in individuals who use marijuana heavily and chronically, drug discontinuation can lead to a withdrawal syndrome. Most marijuana withdrawal symptoms begin within 24 to 48 hours of drug discontinuation, peak within two to six days, and reduce over one to two weeks as Δ9-THC levels decline. HHS Basis for Rec. at 59.

The most commonly reported withdrawal symptoms from clinical investigations are sleep difficulties, decreased appetite and weight loss, craving, irritability, anger, anxiety or nervousness, and restlessness. Less commonly reported withdrawal symptoms include depressed mood, sweating, shakiness, physical discomfort, and chills. HHS described the symptoms of “cannabis withdrawal” listed in the DSM-V as being similar to those reported in the experimental studies, including nervousness or anxiety, irritability or aggression, insomnia or unpleasant dreams, depressed mood, decreased appetite or weight loss, restlessness, abdominal pain, shakiness or tremors, sweating, fever, chills, and headache. HHS Basis for Rec. at 59-60.

HHS reported that up to 40 to 50 percent of individuals who use marijuana on a regular basis may experience physical dependence. A meta-analysis of 23,518 individuals who frequently used marijuana showed that 47 percent of subjects reported symptoms of marijuana withdrawal. The prevalence of physical dependence was 54 percent in outpatient samples, 17 percent in community samples, and 87 percent among inpatients in drug abuse treatment centers. This is consistent with data showing that 90 percent of individuals who were diagnosed with CUD also reported physical dependence. Further, individuals diagnosed with CUD experience more severe and longer lasting withdrawal symptoms when discontinuing marijuana than individuals who do not have a diagnosis of CUD. This may be because individuals with CUD have greater exposure to marijuana. HHS Basis for Rec. at 60.

Symptoms associated with marijuana withdrawal appear to be relatively mild compared to those associated with alcohol withdrawal, which can include agitation, paranoia, seizures, and even death. Multiple studies comparing the withdrawal symptoms associated with tobacco (not scheduled in the CSA) and marijuana demonstrate that the magnitude and time course of the two withdrawal syndromes are similar. Animal studies have shown that after short-term administration of equianalgesic doses of heroin and Δ9-THC to monkeys, withdrawal signs were observed after heroin administration but not after Δ9-THC administration, further demonstrating that withdrawal from marijuana is associated with less severe symptoms than withdrawal from other drug classes. HHS Basis for Rec. at 60.

In conclusion, HHS found experimental and clinical evidence that chronic, but not acute, use of marijuana can produce both psychic and physical dependence in humans. Epidemiological data, discussed in greater detail in the sections describing Factors 4 and 5 in sections VI.4 and VI.5 of this preamble, provide additional evidence of psychic dependence. The symptoms associated with both kinds of dependence are relatively mild for most individuals, although their severity may be greater with increased exposure to marijuana. HHS Basis for Rec. at 61.

In 2016, DEA found that “[l]ong-term, heavy use of marijuana can lead to physical dependence and withdrawal following discontinuation, as well as psychic or psychological dependence.” 81 FR 53740. DEA notes that some physicians have argued that CUD is underdiagnosed and undertreated in the medical setting, 33 and that other medical professionals have noted that CUD needs to be better understood and characterized to better inform users and treatment professionals. 34 DEA anticipates that additional psychic or physiological dependence liability may be appropriate for consideration.

The eighth factor that DOJ and HHS are required to consider under 21 U.S.C. 811(c) is whether marijuana is an immediate precursor of a substance already controlled under the CSA. HHS concluded that marijuana is not an immediate precursor of another controlled substance. HHS Basis for Rec. at 61. This finding is consistent with DEA's finding in 2016. 81 FR 53740. DEA welcomes additional information on this factor.

After conducting the eight-factor analysis in 2023, HHS has recommended three findings regarding the appropriate schedule in which to place marijuana. The three findings relate to: (1) a substance's abuse potential; (2) whether the substance has a CAMU; and (3) the safety or dependence potential of the substance. 21 U.S.C. 812(b); HHS Basis for Rec. at 62-65.

In 2016, HHS found that many factors indicated marijuana's high abuse potential, “including the large number of individuals regularly using marijuana, marijuana's widespread use, and the vast amount of marijuana available for illicit use.” 81 FR 53688 at 53706. As a result of its most recent evaluation, which incorporates post-2016 data into its analysis, HHS has recommended a finding that marijuana has a potential for abuse less than the drugs or other substances in schedules I and II.

Marijuana contains Δ9-THC (also known as dronabinol when specifically referring to (-) - trans -Δ9-THC stereoisomer), the substance responsible for the abuse potential of marijuana. Δ9-THC has agonist properties at CB 1 cannabinoid receptors and produces rewarding responses in animals, as evidenced by its ability to produce self-administration and CPP. When marijuana is administered to humans under experimental conditions, it produces a wide range of positive subjective responses in addition to certain negative subjective responses. Common responses to marijuana when it is used by individuals for nonmedical purposes include euphoria and other positive subjective responses, as well as perceptual changes, sedative responses, anxiety responses, psychiatric, social, and cognitive changes, and physiological changes. HHS Basis for Rec. at 62.

HHS noted that epidemiological data from NSDUH show that marijuana is the most frequently used federally illicit drug in the United States on a past-year and past-month basis among the illicit comparator drugs considered. Although 50 percent of respondents in NSDUH reported using marijuana nonmedically fewer than 5 days per month, another 30 percent reported using it nonmedically for 20 days or more per month. HHS Basis for Rec. at 62.

Despite the high prevalence of nonmedical use of marijuana, HHS observed that an overall evaluation of epidemiological indicators suggests that it does not produce serious outcomes compared to drugs in schedules I or II. HHS found this especially notable given the availability of marijuana and marijuana-derived products that contain extremely high levels of Δ9-THC. Due to such availability, the epidemiological data described in HHS's evaluation inherently include the outcomes from individuals who use marijuana and marijuana-derived products that have doses of Δ9-THC that range from low to very high, and yet the data demonstrate that these products overall are producing fewer negative outcomes than drugs in schedules I or II. HHS Basis for Rec. at 62.

HHS compared the rank ordering of selected drugs that are abused for various epidemiological measures and observed that marijuana was among the drugs at the very lowest ranking for a number of measures, including PC abuse cases, likelihood that any use would lead to a PC call, accidental or unintentional poisoning, utilization-adjusted rates of unintentional exposure, utilization-adjusted and population-adjusted rates for ED visits and hospitalizations, likelihood of being diagnosed with a serious SUD, deaths reported to PCs, and overdose deaths when used with other drugs or as a single substance (as total numbers and when utilization-adjusted). In contrast, comparators such as heroin (schedule I), oxycodone (schedule II), and cocaine (schedule II) typically were in the highest rank ordering on these measures. HHS Basis for Rec. at 62.

For the various epidemiological measures evaluated above, HHS noted that marijuana was also compared to controlled substances in schedule III (ketamine) and schedule IV (benzodiazepines, zolpidem, and tramadol), as well as to other schedule II substances (fentanyl and hydrocodone). The analyses were conducted in this manner to provide a comprehensive assessment of the relative abuse potential of marijuana. However, the rank order of these substances regarding harms does not consistently align with the relative scheduling placement of these drugs in the CSA due to the pharmacological differences between various classes of drugs. HHS Basis for Rec. at 63.

There are a number of confounding factors that likely influence the adverse outcomes measured in various epidemiological databases and account for the rank ordering of the drugs evaluated on these measures. For example, a different population abuses each substance, and each substance has a different prevalence of abuse and a different profile of severe adverse outcomes in a setting of nonmedical use and abuse. Thus, it is challenging to reconcile the ranking of relative harms associated with the comparators used in this evaluation when the rankings differ across various epidemiological databases and when these rankings often do not align with the scheduling placement of these comparators under the CSA. HHS Basis for Rec. at 63.

To address these challenges, HHS evaluated the totality of the available data and has concluded that it supports the placement of marijuana in schedule III. Overall, these data demonstrate that, although marijuana is associated with a high prevalence of abuse, the profile of and propensity for serious outcomes related to that abuse lead to a conclusion that marijuana is most appropriately controlled in schedule III under the CSA. HHS Basis for Rec. at 63.

The Attorney General has considered HHS's recommendations and conclusions and accords HHS's scientific and medical determinations binding weight at this stage of the scheduling process. See OLC Op. at *22 n.6 (“HHS's recommendations with respect to `scientific and medical matters' are binding for all eight factors listed in section 811(c).”). The Attorney General concurs with HHS's recommendation, for purposes of initiation of these rulemaking proceedings, that marijuana has a potential for abuse less than the drugs or other substances in schedules I and II.

In 2016, HHS recommended a finding that marijuana had no CAMU due in part to a lack of adequate safety studies or evidence that qualified experts accepted marijuana for use in treating a specific, recognized disorder. 81 FR 53688 at 53707. As a result of its most recent evaluation, which incorporates post-2016 data into its analysis, HHS recommends a finding that marijuana has a CAMU.

In making that recommendation, HHS analyzed whether there is (1) widespread current experience with medical use of the substance in the United States by licensed health care practitioners operating in accordance with implemented State-authorized programs, where the medical use is recognized by entities that regulate the practice of medicine; and (2) some credible scientific support for a least one of those medical uses. Applying this test, HHS recommended a finding that marijuana has a currently accepted medical use in the United States, specifically for the treatment of anorexia related to a medical condition, nausea and vomiting ( e.g., chemotherapy-induced), and pain. According to HHS, its evaluation also supported a finding that there is accepted safety for the use of marijuana under medical supervision for the treatment of anorexia related to a medical condition, nausea and vomiting ( e.g., chemotherapy-induced), and pain. HHS Basis for Rec. at 63-64.

In the past, DEA has concluded that a substance has a CAMU under the CSA only if one of two tests is satisfied. First, DEA has determined that a substance has a CAMU if the substance has been approved by FDA for marketing under the FDCA, either through the NDA process or by meeting the criteria to be recognized as a “Generally Recognized As Safe and Effective” (“GRASE”) drug. 57 FR 10499, 10503 (March 26, 1992). Second, DEA has determined a substance has a CAMU if the substance satisfies a five-part test established by DEA in 1992 that was based on the “core FDCA standards for acceptance of drugs for medical use”:

1. There must be adequate safety studies;

2. The drug's chemistry must be known and reproducible;

3. There must be adequate and well-controlled studies proving efficacy;

4. The drug must be accepted by qualified experts; and

5. The scientific evidence must be widely available.

In its most recent evaluation, HHS informed DEA of its view that DEA's previous approach to determining whether a drug has a CAMU does not adequately account for certain indicia of medical use that, where present, are relevant to determining whether a substance has a CAMU for purposes of scheduling under the CSA. Specifically, HHS observed that DEA's tests left no room for an evaluation of (1) whether there is widespread medical use of a drug under the supervision of licensed health care practitioners under State-authorized programs and, (2) if so, whether there is credible scientific evidence supporting such medical use. HHS therefore developed an alternative test composed of those two inquiries as a third, independently sufficient approach for determining whether a substance has a CAMU under the CSA. HHS applied this two-part test to marijuana and recommended a finding that marijuana has a CAMU under the CSA. HHS Basis for Rec. at 24-28.

Upon receiving HHS's recommendation, the Attorney General requested that OLC advise on whether HHS's test, if satisfied, established a CAMU “even if the drug has not been approved by FDA and even if the drug does not satisfy DEA's five-part test.” OLC Op. at *3. OLC determined that DEA's current approach to determining whether a drug has a CAMU is impermissibly narrow, because it “ignor[es] widespread clinical experience with a drug that is sanctioned by state medical licensing regulators.” Id. at *13-14; see also id. at *12. OLC further opined that satisfying HHS's two-part inquiry is sufficient to establish that a drug has a CAMU. Id. at * 4, *16-20. And OLC concluded that, while HHS's CAMU recommendation is not binding on DEA, the medical and scientific determinations that underlie its recommendation are binding until the initiation of formal rulemaking proceedings, and that DEA must accord those determinations “significant deference” throughout the rulemaking process. Id. at *4, *20-26.

Under Part 1 of the HHS CAMU test, the Office of the Assistant Secretary for Health (“OASH”) considered whether there is widespread current experience with medical use of marijuana in the United States by licensed HCPs operating in accordance with implemented State-authorized programs, where such medical use is recognized by entities that regulate the practice of medicine under these State jurisdictions. Part 2 of the CAMU test evaluated whether there exists some credible scientific support for at least one of the medical conditions for which the Part 1 test is satisfied. The evaluation in Part 2, undertaken by FDA, was not meant to be, nor is it, a determination of safety and efficacy under the Federal Food, Drug, and Cosmetic Act's drug approval standard for new human or animal drugs. Rather, HHS's two-part test is designed to evaluate whether a substance, in this case marijuana, has a CAMU for purposes of drug scheduling recommendations and placement in a drug schedule consistent with criteria set forth in 21 U.S.C. 812(b). HHS Basis for Rec. at 24.

In the evaluation and assessment under Part 1 of the CAMU test, OASH found that more than 30,000 HCPs are authorized to recommend the use of marijuana for more than six million registered patients, constituting widespread clinical experience associated with various medical conditions recognized by a substantial number of jurisdictions across the United States. For several jurisdictions, these programs have been in place for several years, and include features that actively monitor medical use and product quality characteristics of marijuana dispensed. HHS Basis for Rec. at 24.

Based on OASH's findings in Part 1 of the CAMU test, the Assistant Secretary for Health concluded that an FDA assessment under Part 2 of the CAMU test was warranted to determine if credible scientific support exists for the use of marijuana to treat at least one of the medical conditions identified by OASH under Part 1. HHS Basis for Rec. at 24.

At this stage of initiating a rulemaking, the Attorney General agrees with OASH that there is widespread clinical experience with marijuana for at least one medical condition.

FDA conducted Part 2 of the CAMU test for seven indications, based in part on OASH's findings under Part 1 of the CAMU test  35 and in part on FDA's own analysis of the landscape in which marijuana is currently used medically, including information from State-authorized programs on how and to what extent marijuana is being utilized for medical purposes. The seven indications are: (1) anorexia;  36 (2) anxiety;  37 (3) epilepsy; (4) inflammatory bowel disease (“IBD”); (5) nausea and vomiting; (6) pain; and (7) post-traumatic stress disorder (“PTSD”). FDA's evaluation under Part 2 of the CAMU test was based on systematic reviews of studies investigating the safety and effectiveness of marijuana, relevant professional societies' position statements, data from State medical marijuana programs and United States national surveys, and the labeling of FDA-approved products relevant to the analysis. HHS Basis for Rec. at 25.

In evaluating whether there exists some credible scientific support under Part 2 of the CAMU test for a particular use, factors in favor of a positive finding included whether: (1) favorable clinical studies of the medical use of marijuana, although not necessarily adequate and well-controlled clinical studies that would support approval of an NDA, have been published in peer-reviewed journals or (2) qualified expert organizations ( e.g., academic groups, professional societies, or government agencies) have opined in favor of the medical use or provided guidance to HCPs on the medical use. Factors that weigh against a finding that Part 2 of the CAMU test is met included whether: (1) data or information indicate that medical use of the substance is associated with unacceptably high safety risks for the likely patient population, e.g., due to toxicity concerns; (2) clinical studies with negative efficacy findings for the medical use of marijuana have been published in peer reviewed journals; or (3) qualified expert organizations ( e.g., academic or professional societies, government agencies) recommend against the medical use of marijuana based on the available data at the time of their position statement. HHS Basis for Rec. at 25.

FDA's review of the available information identified mixed findings of effectiveness across indications, ranging from data showing inconclusive findings to considerable evidence in favor of effectiveness, depending on the source. The largest evidence base for effectiveness exists for marijuana use within the pain indication (in particular, neuropathic pain). Numerous systematic reviews concluded that there exists some level of evidence supporting the use of marijuana for chronic pain. The 2017 NASEM report concluded there was “substantial evidence”  38 supporting the use of cannabis products relevant to this review for pain, as have other reviews. The AHRQ living systematic review has concluded that there is some support for the use of marijuana-related products in the treatment of chronic pain, but overall concluded these effects were small and the increased risk of dizziness, nausea, and sedation may limit the benefit. A systematic review of scientific and medical literature was conducted in 2023 by the University of Florida (“UF”) under contract with FDA. UF epidemiologists identified some data supporting effectiveness of marijuana, including some within their own meta-analysis; however, they ultimately concluded the results are inconclusive or mixed. FDA also conducted a separate analysis of published scientific reviews, several of which drew conclusions similar to those of UF. HHS Basis for Rec. at 25-26.

UF evaluated other therapeutic conditions mentioned above, i.e., anorexia, anxiety, epilepsy, IBD, nausea, and PTSD, employing a similar systematic review of scientific and medical literature. UF found that there is low- to moderate-quality evidence supporting the use of marijuana as medical treatment for outcomes in anorexia, nausea and vomiting, and PTSD. FDA's review of systematic reviews showed mixed results for these indications. In particular, FDA found that the potential for psychiatric adverse events associated with treating PTSD with marijuana may be more substantial than any limited benefit in observational studies. Although UF did not conclude that there was evidence in support of the effectiveness of marijuana in IBD, both their review and other systematic reviews found some benefit with respect to subjective symptoms in this condition. With regard to epilepsy and anxiety, both UF's review and FDA's review of other systematic reviews did not find support for marijuana providing benefit in the treatment of these conditions. Where positive results on effectiveness outcome measures were found, the effects and the quality of evidence were generally in the low-to-moderate range. UF did not find high quality evidence supporting worsening of outcomes in any indication. HHS Basis for Rec. at 26.

FDA concluded that none of the evidence from the systematic reviews included in the CAMU test Part 2 analysis identified any safety concerns that would preclude the use of marijuana in the indications for which there exists some credible scientific support for its therapeutic benefit. FDA assessed the clinical safety data identified in the literature from controlled trials as generally consistent between sources but limited in the rigor of safety reporting. FDA also explained that the vast majority of the observational studies evaluated in the context of medical use were excluded from the final synthesis of evidence due to concerns regarding their quality ( e.g., only one observational study for the anxiety indication and one for the PTSD indication were included). According to FDA, data on safety from both clinical trials and observational studies were generally scarce, but the literature shows that marijuana has more AEs when compared to a placebo or active control group, however, typically in the mild to moderate severity range. HHS Basis for Rec. at 26.

FDA also reviewed results from State reporting data from 37 States with medical marijuana programs and surveys of patients using marijuana in Maryland and Minnesota, which had data available for review. Surveys of patients using marijuana in these two States found most patients did not report any side effects and those that did report side effects mostly described them as mild. Neither State's databases included patients who chose to stop using marijuana, which FDA noted might result in an overestimation of positive experiences. HHS Basis for Rec. at 27.

As of August 2023, FDA reported that the real-world data sources available to FDA, in general, lack the necessary elements to identify the exposure ( i.e., to marijuana), to distinguish the reason for use (medical vs. recreational) and, if applicable, the condition that prompted its medical use, and to permit sound inferential analyses. Therefore, they were not included in HHS's review. HHS Basis for Rec. at 27.

According to FDA, data from United States national surveys, in general, lacked details on patient characteristics and factors that prompted the use of marijuana for medical purposes, and data collection for these surveys was impacted by the COVID-19 pandemic. FDA observed that, despite these limitations, the data suggested that medical use of marijuana increases as age increases. Only data from one survey provided information on the intended indication for use, suggesting that individuals often use marijuana to improve or manage conditions such as depression, anxiety, PTSD, pain, headaches or migraines, sleep disorders, nausea and vomiting, lack of appetite, and muscle spasms, but only approximately half of them reportedly had ever asked a health care professional for a recommendation to use medical marijuana. HHS Basis for Rec. at 27.

Additionally, although the safety data obtained from use in a medical context are considered to be the most relevant for the CAMU analysis, FDA evaluated the safety of marijuana in the nonmedical setting to inform the potential for more severe outcomes. Specifically, FDA evaluated safety outcomes related to marijuana use in the setting of nonmedical use, use of uncertain intent, and unintentional exposure through a variety of epidemiological data sources and in relation to several comparator substances controlled under the CSA, including drugs in schedule I: heroin (an illicit opioid drug); schedule II: hydrocodone and oxycodone (approved opioid prescription drug products), cocaine and fentanyl (largely illicitly produced drugs in the nonmedical use setting, although there are approved prescription drugs); schedule III: ketamine (an approved prescription drug); and schedule IV: zolpidem, benzodiazepines, and tramadol (approved prescription drugs). According to FDA, the comparative data demonstrate that, even in the context of nonmedical use, marijuana has a less concerning overall safety profile relative to the comparators for a number of important outcomes ( e.g., single substance use overdose death, hospitalizations). However, FDA observed that in young children, population-adjusted rates of ED visits and hospitalizations involving marijuana poisoning were higher than heroin, cocaine, and benzodiazepines for the periods studied. Of note, some of the comparator substances are approved for use in conditions similar to the indications for which marijuana was evaluated in the CAMU analysis ( e.g., opioids for pain, benzodiazepines for anxiety-related conditions). HHS Basis for Rec. at 27.

FDA also considered position statements from professional organizations relevant to the indications discussed. The vast majority of professional organizations did not recommend the use of marijuana in their respective specialties; however, none specifically recommended against it, with the exception of the American Psychiatric Association, which stated that marijuana is known to worsen certain psychiatric conditions. HHS Basis for Rec. at 27-28.

On balance, FDA found the available data indicated that there is some credible scientific support for the use of marijuana in the treatment of chronic pain, anorexia related to a medical condition, and nausea and vomiting, with varying degrees of support and consistency of findings. Additionally, no safety concerns were identified in FDA's review that would indicate that medical use of marijuana poses unacceptably high safety risks for the indications where there is some credible scientific evidence supporting its therapeutic use. HHS Basis for Rec. at 28.

Based on the totality of the available data, FDA concluded that there exists some credible scientific support for the medical use of marijuana in at least one of the indications for which there is widespread current experience in the United States, as identified by OASH under Part 1 of the CAMU test. The indications evaluated were anorexia related to a medical condition, anxiety, epilepsy, IBD, nausea and vomiting ( e.g., chemotherapy-induced), pain, and PTSD. FDA clarified that the analysis and conclusions on the available data are not meant to imply that safety and effectiveness have been established for marijuana that would support FDA approval of a marijuana drug product for a particular indication. However, FDA determined that the available data do provide some level of support for the way marijuana is being recommended by health care practitioners in clinical practice. Thus, based on the widespread HCP experience and the extent of medical use evaluated by OASH under the Part 1 test, and FDA's evaluation of available credible scientific support described herein for at least some therapeutic uses identified in the Part 1 test, HHS recommended a finding that, for purposes of the drug scheduling criteria in 21 U.S.C. 812(b), marijuana has a CAMU for: anorexia related to a medical condition; nausea and vomiting ( e.g., chemotherapy-induced); and pain. HHS Basis for Rec. at 28.

The Attorney General has considered HHS's recommendations and conclusions and accords HHS's scientific and medical determinations binding weight until the initiation of the formal rulemaking process. See OLC Op. at *24. Applying HHS's two-part test, and in light of OLC's legal opinion that the HHS's test is sufficient under the CSA, the Attorney General concurs with HHS's conclusion, for purposes of the initiation of these rulemaking proceedings, that there is a CAMU for marijuana.

As a result of its most recent evaluation, which incorporates post-2016 data into its analysis, HHS has recommended a finding that abuse of marijuana may lead to moderate or low physical dependence or high psychological dependence. HHS Basis for Rec. at 65.

According to HHS, clinical studies have demonstrated that marijuana produces physical and psychological dependence. Regarding physical dependence, as evidenced by its associated withdrawal symptomology upon abrupt discontinuation of use, the most commonly reported marijuana withdrawal symptoms in clinical investigations are sleep difficulties, decreased appetite and weight loss, craving, irritability, anger, anxiety or nervousness, and restlessness. Marijuana withdrawal symptoms typically peak within two to six days and decline over one to two weeks as Δ9-THC is eliminated. Similarly, the drug labels for the FDA-approved drug products Marinol and Syndros state that, following chronic administration of dronabinol, drug discontinuation leads to irritability, insomnia, and restlessness at 12 hours, and by 24 hours the withdrawal symptoms can include hot flashes, sweating, rhinorrhea, diarrhea, and anorexia. HHS Basis for Rec. at 64.

HHS observes that marijuana withdrawal syndrome has been reported in individuals with heavy, chronic marijuana use, but its occurrence in occasional users of marijuana has not been established. The marijuana withdrawal syndrome appears to be relatively mild compared to the withdrawal syndrome associated with alcohol, which can include more serious symptoms such as agitation, paranoia, seizures and even death. Multiple studies comparing the withdrawal symptoms associated with marijuana and tobacco demonstrate that the magnitude and time course of the two withdrawal syndromes are similar. HHS Basis for Rec. at 64.

HHS also notes that the ability of marijuana to produce psychic dependence is shown through its ability to produce rewarding effects that underlie its nonmedical use and epidemiological outcomes related to abuse, as detailed in the first finding on abuse potential. HHS Basis for Rec. at 64-65.

Based on the evidence, HHS determined that the abuse of marijuana may lead to moderate or low physical dependence, depending on frequency and degree of marijuana exposure. HHS further concluded that marijuana can produce psychic dependence in some individuals, but that the likelihood of serious outcomes is low, suggesting that high psychological dependence does not occur in most individuals who use marijuana. HHS Basis for Rec. at 65.

The Attorney General has considered HHS's recommendations and conclusions and accords HHS's scientific and medical determinations binding weight at this stage of the scheduling process. See OLC Op. at *22 n.6. For purposes of the initiation of these rulemaking proceedings, the Attorney General concurs with HHS's conclusion that the abuse of marijuana may lead to moderate or low physical dependence, depending on frequency and degree of marijuana exposure.

HHS has recommended a finding that marijuana has a CAMU. HHS Basis for Rec. at 63-64. After considering the foregoing facts and data and the recommendation of HHS, and after according binding weight to HHS's scientific and medical determinations, the Attorney General concludes that there is, at present, substantial evidence that marijuana does not warrant control under schedule I of the CSA. Accordingly, the Attorney General is issuing this notice of proposed rulemaking to initiate rulemaking proceedings to reschedule marijuana. 21 U.S.C. 811(b).

HHS has recommended that marijuana be transferred from schedule I to schedule III rather than from schedule I to schedule II based on its evaluation that the drug has a relatively lower level of abuse compared to drugs currently scheduled in schedules I and II and its evaluation that marijuana may lead to moderate or low physical dependence and has a low likelihood of psychic dependence. Consistent with HHS's analysis, the Attorney General has determined at this initial stage that marijuana does not appear to meet the elements of a schedule II drug, which include a high potential for abuse and a likelihood of severe physiological or physical dependence from such abuse. 21 U.S.C. 812(b)(3). Rather, marijuana's profile as a drug with a lower degree of abuse potential than schedule I ( e.g., heroin) and schedule II ( e.g., fentanyl, cocaine) drugs and a moderate to low level of physical dependence militates in favor of rescheduling it in schedule III. Accordingly, in this notice of proposed rulemaking, the Attorney General is proposing to reschedule marijuana in schedule III and solicits comments on these preliminary findings.

This rescheduling of marijuana would apply to marijuana as listed in 21 CFR 1308.11(d)(23). The rescheduling also would apply to marijuana extracts as defined in 21 CFR 1308.11(d)(58) because they meet the statutory definition of marijuana and, prior to 2017, were included in 21 CFR 1308.11(d)(23). See Establishment of a New Drug Code for Marihuana Extract, 81 FR 90194 (Dec. 14, 2016). In addition, this proposal would apply to Δ9-THC derived from the marijuana plant (other than the mature stalks and seeds) that falls outside the definition of hemp, because it meets the statutory definition of marijuana.

This proposal would not apply to synthetically derived THC, which is outside the CSA's definition of marijuana. Those tetrahydrocannabinols that can be derived only through a process of artificial synthesis ( e.g., delta-10-tetrahydrocannabinol) are excluded. HHS provided a recommendation only relating to “marijuana” as defined in the CSA. That definition is limited to the plant (other than the mature stalks and seeds) and derivatives of the plant. Therefore, synthetic THC will remain in schedule I. This rulemaking would not affect the status of hemp (as defined in 7 U.S.C. 1639 o ), because hemp is excluded from the definition of marijuana. This rulemaking is not proposing to reschedule any drug product containing marijuana or THC that previously has been rescheduled out of schedule I ( e.g., Marinol and Syndros). Nor does it impact the status of any previously scheduled synthetic cannabinoids.

In proposing an appropriate schedule for marijuana, the Attorney General must also consider compliance with the treaty obligations of the United States. As the CSA recognizes, the United States is a party to the Single Convention. 21 U.S.C. 801(7). Parties to the Single Convention are obligated to maintain various control provisions related to the drugs that are covered by the treaty. See, e.g., Single Convention arts. 2, 4. Congress enacted many of the CSA's provisions for the specific purpose of ensuring U.S. compliance with the treaty. See OLC Op. at *27. Among these is a scheduling provision, 21 U.S.C. 811(d)(1). Section 811(d)(1) provides that, where a drug is subject to control under the Single Convention, the Attorney General must “issue an order controlling such drug under the schedule he deems most appropriate to carry out such [treaty] obligations, without regard to the findings required by [21 U.S.C. 811(a) or 812(b)] and without regard to the procedures prescribed by [21 U.S.C. 811(a) and (b)].”  39

Marijuana is a drug covered in the Single Convention under the term “cannabis.”  40 OLC initially advised in 1972 that controls under Article 21 of the Single Convention would not be satisfied if marijuana were listed in schedule III, IV, or V of the CSA. Memorandum for John E. Ingersoll, Director, Bureau of Narcotics and Dangerous Drugs, from Mary C. Lawton, Deputy Assistant Attorney General, Office of Legal Counsel, Re: Petition to Decontrol Marihuana; Interpretation of Section 201 of the Controlled Substances Act of 1970 at 12-13 (Aug. 21, 1972). However, OLC has reexamined the conclusion of its 1972 memorandum, taking into account statutory amendments since 1972 and a possibility it did not consider in 1972: placing marijuana into schedule III while issuing regulations that would enable the United States to comply with its international obligations. OLC Op. at *4, 26-35. OLC has concluded that both the Single Convention and the CSA allow the Attorney General to satisfy the treaty obligations of the United States with respect to marijuana by supplementing scheduling decisions with additional controls under the CSA. Id.

If marijuana were listed in schedule III, most of the Single Convention's obligations would continue to be met by CSA statutory authorities and associated regulations. See OLC Op. at *33-34. One potential gap concerns the quota on manufacturing cannabis required by Article 21 of the Convention, but that gap can be filled using the CSA's regulatory authorities. See id. at *34; see also, e.g., 21 U.S.C. 821 (authorizing the Attorney General to impose restrictions “relate[ed] to the . . . control of the manufacture” of a drug); id. 871(b) (authorizing the Attorney General to issue regulations “necessary and appropriate for the efficient execution of his functions under this subchapter”); id. 822(b) (allowing the Attorney General to regulate “the extent” of manufacture of a drug through registration); id. 823(e) (requiring the Attorney General to register an applicant to manufacture a schedule III drug “unless he determines that the issuance of such registration is inconsistent with the public interest”).

In addition, if marijuana is transferred into schedule III, DEA will continue to have authority to maintain its existing regulatory scheme, located at 21 CFR part 1318, governing the registration of manufacturers seeking to plant, grow, cultivate, or harvest marijuana, as required to comply with Articles 23 and 28 of the Single Convention. Authority for those regulations currently flows from 21 U.S.C. 823(a), which is applicable to drugs in schedules I and II. OLC has concluded, however, that 21 U.S.C. 823(e), which is applicable to drugs in schedules III, IV, and V, provides an alternative source of authority for complying with Articles 23 and 28 of the Single Convention. See OLC Op. at *34 n.9. The CSA also recognizes that the United States is also a party to the Convention on Psychotropic Substances, Feb. 21, 1971, 32 U.S.T. 543, 1019 U.N.T.S. 175 (“Convention on Psychotropic Substances”). See also 21 U.S.C. 801a(2). As with the Single Convention, parties to the Convention on Psychotropic Substances are obligated to take various control measures related to the drugs that are covered by the treaty. Id. Congress implemented the additional authority necessary to comply with the Convention on Psychotropic Substances through various amendments to the CSA. Id. 801a(2)-(3).

Δ9-THC is a substance covered by schedule II of the Convention on Psychotropic Substances. In this rule, DOJ proposes to reschedule Δ9-THC that falls within the CSA's definition of marijuana into CSA schedule III. As is the case for marijuana under the Single Convention, the controls available under CSA schedule III are sufficient to comply with the requirements of the Convention on Psychotropic Substances with respect to Δ9-THC, although additional regulatory action may be necessary to implement certain Convention requirements, such as the export and import authorizations required by Article 12. See, e.g., Schedules of Controlled Substances: Rescheduling of the Food and Drug Administration Approved Product Containing Synthetic Dronabinol [(-)-D  9 -(trans)-Tetrahydrocannabinol] in Sesame Oil and Encapsulated in Soft Gelatin Capsules From Schedule II to Schedule III, 64 FR 35928, 35928 (July 2, 1999). Compare, e.g., Convention on Psychotropic Substances art. 12(1) (requiring export and import authorizations for substances in Convention Schedule II), with 21 U.S.C. 952(b)(2) (authorizing import permits for CSA schedule III substances), and id. 953(e)(2) (authorizing export permits for CSA schedule III substances).

Accordingly, concurrent with this rulemaking, DEA will consider the marijuana-specific controls that would be necessary to meet U.S. obligations under the Single Convention and the Convention on Psychotropic Substances in the event that marijuana is rescheduled to schedule III, and, to the extent they are needed if marijuana is rescheduled, will seek to finalize any such regulations as soon as possible.

If marijuana is transferred to schedule III, the regulatory controls applicable to schedule III controlled substances would apply, as appropriate, along with existing marijuana-specific requirements and any additional controls that might be implemented, including those that might be implemented to meet U.S. treaty obligations. The manufacture, distribution, dispensing, and possession of marijuana would also remain subject to applicable criminal prohibitions under the CSA. 21 U.S.C. 841-844.

In addition, marijuana would remain subject to applicable provisions of the FDCA. For example, under the FDCA, a drug containing a substance within the CSA's definition of “marijuana” would need FDA approval to be lawfully “introduce[d] or deliver[ed] for introduction into interstate commerce,” unless an IND is in effect for that drug. See 21 U.S.C. 355(a), 355(i), 331(d). To date, although there have been INDs for drugs containing a substance within the CSA's definition of “marijuana,” no such drugs have been approved by FDA.

DOJ is seeking comment on the practical consequences of rescheduling marijuana into schedule III under the relevant statutory frameworks.

Based on the legal opinion of OLC and consideration of the scientific and medical evaluation and accompanying recommendation of HHS, the Attorney General is initiating a rulemaking that proposes the placement of marijuana in schedule III of the CSA. DOJ is soliciting comments on this proposal.

In accordance with 21 U.S.C. 811(a), this scheduling action is subject to formal rulemaking procedures done “on the record after opportunity for a hearing,” which are conducted pursuant to the provisions of 5 U.S.C. 556 and 557. The CSA sets forth the criteria for removing a drug or other substance from the list of controlled substances. Such actions are exempt from review by the Office of Management and Budget pursuant to section 3(d)(1) of Executive Order 12866 and the principles reaffirmed in Executive Order 13563 and 14094.

While this scheduling action is exempt from review under Executive Order 12866, DOJ recognizes this action may have unique economic impacts. As stated above, marijuana is subject to a number of State laws that have allowed a multibillion dollar industry to develop. DOJ acknowledges that there may be large impacts related to Federal taxes and research and development investment for the pharmaceutical industry, among other things. DOJ is specifically soliciting comments on the economic impact of this proposed rule. DOJ will revise this section at the final rule stage if warranted after consideration of any comments received.

This proposed regulation meets the applicable standards set forth in sections 3(a) and 3(b)(2) of Executive Order 12988 to eliminate drafting errors and ambiguity, minimize litigation, provide a clear legal standard for affected conduct, and promote simplification and burden reduction.

This rulemaking does not have federalism implications warranting the application of Executive Order 13132. The rule does not have substantial direct effects on the States, on the relationship between the Federal Government and the States, or the distribution of power and responsibilities among the various levels of government.

This proposed rule does not have Tribal implications warranting the application of Executive Order 13175. This rule does not have substantial direct effects on one or more Indian Tribes, on the relationship between the Federal Government and Indian Tribes, or on the distribution of power and responsibilities between the Federal Government and Indian Tribes.

DOJ has concluded that this action may have a significant economic impact on a substantial number of small entities under the Regulatory Flexibility Act, 5 U.S.C. 601 et seq. For example, section 280E of the Internal Revenue Code bars businesses from claiming tax deductions for otherwise allowable expenses where the business “consists of trafficking in controlled substances (within the meaning of schedule I and II of the Controlled Substances Act).” 26 U.S.C. 280E. If marijuana is ultimately transferred to schedule III, section 280E would no longer serve as a statutory bar to claiming deductions for those expenses. In addition, small entities engaged in research on marijuana may be subject to different research protocols set by DEA if the research is conducted on a schedule III substance rather than a schedule I substance. 41 However, DOJ is currently not in a position to estimate the number of small entities affected by these or other potential effects of this action. DOJ seeks comment and additional information to inform its analysis.

In accordance with the Unfunded Mandates Reform Act of 1995 (“UMRA”), 2 U.S.C. 1501 et seq., DOJ has determined that this action would not result in any Federal mandate that may result “in the expenditure by State, local, and Tribal governments, in the aggregate, or by the private sector, of $100,000,000 or more (adjusted annually for inflation) in any 1 year.” See 2 U.S.C. 1532(a). Therefore, neither a Small Government Agency Plan nor any other action is required under UMRA.

This action does not impose any new or revised “collection[s] of information” as defined by the Paperwork Reduction Act of 1995, 44 U.S.C. 3502(3).

For the reasons set out above, 21 CFR part 1308 is proposed to be amended as follows:

The revision reads as follows:

I. Table of Abbreviations CFR Code of Federal Regulations DHS Department of Homeland Security FR Federal Register NPRM Notice of Proposed Rulemaking § Section U.S.C. United States Code II. Background, Purpose, and Legal Basis

On December 17, 2023, Goettle Construction notified the Coast Guard that they will be conducting waterside construction associated with a bank stabilization project on the Cuyahoga River in Cleveland, Ohio from August 15, 2024, through November 30, 2025. Construction is intended to shore-up and replace approximately 2,400 linear feet of corrugated steel bulkhead located on the western (left descending) bank of the Cuyahoga River between the Detroit-Superior Bridge and the Columbus Road Bridge. The Captain of the Port Sector Eastern Great Lakes (COTP) has determined that potential hazards associated with the equipment used to complete this project would be a safety concern for any craft intending to navigate near the project area during construction hours. Furthermore, additional safety measures are necessary to keep workers on the construction barges safe while completing the construction project.

The purpose of this rulemaking is to ensure the safety of vessels and the navigable waters within 10 feet of the construction barges during the scheduled construction hours of 8 a.m. through 4 p.m., Monday through Friday. The Coast Guard is proposing this rulemaking under authority in 46 U.S.C. 70034.

The Captain of the Port Eastern Great Lakes (COTP) is proposing to establish a Regulated Navigation Area (RNA) from 8 a.m. through 4 p.m. Monday through Friday starting on August 15, 2024, with an anticipated completion date of November 30, 2025. The RNA would cover all navigable waters within 10 feet of construction barges in the Cuyahoga River located on the western bank (left descending bank) of the Cuyahoga River between the Detroit-Superior Bridge and the Columbus Road Bridge in Cleveland, Ohio. The duration of the zone is intended to ensure the safety of vessels and these navigable waters during the scheduled hours of the construction project. No vessel or craft would be permitted to be operated within 10 feet of the construction barges without obtaining permission from the COTP or a designated representative. The regulatory text we are proposing appears at the end of this document.

We developed this proposed rule after considering numerous statutes and Executive orders related to rulemaking. Below we summarize our analyses based on a number of these statutes and Executive orders, and we discuss First Amendment rights of protestors.

Executive Orders 12866 and 13563 direct agencies to assess the costs and benefits of available regulatory alternatives and, if regulation is necessary, to select regulatory approaches that maximize net benefits. This NPRM has not been designated a “significant regulatory action,” under section 3(f) of Executive Order 12866, as amended by Executive Order 14094 (Modernizing Regulatory Review). Accordingly, the NPRM has not been reviewed by the Office of Management and Budget (OMB).

This regulatory action determination is based on size, location, and duration of the proposed rule.

The Regulatory Flexibility Act of 1980, 5 U.S.C. 601-612, as amended, requires Federal agencies to consider the potential impact of regulations on small entities during rulemaking. The term “small entities” comprises small businesses, not-for-profit organizations that are independently owned and operated and are not dominant in their fields, and governmental jurisdictions with populations of less than 50,000. The Coast Guard certifies under 5 U.S.C. 605(b) that this proposed rule would not have a significant economic impact on a substantial number of small entities.

While some owners or operators of vessels intending to transit the safety zone may be small entities, for the reasons stated in section IV.A above, this proposed rule would not have a significant economic impact on any vessel owner or operator.

If you think that your business, organization, or governmental jurisdiction qualifies as a small entity and that this proposed rule would have a significant economic impact on it, please submit a comment (see ADDRESSES ) explaining why you think it qualifies and how and to what degree this rule would economically affect it.

Under section 213(a) of the Small Business Regulatory Enforcement Fairness Act of 1996 (Pub. L. 104-121), we want to assist small entities in understanding this proposed rule. If the proposed rule would affect your small business, organization, or governmental jurisdiction and you have questions concerning its provisions or options for compliance, please call or email the person listed in the FOR FURTHER INFORMATION CONTACT section. The Coast Guard will not retaliate against small entities that question or complain about this proposed rule or any policy or action of the Coast Guard.

This proposed rule would not call for a new collection of information under the Paperwork Reduction Act of 1995 (44 U.S.C. 3501-3520).

A rule has implications for federalism under Executive Order 13132 (Federalism), if it has a substantial direct effect on the States, on the relationship between the National Government and the States, or on the distribution of power and responsibilities among the various levels of government. We have analyzed this proposed rule under that Order and have determined that it is consistent with the fundamental federalism principles and preemption requirements described in Executive Order 13132.

Also, this proposed rule does not have tribal implications under Executive Order 13175 (Consultation and Coordination with Indian Tribal Governments) because it would not have a substantial direct effect on one or more Indian tribes, on the relationship between the Federal Government and Indian tribes, or on the distribution of power and responsibilities between the Federal Government and Indian tribes. If you believe this proposed rule has implications for federalism or Indian tribes, please call or email the person listed in the FOR FURTHER INFORMATION CONTACT section.

The Unfunded Mandates Reform Act of 1995 (2 U.S.C. 1531-1538) requires Federal agencies to assess the effects of their discretionary regulatory actions. In particular, the Act addresses actions that may result in the expenditure by a State, local, or tribal government, in the aggregate, or by the private sector of $100,000,000 (adjusted for inflation) or more in any one year. Though this proposed rule would not result in such an expenditure, we do discuss the potential effects of this proposed rule elsewhere in this preamble.

We have analyzed this proposed rule under Department of Homeland Security Directive 023-01, Rev. 1, associated implementing instructions, and Environmental Planning COMDTINST 5090.1 (series), which guide the Coast Guard in complying with the National Environmental Policy Act of 1969 (42 U.S.C. 4321-4370f), and have made a preliminary determination that this action is one of a category of actions that do not individually or cumulatively have a significant effect on the human environment. This proposed rule involves a Regulated Navigation Area lasting 8 hours each day, Monday through Friday that would prohibit vessels from operating within 10 feet of the construction barges. Normally such actions are categorically excluded from further review under paragraph L60(a) of Appendix A, Table 1 of DHS Instruction Manual 023-01-001-01, Rev. 1. A preliminary Record of Environmental Consideration supporting this determination is available in the docket. For instructions on locating the docket, see the ADDRESSES section of this preamble. We seek any comments or information that may lead to the discovery of a significant environmental impact from this proposed rule.

The Coast Guard respects the First Amendment rights of protesters. Protesters are asked to call or email the person listed in the FOR FURTHER INFORMATION CONTACT section to coordinate protest activities so that your message can be received without jeopardizing the safety or security of people, places, or vessels.

We view public participation as essential to effective rulemaking and will consider all comments and material received during the comment period. Your comment can help shape the outcome of this rulemaking. If you submit a comment, please include the docket number for this rulemaking, indicate the specific section of this document to which each comment applies, and provide a reason for each suggestion or recommendation.

Submitting comments. We encourage you to submit comments through the Federal Decision-Making Portal at https://www.regulations.gov. To do so, go to https://www.regulations.gov, type USCG-2024-0393 in the search box and click “Search.” Next, look for this document in the Search Results column, and click on it. Then click on the Comment option. If you cannot submit your material by using https://www.regulations.gov, call or email the person in the FOR FURTHER INFORMATION CONTACT section of this proposed rule for alternate instructions.

Viewing material in docket. To view documents mentioned in this proposed rule as being available in the docket, find the docket as described in the previous paragraph, and then select “Supporting & Related Material” in the Document Type column. Public comments will also be placed in our online docket and can be viewed by following instructions on the https://www.regulations.gov Frequently Asked Questions web page. Also, if you click on the Dockets tab and then the proposed rule, you should see a “Subscribe” option for email alerts. The option will notify you when comments are posted, or a final rule is published.

We review all comments received, but we will only post comments that address the topic of the proposed rule. We may choose not to post off-topic, inappropriate, or duplicate comments that we receive.

Personal information. We accept anonymous comments. Comments we post to https://www.regulations.gov will include any personal information you have provided. For more about privacy and submissions to the docket in response to this document, see DHS's eRulemaking System of Records notice (85 FR 14226, March 11, 2020).

For the reasons discussed in the preamble, the Coast Guard is proposing to amend 33 CFR part 165 as follows: