S. 5047 (Introduced-in-Senate)

To require the Director of the National Institute of Standards and Technology to establish a working group to facilitate coordination among relevant stakeholders regarding compound risks relating to artificial intelligence physical systems, and for other purposes.

July 21, 2026

Mr. Sheehy (for himself and Mr. Kim) introduced the following bill; which was read twice and referred to the Committee on Commerce, Science, and Transportation

Be it enacted by the Senate and House of Representatives of the United States of America in Congress assembled,

This Act may be cited as the “Artificial Intelligence Physical Systems Coordination Act of 2026”.

SEC. 2. National Institute of Standards and Technology working group on compound risks relating to artificial intelligence physical systems.

(a) Establishment.—The Director of the National Institute of Standards and Technology (in this section referred to as the “Director”) shall establish a working group (in this section referred to as the “Working Group”) to facilitate coordination among relevant stakeholders regarding effects on the risk landscape arising from concurrent advances in—

(3) actuator or robotic technologies that enable autonomous physical action; and

(4) modeling, simulation, and digital twin technologies that enable the testing, validation, and optimization of artificial intelligence-enabled physical systems.

(b) Membership.—The Working Group shall include not fewer than 12 and not more than 20 members, appointed by the Director from among the following:

(1) Industries, including at least 2 from each of the following, but not more than 6 from any of the following:

(E) Automation and manufacturing technology providers that integrate artificial intelligence and robotics.

(2) Faculty of accredited institutions of higher education, as well as representatives of independent research organizations with expertise in artificial intelligence, robotics, physical systems safety, cybersecurity, or security evaluation.

(3) Civil society organizations with demonstrated technical or public-interest expertise relevant to artificial intelligence, robotics, cybersecurity, or physical systems safety.

(4) Such other stakeholders with demonstrated subject-matter expertise relevant to the duties of the Working Group, including stakeholders in cybersecurity or critical infrastructure security, as the Director considers relevant.

(1) APPOINTMENT.—Members of the Working Group shall be appointed by the Director for terms of not more than 2 years.

(2) REAPPOINTMENT.—A member may be reappointed at the discretion of the Director.

(3) VACANCIES.—Any vacancy in the Working Group shall not affect its powers and shall be filled in the same manner as the original working group.

(1) identify and describe, using publicly available information to the extent practicable, plausible pathways by which the advances described in subsection (a) may introduce, alter, reduce, or intensify particular classes of physical systems safety or security concerns, including risks that may arise from the combination of model capability and the availability, affordability, and accessibility of relevant sensors and actuators;

(2) engage in technical discussions and research for the development of best practices and recommendations regarding whether, and under what conditions, analytical tools, benchmarks, or guidelines would be useful for assessing such physical safety and security concerns, including—

(A) approaches for collecting, synthesizing, and presenting publicly available trend data relating to the factors described in paragraphs (1) through (3) of subsection (a);

(B) potential evaluation methods or benchmark concepts for characterizing safety- and security-related behaviors, emergent behaviors, operational characteristics, and implications and risks for systems at the interface of software models and physical sensors or actuators; and

(C) potential indicators of evolving physical safety and security concerns that may warrant heightened attention by policymakers, developers, or deployers;

(3) identify voluntary best practices, where appropriate, to support the responsible development and deployment of artificial intelligence physical systems, including best practices relating to human-machine collaboration in operational environments, and in a manner that supports the identification, assessment, or mitigation of the safety and security concerns identified in paragraph (1); and

(4) identify opportunities to align the work of the Working Group with existing voluntary standards development activities of the National Institute of Standards and Technology.

(1) shall be carried out in a transparent, consultative, and nonregulatory manner consistent with the mission of the National Institute of Standards and Technology to advance innovation and industrial competitiveness;

(2) shall not require the disclosure of proprietary information, trade secrets, or classified information; and

(1) INITIAL REPORT.—Not later than 540 days after the date of the enactment of this Act, the Director shall publish a report summarizing the activities of the Working Group and any recommendations developed under subsection (c), including recommendations, if any, regarding the development of analytical tools, benchmarks, or guidelines described in subsection (c)(2).

(2) UPDATES.—Not less frequently than once every 2 years after publication of the initial report under paragraph (1), the Director may publish an updated report reflecting new stakeholder input, emerging research, and developments relevant to the concurrent advances described in subsection (a).

(3) PUBLIC AVAILABILITY.—Each report published under this subsection shall be made publicly available in written and digital formats and may include data visualizations or other resources useful to policymakers, researchers, and industry participants.