[House Hearing, 119 Congress]
[From the U.S. Government Publishing Office]
THE DIGITAL BATTLEFIELD: HOW ON-LINE
TERRORISTS USE THE INTERNET AND ON-
LINE NETWORKS FOR RECRUITMENT AND
RADICALIZATION
=======================================================================
HEARING
BEFORE THE
SUBCOMMITTEE ON
COUNTERTERRORISM,
LAW ENFORCEMENT, AND
INTELLIGENCE
OF THE
COMMITTEE ON HOMELAND SECURITY
HOUSE OF REPRESENTATIVES
ONE HUNDRED NINETEENTH CONGRESS
FIRST SESSION
__________
MARCH 4, 2025
__________
Serial No. 119-5
__________
Printed for the use of the Committee on Homeland Security
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]
Available via the World Wide Web: http://www.govinfo.gov
__________
U.S. GOVERNMENT PUBLISHING OFFICE
60-837 PDF WASHINGTON : 2025
-----------------------------------------------------------------------------------
COMMITTEE ON HOMELAND SECURITY
Mark E. Green, MD, Tennessee, Chairman
Michael T. McCaul, Texas, Vice Bennie G. Thompson, Mississippi,
Chair Ranking Member
Clay Higgins, Louisiana Eric Swalwell, California
Michael Guest, Mississippi J. Luis Correa, California
Carlos A. Gimenez, Florida Shri Thanedar, Michigan
August Pfluger, Texas Seth Magaziner, Rhode Island
Andrew R. Garbarino, New York Daniel S. Goldman, New York
Marjorie Taylor Greene, Georgia Delia C. Ramirez, Illinois
Tony Gonzales, Texas Timothy M. Kennedy, New York
Morgan Luttrell, Texas LaMonica McIver, New Jersey
Dale W. Strong, Alabama Julie Johnson, Texas, Vice Ranking
Josh Brecheen, Oklahoma Member
Elijah Crane, Arizona Pablo Jose Hernandez, Puerto Rico
Andrew Ogles, Tennessee Nellie Pou, New Jersey
Sheri Biggs, South Carolina Sylvester Turner, Texas
Gabe Evans, Colorado Vacant
Ryan Mackenzie, Pennsylvania Vacant
Brad Knott, North Carolina
Eric Heighberger, Staff Director
Hope Goins, Minority Staff Director
Sean Corcoran, Chief Clerk
------
SUBCOMMITTEE ON COUNTERTERRORISM, LAW ENFORCEMENT, AND INTELLIGENCE
August Pfluger, Texas, Chairman
Marjorie Taylor Greene, Georgia Seth Magaziner, Rhode Island,
Tony Gonzales, Texas Ranking Member
Morgan Luttrell, Texas J. Luis Correa, California
Gabe Evans, Colorado Daniel S. Goldman, New York
Ryan Mackenzie, Pennsylvania Pablo Jose Hernandez, Puerto Rico
Mark E. Green, MD, Tennessee (ex Nellie Pou, New Jersey
officio) Bennie G. Thompson, Mississippi
(ex officio)
Michael Koren, Subcommittee Staff Director
Brittany Carr, Minority Subcommittee Staff Director
C O N T E N T S
----------
Page
Statements
The Honorable August Pfluger, a Representative in Congress From
the State of Texas, and Chairman, Subcommittee on
Counterterrorism, Law Enforcement, and Intelligence:
Oral Statement................................................. 1
Prepared Statement............................................. 3
The Honorable Seth Magaziner, a Representative in Congress From
the State of Rhode Island, and Ranking Member, Subcommittee on
Counterterrorism, Law Enforcement, and Intelligence:
Oral Statement................................................. 5
Prepared Statement............................................. 6
The Honorable Bennie G. Thompson, a Representative in Congress
From the State of Mississippi, and Ranking Member, Committee on
Homeland Security:
Prepared Statement............................................. 7
Witnesses
Mr. Daveed Gartenstein-Ross, Ph.D., Senior Advisor on Asymmetric
Warfare, Foundation for Defense of Democracies:
Oral Statement................................................. 9
Prepared Statement............................................. 10
Mr. Aaron Zelin, Ph.D., Senior Research Fellow, The Washington
Institute for Near East Policy:
Oral Statement................................................. 22
Prepared Statement............................................. 24
Mr. Daniel Flesch, Senior Policy Analyst, Middle East and North
Africa, Allison Center for National Security, The Heritage
Foundation:
Oral Statement................................................. 29
Prepared Statement............................................. 30
Mr. Kurt Braddock, Ph.D., Assistant Professor, Public
Communication, American University:
Oral Statement................................................. 34
Prepared Statement............................................. 36
THE DIGITAL BATTLEFIELD: HOW ON-LINE TERRORISTS USE THE INTERNET AND
ON-LINE NETWORKS FOR RECRUITMENT AND RADICALIZATION
----------
Tuesday, March 4, 2025
U.S. House of Representatives,
Committee on Homeland Security,
Subcommittee on Counterterrorism,
Law Enforcement, and Intelligence,
Washington, DC.
The subcommittee met, pursuant to notice, at 2:14 p.m., at
Room 310, Cannon House Office Building, Hon. August Pfluger
[Chairman of the subcommittee] presiding.
Present: Representatives Pfluger, Greene, Luttrell,
Magaziner, Correa, Goldman, and Pou.
Also present: Representative Crane.
Mr. Pfluger. The Committee on Homeland Security,
Subcommittee on Counterterrorism and Intelligence will come to
order. Without objection, this committee may recess at any
point. Without objection, the gentleman from Arizona, Mr. Eli
Crane, is permitted to sit on the dais and ask questions of
witnesses.
The purpose of this hearing is to identify how foreign
terrorist organizations, like ISIS and other nefarious actors,
use the internet, on-line networks, and generative AI to
recruit and radicalize individuals to commit violence and
terrorist acts. This hearing will also explore policy solutions
to ensure law enforcement agencies have the necessary tools and
training to combat these threats. I now recognize myself for an
opening statement.
Good afternoon. Welcome to the first Counterterrorism and
Intelligence Subcommittee hearing of the 119th Congress. I
would like to begin by welcoming all the Members here today. I
would also like to take a second to welcome several new Members
to the committee: Representatives Gabe Evans, Ryan Mackenzie,
Nellie Pou, and Pablo Jose Hernandez. These Members will be
coming in after the vote series has just ended.
But last Congress, this subcommittee held various hearings
on pressing national security issues. Those national security
threats range from vulnerabilities posed by transnational
criminal organizations and known or suspected terrorists at our
Southern Border to the threats posed by the Chinese Communist
Party and the Iranian regime to U.S. homeland security. We will
continue to work on those issues and expand our scope to
include new and emerging threats.
This subcommittee also conducted critical oversight on
important DHS entities, such as the Office of Intelligence and
Analysis, and the ramifications of policy decisions made by the
Biden administration, such as our Nation's withdrawal from
Afghanistan. Our Nation faces many challenges and this
subcommittee must examine these threats through a clear and
objective lens. Since 9/11 and the formation of this committee,
this subcommittee has worked tirelessly to ensure that our
Nation understands the threats that face the United States and
that our law enforcement agencies have the proper tools to
safeguard our country and the American people. As Chairman of
this subcommittee, I am committed to working with everyone on
this committee to advance these goals.
Unfortunately, we did begin this year with a sober reminder
that the same ideology that radicalized 19 individuals to
hijack commercial airliners and fly them into the World Trade
Center, the Pentagon, and in Shanksville, Pennsylvania, drove
an individual to commit a heinous ISIS-inspired terror attack
that killed 14 people and injured dozens more on Bourbon Street
in New Orleans, Louisiana. Our thoughts and prayers are with
all of those affected by this tragedy.
Since the terror attack, the Nation has learned more about
that particular perpetrator's background. The individual,
Shamsud-Din Jabbar, is believed to have self-radicalized on-
line through various propaganda channels affiliated with ISIS.
The FBI has also confirmed that Jabbar's on-line search history
indicated he conducted extensive research into the terror
attacks that took place last year at a German Christmas market
and that he pledged his support to ISIS on his personal on-line
media accounts, including Facebook.
The case related to Jabbar is not an isolated incident of
an individual within the United States becoming radicalized on-
line by a foreign terrorist organization. On-line
radicalization is a significant terrorism threat that our
Nation currently faces. In fact, the Ranking Member and myself
have agreed that this would be the first hearing in a
bipartisan way that we want to tackle this issue. I hope
everyone knows that we have worked together on this and I hope
that we can focus and hone in on these particular issues right
here in a bipartisan way.
But at a recent Homeland Security and Government Affairs
Committee hearing, former FBI Director Wray highlighted the
severity of the threat when he stated, ``The greatest terrorism
threat to our homeland is posed by lone actors or small cells
of individuals who typically radicalize to violence on-line and
who primarily use easily accessible weapons to attack soft
targets.''
To provide greater context, between April 2021 and January
2025, there were over 50 indictments against individuals who
have worked to provide material support or carry out an attack
on behalf of foreign terrorist organizations like ISIS, al-
Qaeda, and Hezbollah. Almost all of these individuals have been
radicalized by terrorist media propagations.
Americans' ability to access propaganda from foreign
terrorist organizations is easier than ever. Groups like al-
Qaeda, ISIS, and Boko Haram are disseminating their propaganda
at lightning speed across the globe on platforms such as
Telegram and Redz. Hamas, Hezbollah, and other Iranian proxies
have utilized TikTok and other platforms to spread propaganda
in the West related to Israel's war with Hamas in hopes of
radicalizing individuals to join their quest to launch a global
intifada.
We saw this hateful propaganda mold the actions of
radicalized protesters at Columbia University and other
colleges across the United States. Another stark example of on-
line radicalization was seen last year when thousands of
Americans went on TikTok and attempted to rationalize Osama Bin
Laden's letter to America and victim blame our Nation for the
atrocities that took place on 9/11. Moreover, we have now seen
an increase in juvenile home-grown violent extremists inspired
by foreign terrorist organizations' media being shared on
youth-oriented platforms. For instance, a 16-year-old in Las
Vegas, Nevada, was arrested after threatening a lone-wolf
terrorist attack in support of ISIS and possessing components
and instructions to build a bomb. He wrote in a chat room that
he would be starting lone-wolf operations in Las Vegas against
the enemies of Allah.
There is no doubt that ISIS actively looks to inspire
younger individuals and frequently produces media tailored at
juveniles because they think they are more susceptible to
terrorist ideologies and more accessible due to their on-line
presence on a range of platforms. We must acknowledge and
confront this urgent threat head-on or we risk having another
incident like we did in New Orleans.
I hope today's discussion is the first of many productive
conversations on the enduring terrorism threats our Nation
faces and how this subcommittee can find legislative solutions
to overcome these challenges. I also hope that while we may
disagree here in a partisan way on some of the things that have
happened in the last couple of years, I hope that we can really
focus and limit blame, that we can take the context and the
facts. I hope that the Ranking Member knows that I took
seriously his inputs to having this hearing, to holding this
hearing, to focusing on this not just as a sign of goodwill,
but also as a sign that this committee, of all the committees
in Congress, that this committee right here was formed after
the worst terrorist attack in our country, and that the sole
reason we have this committee is to ensure that the oversight
of those entrusted with national security, of homeland
security, do it correctly.
[The statement of Chairman Pfluger follows:]
Statement of Chairman August Pfluger
March 4, 2025
Good afternoon, and welcome to the first Counterterrorism and
Intelligence Subcommittee hearing of the 119th Congress.
I want to begin by welcoming all the Members here today. I'd also
like to take a second to welcome several new Members to the committee.
Representatives Gabe Evans, Ryan Mackenzie, Nellie Pou, and Pablo Jose
Hernandez.
Last Congress, this subcommittee held various hearings on pressing
national security issues.
Those national security threats ranged from the vulnerabilities
posed by transnational criminal organizations and known or suspected
terrorists at our Southern Border to the threats posed by the Chinese
Communist Party and the Iranian regime to U.S. homeland security.
We will continue to work on those issues and expand our scope to
include new and emerging threats.
This subcommittee also conducted critical oversight work on
important DHS entities, such as DHS's Office of Intelligence and
Analysis, and the ramifications of policy decisions made by the Biden
administration, such as our Nation's withdrawal from Afghanistan.
Our Nation faces many challenges, and this subcommittee must
examine these threats through a clear and objective lens.
Since 9/11 and the formation of this committee, this subcommittee
has worked tirelessly to ensure that our Nation understands the threats
that face the United States and that our law enforcement agencies have
the proper tools to safeguard our country and the American people.
As Chairman of this subcommittee, I am committed to working with
everyone to advance these goals.
Unfortunately, we began this year with a sober reminder that the
same ideology that radicalized 19 individuals to hijack commercial
airliners and fly them into the World Trade Center, the Pentagon, and
Shanksville, Pennsylvania, drove an individual to commit a heinous
ISIS-inspired terror attack that killed 14 people and injured dozens
more on Bourbon Street in New Orleans, Louisiana.
Our thoughts and prayers are with all of those affected by this
tragedy.
Since the New Orleans terror attack, the Nation has learned more
about the perpetrator's background.
The individual, Shamsud-Din Jabbar, is believed to have self-
radicalized on-line through various propaganda channels affiliated with
ISIS.
The FBI has also confirmed that Jabbar's on-line search history
indicated he conducted extensive research into the terror attacks that
took place last year at a German Christmas market and that he pledged
his support for ISIS on his personal Facebook account.
The case related to Jabbar is not an isolated incident of an
individual within the United States becoming radicalized on-line by a
foreign terrorist organization.
On-line radicalization is a significant terrorism threat our Nation
currently faces.
At a recent Homeland Security and Government Affairs Committee
hearing, former FBI Director Wray highlighted the severity of this
threat when he stated: ``[t]he greatest terrorism threat to our
homeland is posed by lone actors or small cells of individuals who
typically radicalize to violence on-line, and who primarily use easily
accessible weapons to attack soft targets.''
To provide greater context, between April 2021 and January 2025,
there were over 50 indictments against individuals who have worked to
provide material support or carry out an attack on behalf of foreign
terrorist organizations like ISIS, al-Qaeda, and Hezbollah.
Almost all of these individuals had been radicalized by terrorist
media propaganda.
Americans' ability to access propaganda from foreign terrorist
organizations is easier than ever.
Groups like al-Qaeda, ISIS, and Boko Haram are disseminating their
propaganda at lightning speed across the globe on platforms, such as
Telegram and Redz.
Hamas, Hezbollah, and other Iranian proxies have utilized TikTok
and other platforms to spread propaganda in the West related to
Israel's war with Hamas in hopes of radicalizing individuals to join
their quest to launch a global intifada.
We saw this hateful propaganda mold the actions of radicalized
protesters at Columbia University and other colleges across the United
States.
Another stark example of on-line radicalization was seen last year
when thousands of Americans went on TikTok and attempted to rationalize
Osama bin Laden's letter to America and victim-blame our Nation for the
atrocities that took place on 9/11.
Moreover, we have now seen an increase in juvenile home-grown
violent extremists inspired by foreign terrorist organizations' media
being shared on youth-oriented platforms.
For instance, a 16-year-old in Las Vegas, Nevada, was arrested
after threatening a lone-wolf terrorist attack in support of ISIS and
possessing components and instructions to build a bomb. He wrote in a
chatroom that he would be ``starting lone-wolf operations in Las Vegas
against enemies of Allah.''
There is no doubt that ISIS actively looks to inspire younger
individuals and frequently produces media tailored at juveniles because
they think they are more susceptible to terrorist ideologies and more
accessible due to their on-line presence on a range of platforms.
We must acknowledge and confront this urgent threat head-on, or we
risk having another terrorist attack like New Orleans take place here
in the United States.
I hope today's discussion is the first of many productive
conversations on the enduring terrorism threats our Nation faces and
how this subcommittee can find legislative solutions to overcome these
challenges.
I thank our distinguished panel for their testimony today.
With that, I yield to the gentleman from Rhode Island, Ranking
Member Magaziner, for his opening statement.
Mr. Pfluger. With that, I thank our distinguished panel. I
will introduce you shortly. I would like to yield to the
gentleman from Rhode Island, my friend and the Ranking Member,
Mr. Magaziner, for his opening statement.
Mr. Magaziner. I thank Chairman Pfluger for his kindness
and also for having this hearing today. I think it is
appropriate that this is our first hearing in this subcommittee
of the new Congress. It is a topic I have been interested in
addressing for some time and I know that he is very interested
in it as well, because increasingly, on-line radicalization is
one of the largest threats that we face to the homeland. I want
to thank our witnesses for being here as well.
On New Year's Day, 14 innocent people lost their lives and
more than 50 were injured when a terrorist drove a pickup truck
into a crowd on Bourbon Street in New Orleans. In their
investigation after the attack, the FBI found that the attacker
had been radicalized on-line through information about ISIS and
engagement with ISIS content. According to the Department of
Homeland Security's Office of Intelligence and Analysis, since
August 2023, so just a year-and-a-half, law enforcement has
disrupted 5 plots in which juveniles were radicalized by
foreign terrorist organizations on-line and mobilized to plan
terror attacks. Five times in just the last 5 years--sorry,
just the last year-and-a-half.
While foreign terrorist organizations like ISIS and al-
Qaeda exploit the internet to spread propaganda, they,
unfortunately, are not the only bad actors who use on-line
platforms to inspire individuals to violence. In September, the
Justice Department indicted 2 individuals who are leaders of
the so-called Terrorgram Collective and a domestic neo-fascist
terrorist organization that used the Telegram platform to plot
the murder of Federal officials, including a U.S. senator, U.S.
district court judge, and a U.S. attorney, who they considered
to be ``enemies of the white race.'' Terrorgram provided
guidance and instructions for terror attacks, encouraging the
use of bombs to attack critical infrastructure, such as
government buildings and energy facilities as well.
In December 2023, the FBI arrested an Arizona man for using
social media to incite a religiously-motivated terror attack in
Australia, an ambush shooting that killed 6 people, including 2
police officers. This individual and his wife and brother were
motivated by an extreme religious ideology and attacked and
killed victims simply for being law enforcement officers. We
also know that on-line radicalization played a key role in the
white supremacist extremist attack in Buffalo in 2022 when an
18-year-old terrorist killed 10 people, all of whom were
African American. An investigative report by the Office of the
New York State Attorney General found that this individual
learned about the so-called Great Replacement Theory via memes,
an ideology that he himself cited as the basis for his racist
terror attack.
So these are just a few examples of domestic violent
extremists' use of the internet and digital content for
recruitment and radicalization. Of course, foreign terrorist
organizations, as has been noted, are using the same exact
strategies, unfortunately, often to lethal effect.
Congress is not powerless to stop this. Social media
companies and gaming companies can and should moderate violent
extremist content on-line with proper consideration for
protected speech. They should also cooperate with law
enforcement investigations.
To be clear, American citizens have a First Amendment right
to hold and espouse extreme ideologies. However, there is no
First Amendment right to plot acts of violence. There is no
First Amendment right to recruit others into committing acts of
violence. Foreign terrorists have no First Amendment rights of
any sort. So social media companies and gaming companies are in
no way compelled to allow terrorists or criminals to use their
platforms to incite violence. In fact, they should face
consequences if they knowingly do so. That is something that
Congress needs to act on.
I worked with Ranking Member Thompson to request that the
Government Accountability Office examine how social media and
gaming companies report and mitigate on-line content that
promote violent extremism, and the extent to which the FBI and
DHS have developed goals and strategies for sharing information
on the threat of violent extremism on social media and gaming
companies. The GAO produced its findings in January 2024, which
included neither--but neither DHS nor FBI have developed such a
strategy. They must do so, and the strategy must be
comprehensive to include all forms of on-line radicalization
and incitement of violence, whatever the ideology behind it may
be.
It is my hope that this subcommittee will continue today's
conversation with today's experts, but also with
representatives of the social media and on-line gaming
industries and the relevant law enforcement agencies so that we
can explore real answers, including legislative solutions, in
combating on-line radicalization.
I thank the Chairman again for his bipartisan approach to
this topic. I am excited to work together on it and I yield
back.
[The statement of Ranking Member Magaziner follows:]
Statement of Ranking Member Seth Magaziner
March 4, 2025
On New Year's Day, 14 innocent people lost their lives and over 50
were injured when a terrorist drove a pickup truck into a crowd on
Bourbon Street in New Orleans. In their investigation after the attack,
the FBI found that the attacker had been radicalized on-line through
information about ISIS and engagement with ISIS content.
According to the Department of Homeland Security's Office of
Intelligence and Analysis, since August 2023, law enforcement has
disrupted 5 plots in which juveniles were radicalized by foreign
terrorist organizations on-line and mobilized to plan terror attacks.
While foreign terrorist organizations like ISIS and al-Qaeda exploit
the internet to spread propaganda, they unfortunately aren't the only
bad actors who use on-line platforms to inspire individuals to
violence.
In September, the Justice Department indicted Dallas Humber and
Matthew Allison, 2 leaders of the Terrorgram Collective--a far-right,
neo-fascist terrorist organization--that used the Telegram platform to
plot the murder of Federal officials such as a U.S. Senator, a U.S.
district court judge, and a U.S. attorney who they considered enemies
of the white race. Terrorgram provided guidance and instructions for
terrorist attacks encouraging the use of bombs to attack critical
infrastructure such as government buildings and energy facilities.
In December 2023, the FBI arrested an Arizona man for using social
media to incite a ``religiously-motivated terrorist attack,'' in
Australia--an ambush shooting that killed 6 people including 2 police
officers. Gareth Train, his brother Nathaniel Train and Nathanial's
wife Stacey Train, motivated by a Christian extremist ideology--
attacked and killed the victims simply for being law enforcement
officers.
We also know that on-line radicalization played a key role in the
white supremacist extremist attack in Buffalo in 2022, when an 18-year-
old terrorist killed 10 people--all of whom were African American. An
investigative report by the Office of the New York State Attorney
General found that the Buffalo terrorist learned about the so-called
``Great Replacement Theory'' via memes--an ideology he cited as the
basis for his racist terrorist attack.
These are just a few examples of domestic violent extremists' use
of the internet and digital content for recruitment and radicalization.
Congress is not powerless to stop this.
Social media companies can and should moderate violent extremist
content on-line with proper consideration for protected speech. They
should also cooperate with law enforcement investigations.
To be clear: American citizens have a first amendment right to hold
and espouse extreme ideologies.
However:
There is no first amendment right to plot acts of violence
There is no first amendment right to recruit others to
commit acts of violence
Foreign terrorists have no first amendment rights of any
sort
And social media companies are in no way compelled to allow
terrorists or criminals to use their platforms.
In fact, they should face consequences if they knowingly do so, and
that is something that Congress needs to act on.
I worked with Ranking Member Thompson to request that the
Government Accountability Office (GAO) examine how social media and
gaming companies report mitigating on-line content that promotes
domestic violent extremism and the extent to which the FBI and DHS have
developed goals and strategies for sharing information on the threat of
domestic violent extremists with social media and gaming companies.
GAO produced its findings in January 2024, which included that
neither DHS or FBI have developed such a strategy. They must do so, and
the strategy must be comprehensive to include all forms of on-line
radicalization, any incitement of violence, whatever the ideology
behind it may be. It is my recommendation that the subcommittee
continue today's conversation with both representatives from the social
media and gaming industry and the agencies so that we can explore real
answers in combating on-line radicalization.
Mr. Pfluger. I thank the Ranking Member. Other Members of
the committee are reminded that opening statements may be
submitted for the record.
[The statement of Ranking Member Thompson follows:]
Statement of Ranking Member Bennie G. Thompson
March 4, 2025
Terrorists' use of on-line platforms to spread ideology, recruit
members, and plan attacks is not novel. Neither is it new that
individuals with extreme views seek out virtual communities of like-
minded people and use the internet to express grievances and plan and
live-stream attacks.
Nearly 10 years ago, minutes prior to a failed terrorist attack at
the Curtis Culwell Center in Garland, Texas, one of the perpetrators
posted a tweet with the hashtag #texasattack. One month following that
attack, 9 African-American parishioners at Mother Emanuel in South
Carolina were killed by a white-supremacist terrorist who self-
radicalized on-line.
Next week will mark 5 years since a terrorist live-streamed on
Facebook the attacks in which he heinously murdered 51 people at
mosques in Christchurch, New Zealand. Unfortunately, the list of
examples where extremism on-line and real-world violence converge goes
on and on.
Although extremists have recently migrated to encrypted platforms,
they largely use the same social media platforms as the average
internet user. Research indicates that up to a third of internet users
experience hate speech on-line.
The data is even more grim when it comes to on-line multiplayer
games. According to a NYU study on extremists' exploitation of gaming
platforms, 41 percent of survey respondents came across statements
portraying a particular ethnic, gender, or religious group as inferior.
People encountered statements that included support for
reprehensible ideas such as: violence against women is justified, a
particular race or ethnicity should be expelled or eliminated, and
using violence is justified or necessary to achieve a political aim.
A separate study by ADL found that 75 percent of teens and pre-
teens experienced harassment in on-line multiplayer games in 2023 and
that women and Black or African American gamers were the most harassed
because of their identity.
This problem will only become more acute as the current owner of X,
Elon Musk, is an extremist himself who has created an environment on
his platform where pro-Nazi accounts that share Hitler speeches
flourish.
Moreover, Musk profits from this kind of hateful content. The
Center for Countering Digital Hate found that X ``stands to make up to
$19 million a year from ads on just 10 toxic reinstated accounts,'' and
that accounts spewing hateful anti-LGBTQ rhetoric generate up to $6.4
million per year for X in ad revenue.
Musk will not adopt any counter-extremist content policies or
measures because that would affect his bottom line--never mind the fact
that he also holds the keys to the Federal Government and will block
any engagement between Government and industry on the issue.
Nevertheless, Committee Democrats will continue to examine the
threats posed by on-line radicalization and engage the willing and
socially responsible on-line and gaming platforms on solutions.
Mr. Pfluger. I am pleased to have 4 distinguished witnesses
before us today on this very important topic, some of whom have
testified before this committee multiple times. I ask that our
panel of witnesses please rise and raise their right hand.
[Witnesses sworn.]
Mr. Pfluger. Thank you. Let the record reflect that the
witnesses have answered in the affirmative.
I will now formally introduce our panel of witnesses,
starting with Dr. Daveed Gartenstein-Ross serves as a senior
advisor on asymmetric warfare at the Foundation for Defense of
Democracies. He is also the founder and chief executive officer
of Valens Global. Welcome.
I would like to introduce now Dr. Aaron Zelin, who
currently serves as a senior research fellow at the Washington
Institute where he directs the Islamic State Worldwide Activity
Map Project. I am also a graduate. As a fellow of the
Washington Institute, it is good to see you.
I would like to introduce Mr. Daniel Flesch, who serves as
a senior policy analyst for Middle East and North Africa at the
Heritage Foundation. He previously served as an IDF
paratrooper. Welcome.
Last, Dr. Kurt Braddock is an assistant professor of public
communication in the School of Communication at American
University. His research focuses on the strategies used by
violent extremist groups to recruit and radicalize individuals
targeted by propaganda. I thank everyone for being here. Thank
you for your expertise. Thank you for your time.
We will now go to your opening statements. I know you have
submitted written statements and if you would be so kind to
summarize that at the 5-minute mark you will see the timer that
counts down on that.
With that, Dr. Gartenstein-Ross, you are recognized for
your opening statement.
STATEMENT OF DAVEED GARTENSTEIN-ROSS, PH.D., SENIOR ADVISOR ON
ASYMMETRIC WARFARE, FOUNDATION FOR DEFENSE OF DEMOCRACIES
Mr. Gartenstein-Ross. Thank you. Chairman Pfluger, Ranking
Member Magaziner, distinguished Members of the subcommittee, on
behalf of the Foundation for Defense of Democracies, thank you
for the opportunity to testify on the important topic that the
subcommittee is addressing today.
In my written testimony, I discuss what my colleagues at
Valens Global and I call composite violent extremism, or CoVE.
It describes individuals who blend elements from multiple
ideologies, grievances, and sentiments without adhering to a
single well-defined belief system. CoVE was not created by the
on-line space. Ideologically-fluid extremists have existed for
decades, but CoVE's increasing prevalence is tied to the
digital age. The internet has transformed how violent extremist
beliefs form, spread, and evolve.
We've seen this before. Technological shifts have
repeatedly reshaped terrorism and we are now living through the
next great technological shift, one that is more momentous than
the advent of the internet itself. This shift is the rapid
development of artificial intelligence, especially generative
AI.
In 2018, I wrote an article titled ``Terrorists are Going
to Use Artificial Intelligence.'' At the time the idea was
debated, but from our vantage point today, it's not a
prediction. It's happening. We have seen a recurring pattern in
terrorists employing new technologies. Terrorists first fail
with a new technology, often giving way to iterative
experimentation, adaptation, and eventual success. Social
media, encryption, and drones all followed this trajectory,
where initial crude attempts evolved into sophisticated
applications, forcing governments to scramble for
countermeasures. I call this pattern the violent non-state
actor technology adoption curve, a consistent cycle in which
terrorists initially struggle with new tools, but grow more
adept at using them over time. We have seen this cycle unfold
time and again. AI is not an exception. It is the next phase.
Terrorists don't need cutting-edge AI research labs to
weaponize artificial intelligence. They only need access to the
same widely-available AI tools that businesses and individuals
are already using. As AI becomes more powerful, it will lower
barriers to entry even further.
These are not hypothetical threats for the distant future.
They are real today. Bad actors are already experimenting with
AI to illustrate ways that AI and large language models can
influence terrorist operations, there's propaganda and
psychological warfare. AI can generate hyper-personalized
extremist content, create deepfake recruitment videos, and use
generative AI in recruitment agents that engage recruits
dynamically. There's training and operational planning. Large
language models can generate tactical handbooks, simulate
security vulnerabilities, and coach extremists step by step
through attack planning.
Some may assume that safeguards will prevent AI from being
exploited. They won't. Large language models are already easy
to jailbreak, that is to manipulate into bypassing built-in
safeguards against generating prohibited or harmful conduct.
Even platforms with thoughtful built-in restrictions can be
manipulated. I don't want to explain the tactics to do so
openly in this hearing, but I would be happy to generate basic
jailbreaking techniques to any member who is interested.
Also, terrorists don't need ChatGPT. Other models, like
China's DeepSeek, have fewer safeguards and are easier to
weaponize. Governments and tech companies cannot rely on safety
filters alone. A reactive approach is a losing strategy.
In the over 20 years that I have worked on
counterterrorism, one pattern has been painfully clear: we are
too often reactive rather than proactive. The same failure of
imagination that the 9/11 Commission warned of is still at
play. If we fail to recognize how AI will reshape terrorism and
extremism, we risk being blindsided in ways that could dwarf
the digital era mistakes of the past. That's why legislative
efforts, such as the Gen AI Terrorism Threat Assessment Act are
important. This bill would ensure that DHS conducts regular
assessments of how terrorist organizations use encrypted,
cloud-based communication platforms.
We need to monitor these trends rather than react when it's
too late. If we aren't tracking how terrorists adapt to the
technologies that are reshaping the world we will always be
behind. We should act before the next breakthrough becomes the
next disaster.
Again, I appreciate the opportunity to testify and I look
forward to your questions.
[The prepared statement of Mr. Gartenstein-Ross follows:]
Prepared Statement of Daveed Gartenstein-Ross
March 4, 2025
Chairman Pfluger, Ranking Member Magaziner, and distinguished
Members of the subcommittee, on behalf of the Foundation for Defense of
Democracies, I am honored to appear before you to discuss the important
topic of the digital battlefield, exploring how terrorists use the
internet.
This testimony will focus on an important trend in on-line
radicalization that my colleagues and I have been tracking for several
years, which we refer to as composite violent extremism (CoVE).
Composite violent extremism describes cases where individuals engage in
or support acts of terrorism or targeted violence despite adhering to a
mixture of ideological beliefs, grievances, and prejudices that do not
fit neatly into traditional categories of extremism. Unlike
conventional ideological radicalization, where individuals are drawn
into well-defined extremist movements, CoVE involves the blending of
disparate ideological elements, often shaped by on-line environments,
personal grievances, and cultural influences.
Understanding composite violent extremism is important because its
growing prevalence presents new challenges for counterterrorism
efforts, law enforcement, and policy makers--and because it provides
new opportunities for violent extremists. Traditional frameworks for
identifying and countering violent extremism often rely on clear
ideological categories, yet instances of composite violent extremism do
not fit neatly into these classifications. Extremists increasingly draw
from multiple, sometimes contradictory, ideological sources.
This testimony will proceed in two major parts. First, it explains
the concept of composite violent extremism in greater detail, outlining
its defining characteristics and how it differs from traditional forms
of ideological radicalization. Second, the testimony will explore the
specific challenges that CoVE presents for terrorism prevention
efforts. Understanding these dynamics is essential for developing more
effective strategies to counter evolving threats in the digital
battlefield.
______
In April 2022, Frank James opened fire on a New York subway train
during rush hour, injuring 29 people. James's voluminous on-line
writings left authorities, experts, and the media scratching their
heads about his motives. Some reports called him a black nationalist,
while others pointed to more disparate racist and misogynist ideas.
James expressed hatred toward white people, Jews, and Latinos as well
as anti-American sentiments and political grievances--none of which
amounted to a coherent ideology or aligned with any distinct
ideological movement. Complicating matters, James also struggled with
mental illness.
Was James's shooting spree an act of violent extremism? If so, what
kind? Scholars and practitioners have grappled with the increasing
prominence of attackers like James who are ideologically idiosyncratic
or even incoherent. In Congressional testimony from 2022, for example,
FBI Director Christopher Wray spotlighted extremists who hold a ``weird
hodgepodge blend of ideologies,'' noting that this trend is producing
challenges in ``trying to unpack what are often sort-of incoherent
belief systems, combined with kind-of personal grievances.'' Indeed, it
can be difficult to unpack such belief systems because practitioners
have often lacked the conceptual tools necessary to comprehend
extremists who defy neat categorization.
FBI Director Wray described this phenomenon as ``salad bar''
extremism.\1\ For well over half a decade, researchers have been trying
to explain this phenomenon, employing terms like ``ideological
convergence,'' ``fused extremism,'' ``hybrid ideologies,'' ``fringe
fluidity,'' ``ideology a la carte,'' and ``choose your own adventure''
extremism--all of which have slightly different meanings and some of
which only loosely relate to the FBI's concept of salad bar
extremism.\2\ Indeed, cases that fit the so-called salad bar paradigm
(for which I offer an alternative terminology and concrete ways to
understand) are challenging to conceptualize and categorize in large
part because it can be difficult to discern motives amid complex
interplays of disparate beliefs, interests, prejudices, grievances, and
personal risk factors. This testimony thus explains the concept of
composite violent extremism (CoVE).\3\
---------------------------------------------------------------------------
\1\ FBI Director Christopher A. Wray, ``A Review of the President's
Fiscal Year 2023 Funding Request for the Federal Bureau of
Investigation,'' Testimony Before the Senate Appropriations
Subcommittee on Commerce, Justice, Science, and Related Agencies, May
25, 2022. (https://www.appropriations.senate.gov/hearings/a-review-of-
the-presidents-fiscal-year-2023-funding-request-for-the-federal-bureau-
of-investigation); FBI Director Christopher A. Wray, ``Threats to the
Homeland, Testimony Before the Senate Homeland Security and Government
Affairs Committee, September 24, 2020. (https://www.hsgac.senate.gov/
hearings/--14-2020-threats-to-the-homeland).
\2\ See Julien Bellaiche, ``Connecting the Fringes: Neo-Nazi
Glorification of Salafi-Jihadi Representations Online,'' Global Network
on Extremism and Terrorism, August 24, 2021. (https://gnet-
research.org/2021/08/24/connecting-the-fringes-neo-nazi-glorification-
of-salafi-jihadi-representations-on-line); Ariel Koch, ``The ONA
Network and the Transnationalization of Neo-Nazi-Satanism,'' Studies in
Conflict & Terrorism, January 12, 2022, pages 1,172-1,199. (https://
www.tandfon-line.com/doi/full/10.1080/1057610X.2021.2024944); Jesse J.
Norris, ``Idiosyncratic Terrorism: Disaggregating an Undertheorized
Concept,'' Perspectives on Terrorism, Volume 14, Number 3, June 2020.
(https://www.jstor.org/stable/26918296); Milo Comerford and Sasha
Havlicek, ``Mainstreamed Extremism and the Future of Prevention''
(London: Institute for Strategic Dialogue, 2021). (https://
www.isdglobal.org/wp-content/uploads/2021/10/ISD-Mainstreamed-
extremism-and-the-future-of-prevention-3.pdf); Daveed Gartenstein-Ross
and Madeleine Blackman, ``Fluidity of the Fringes: Prior Extremist
Involvement as a Radicalization Pathway,'' Studies in Conflict &
Terrorism, January 7, 2019, pages 555-578. (https://www.tandfon-
line.com/doi/full/10.1080/1057610X.2018.1531545); Paige Pascarelli,
``Ideology a la Carte: Why Lone Actor Terrorists Choose and Fuse
Ideologies,'' Lawfare, October 2, 2016. (https://www.lawfareblog.com/
ideology-%C3%A0-la-carte-why-lone-actor-terrorists-choose-and-fuse-
ideologies); Kurt Braddock, Brian Hughes, and Cynthia Miller-Idriss,
``The Post-9/11 Fight Against Extremism Must Take On Propagandists'
Tricks, Not Just Ideology,'' MarketWatch, September 11, 2021. (https://
www.marketwatch.com/story/the-post-9-11-fight-against-extremism-must-
expand-to-attitudinal-inoculation-11631285779).
\3\ I would like to acknowledge the work of 3 colleagues with whom
I developed the concept of CoVE: Emelie Chace-Donahue, Madison Urban,
and Andrew Zammit. See Daveed Gartenstein-Ross, Andrew Zammit, Emilie
Chace-Donahue & Madison Urban, ``Composite Violent Extremism:
Conceptualizing Attackers Who Increasingly Challenge Traditional
Categories of Terrorism,'' Studies in Conflict & Terrorism, March 29,
2023. (https://www.tandfon-line.com/doi/full/10.1080/
1057610X.2023.2194133).
---------------------------------------------------------------------------
The increased prominence of these attacks that challenge
established categories of violent extremism are of clear interest to
this subcommittee. My contribution today is to offer a
conceptualization of these vexing varieties of violent extremism and
offer potential explanations for their apparently increased frequency.
background
In recent years, governments have begun expanding the scope of
counterterrorism and prevention efforts to address ideologically
ambiguous cases. DHS's 2019 Strategic Framework for Countering
Terrorism and Targeted Violence (the CTTV framework) and the United
Kingdom's mixed, unstable, and unclear (MUU) classification enable
discussion of traditional terrorism alongside cases where an attacker
lacks a clearly discernible ideology but where the intent and tactics
resemble terrorism. Coupling terrorism with this more ambiguously
motivated violence is in part intended to strengthen prevention
efforts. As DHS's strategy stated, these phenomena ``overlap,
intersect, and interact as problems'' and thus ``necessitate a shared
set of solutions.''\4\
---------------------------------------------------------------------------
\4\ U.S. Department of Homeland Security, Strategic Framework for
Countering Terrorism and Targeted Violence, September 2019, page 11.
(https://www.dhs.gov/sites/default/files/publications/
19_0920_plcy_strategic-framework-countering-terrorism-targeted-
violence.pdf).
---------------------------------------------------------------------------
DHS's 2019 CTTV framework was the first time a department-level
U.S. strategy recognized terrorism and targeted violence within the
same threat landscape. According to the framework, ``targeted violence
refers to any incident of violence in which a known or knowable
attacker selects a particular target prior to the violent attack''
(though it is worth noting that the framework also recommends the
promulgation of an updated definition of the phenomenon of targeted
violence).\5\ The CTTV framework goes on to note that ``unlike
terrorism, targeted violence includes attacks otherwise lacking a
clearly discernible political, ideological, or religious motivation,
but that are of such severity and magnitude as to suggest an intent to
inflict a degree of mass injury, destruction, or death commensurate
with known terrorist tactics.''\6\
---------------------------------------------------------------------------
\5\ Ibid., page 4.
\6\ Ibid.
---------------------------------------------------------------------------
The U.K.'s Prevent Programme addresses traditional terrorism but
goes a step beyond the CTTV framework by creating a specific category
for less clearly discernible ideologies. The Prevent Programme includes
the MUU classification for individuals referred to the program whose
ideology or motivations challenge traditional categorizations.\7\
According to the program, MUU applies to cases ``where the ideology
presented involves a combination of elements from multiple ideologies
(mixed), shifts between different ideologies (unstable), or where the
individual does not present a coherent ideology yet may still be
vulnerable to being drawn into terrorism (unclear).''\8\
---------------------------------------------------------------------------
\7\ Mark Townsend, ``Anti-Terrorism Programme Must Keep Focus on
Far Right, Say Experts,'' Guardian (UK), May 22, 2022. (https://
www.theguardian.com/uk-news/2022/may/22/anti-terrorism-programme-must-
keep-focus-on-far-right-say-experts)
\8\ United Kingdom Home Office, ``User Guide to: Individuals
Referred to and Supported Through the Prevent Programme, England and
Wales,'' December 5, 2024. (https://www.gov.uk/government/publications/
user-guide-to-individuals-referred-to-and-supported-through-the-
prevent-programme-england-and-wales/user-guide-to-individuals-referred-
to-and-supported-through-the-prevent-programme-england-and-wales).
---------------------------------------------------------------------------
Australian authorities have similarly demonstrated concern about
ideologically unclear attacks, as shown by the proliferation of fixated
threat assessment centers within State police counterterrorism
functions. Fixated threat assessment centers aim to protect the public
from individuals with an ``obsessive preoccupation with a person or
some idiosyncratic cause, which is pursued to a pathological degree''
that can result in violence.\9\ These centers were not initially viewed
as having a counterterrorism function, as they were largely concerned
with individuals who were ``fixated on a highly personal cause or
grievance'' rather than being ideologically motivated.\10\ The United
Kingdom established a Fixated Threat Assessment Centre (FTAC) in 2006
due to persistent threats to the Royal Family from unstable
individuals. This FTAC remained largely separate from the country's
counterterrorism efforts. Australia's adoption of fixated threat
assessment centers, which began with the establishment of a center in
Queensland in 2013, was initially modeled on the United Kingdom's
approach but increasingly came to involve a more explicit
counterterrorism role.\11\ For example, in 2017 both the New South
Wales Police and Victoria Police established fixated threat assessment
centers within their counterterrorism commands, showing that Australian
counterterrorism approaches were broadening beyond a concern over
individuals with clear ideological motivations.\12\ In this way, the
recognition that counterterrorism tools were relevant in this parallel
context resembles the U.S. Department of Homeland Security's inclusion
of targeted violence in the CTTV framework.
---------------------------------------------------------------------------
\9\ Michele T. Pathe, Timothy Lowry, Debbie J. Haworth, Danae M.
Webster, Melodie J. Mulder, Paul Winterbourne, and Colin J. Briggs,
``Assessing and Managing the Threat Posed by Fixated Persons in
Australia,'' The Journal of Forensic Psychiatry & Psychology, Volume
26, Number 4, May 5, 2015, pages 425-438. (https://www.tandfon-
line.com/doi/full/10.1080/14789949.2015.1037332).
\10\ Ibid., page 426.
\11\ Ibid., pages 425-438.
\12\ Paul Farrell, ``NSW Police Establish `Fixated Persons' Unit to
Help Counter Lone Wolf Terror Attacks,'' The Guardian (UK), April 25,
2017. (https://www.theguardian.com/australia-news/2017/apr/26/nsw-
police-establish-fixated-persons-unit-to-help-counter-lone-wolf-terror-
attacks); Premier of Victoria, ``New Threat Assessment Centre to Keep
Victorians Safe,'' October 3, 2017. (https://www.premier.vic.gov.au/
new-threat-assessment-centre-to-keep-victorians-safe)
---------------------------------------------------------------------------
These policy and operational frameworks are significant steps, but
more attention needs to be paid to understanding trends and common
characteristics within targeted violence, MUU, and fixated threats. The
CoVE framework, which I will now explain, accomplishes this. In 2022-
23, I worked with a team of highly competent researchers--with the
principal contributors being Emelie Chace-Donahue, Madison Urban, and
Andrew Zammit--to better understand how cases like these fit within the
violent extremism landscape. This team sought to develop a new
conceptual framework with a clear overarching concept and a typology of
subordinate concepts that disaggregated the different forms of violent
extremism being observed. We sought to engage in a careful process of
conceptualization, which began by identifying 94 cases of interest
through existing datasets of violent extremists and mass shooters
(including PIRUS and The Violence Project) along with less structured
searches through Google. I will now outline our overarching
conclusions.\13\
---------------------------------------------------------------------------
\13\ I believe that methodological explanations are out of place in
Congressional testimony. For those who are interested in our
methodological approach, see Daveed Gartenstein-Ross, Andrew Zammit,
Emelie Chace-Donahue & Madison Urban, ``Composite Violent Extremism:
Conceptualizing Attackers Who Increasingly Challenge Traditional
Categories of Terrorism,'' Studies in Conflict & Terrorism, March 29,
2023, pages 5-7. (https://www.tandfon-line.com/doi/full/10.1080/
1057610X.2023.2194133).
---------------------------------------------------------------------------
composite violent extremism and its subordinate categories
The framework's overarching concept is composite violent extremism,
or CoVE. Composite, a term denoting something made up of various parts
or elements, encompasses the concept of an amalgamated extremist
outlook at the broadest level. The three subordinate categories in the
typology demonstrate the different types of amalgamation that composite
violent extremism covers. However, it is important to clarify what the
concept of composite violent extremism does not cover.
At one end of a spectrum, composite violent extremism does not
encompass violent extremists who possess a discernible ideology that is
not combined with other sentiments or cooperation with adherents of
other ideologies. This means the concept does not cover the phenomenon
of fringe fluidity, wherein an extremist switches in full from one
ideology to another. It also means the concept potentially does not
cover cases that would be examples of unstable under the MUU criteria,
unless the ideological shifts did not encompass the wholesale adoption
of new ideology but instead involved amalgamation in some form.
At the other end, composite violent extremism does not cover
violent actors who do not remotely demonstrate any ideological
adherence, which meant that many mass shooters are excluded. The
purpose of developing the overarching concept was to capture the
apparent new paradigm of violent extremism that has clashed with
existing categories used by scholars and practitioners, not to simply
broaden the concept of violent extremism so much that it would
encompass all mass killers with multiple non-ideological grievances or
motivations. The concept is therefore deliberately described as
composite violent extremism rather than composite violence.
Within this conception of composite violent extremism are 3
distinct categories based on levels of ideological discernibility and
the centrality of beliefs to an individual's worldview. Rather than
attempting to categorize based on motive--which can be especially
difficult to discern for extremists swayed by multiple beliefs--the
CoVE framework categorizes based on expressed or exhibited beliefs that
appear to influence an individual's worldview and outlook. Many of the
cases our team analyzed did not exhibit clear motives.
The categories of the typology are thus based on whether the
individual expresses or exhibits easily discernible ideologies and the
level to which expressed beliefs appear to be central to the
individual's worldview. Inferences about how central the beliefs
expressed by an individual are to their worldview depend on factors
such as the extent to which an individual expresses the belief (e.g.,
posting about it on-line once versus repeatedly), whether the
individual is connected to groups or movements sharing the belief,
whether their chosen target aligns with the belief, and whether the
individual self-identifies as an adherent of the belief. Though
categorizing in this way still involves a significant amount of
subjectivity, we assess this process as a clear and rigorous way to
discuss cases with high levels of ambiguity that might otherwise be
left under-examined and uncategorized.
One reason that a level of subjectivity invariably remains in play
is that the concepts used to develop these categories are themselves
contested. This is most evident with the concept of ideology itself,
which is ``infamous for its superfluity of meanings.''\14\ For the CoVE
framework, ideology is defined as a set of beliefs that form a coherent
outlook. This includes clearly-defined and discernible ideologies
(e.g., neo-Nazism, jihadism) or any prejudice (e.g., racism, extreme
misogyny) that shapes a worldview or lifestyle. In some cases, a
subculture interest (e.g., extreme violence, ``Columbiners'') can
function as an ideology when it rises to the level of fixation and
clearly shapes an individual's behavioral patterns, worldview, and
identity.\15\ Functional ideologies were only evident in roughly 10
percent of cases in the dataset. This is an unconventional definition
of ideology, aimed at highlighting the reality that there are
traditional ideologies as well as functional ideologies.
---------------------------------------------------------------------------
\14\ Jonathan Leader Maynard, ``Ideology and Armed Conflict,''
Journal of Peace Research 56, Volume 56, Issue 5, April 8, 2019, page
637. (https://journals.sagepub.com/doi/10.1177/0022343319826629).
\15\ ``Columbiners'' refers to the subcultures that display an
obsession with the 1999 Columbine High School shooting in Littleton,
Colorado. Columbiners generally evince a fixation with shooters Eric
Harris and Dylan Klebold, expressing admiration for their actions and,
at times, a desire to emulate them. See: Manny Fernandez, Julie
Turkewitz, and Jess Bidgood, ``For `Columbiners,' School Shootings Have
a Deadly Allure,'' The New York Times, May 30, 2018. (https://
www.nytimes.com/2018/05/30/us/school-shootings-columbine.html).
---------------------------------------------------------------------------
The CoVE framework also relies on specific definitions for
sentiments, prejudices, grievances, subcultures, and fixations, as
these terms are central to the defining requirements of the 4
categories of composite violent extremism. A sentiment refers to an
expressed prejudice, grievance, or subculture interest. A prejudice
refers to a distinct negative opinion that is expressed toward some
defined outgroup. A grievance refers to a real or imagined wrong or
other cause for complaint or protest, especially unfair treatment. A
subculture refers to a community, often on-line, centered around a
particular aesthetic or shared interest at odds with accepted norms
(e.g., glorification of mass violence, Satanism). And a fixation refers
to an obsessive interest in or feeling about someone or something.
With these underlying definitions provided, I will now elaborate on
the different types of composite violent extremism. As I noted earlier,
the 3 categories are ambiguous, mixed, and fused. The ambiguous
category applies to cases where the perpetrator does not exhibit an
easily discernible ideology based on existing buckets (e.g., anti-
government extremism), but rather an amalgamation of sentiments. The
mixed category applies to persons who appear to hold multiple easily
discernible ideologies, potentially alongside other sentiments. The
fused category applies to persons who appear to hold a core ideology
but also exhibit other sentiments that make the case difficult to
neatly categorize using existing buckets.
Ambiguous
This category applies to violent extremists whose worldview does
not appear to be influenced by any clearly discernible ideologies, but
rather by an amalgamation of prejudices, grievances, and subcultures
that may undergird various extremist ideologies (e.g., misogyny,
racism, antisemitism, conspiracy theories, or mass violence). This
includes individuals who express some level of support for an ideology
(e.g., posting Nazi symbols) alongside other prejudices that make it
difficult or impossible to discern a central belief system.
One example is an August 2022 attack in Bend, Oregon. Ethan
Miller--who opened fire in a grocery store, killing 2 people before
taking his own life--exhibited a range of racist and misogynistic
prejudices alongside other extreme sentiments. His journal and social
media exhibited racist terms (against white people, black people, Jews,
Asians, and Latinos) and expressions of hatred for ``EVERYONE &
EVERYTHING.'' Though Miller rejected being labeled an incel
(involuntary celibate), his writings exhibited hatred and threats
toward women and an unfulfilled desire for a partner common among incel
communities. Miller also railed against the government, police,
religion, and technology. He claimed inspiration from the 1999
Columbine school shooting.\16\ The dizzying array of sentiments
expressed by Miller, which resembled multiple ideologies but never
amounted to coherent expressions of these ideologies, places him in the
ambiguous category.
---------------------------------------------------------------------------
\16\ ``Oregon Safeway Shooter Claimed Inspiration from Columbine,
Expressed Harsh Resentment Toward Women,'' SITE Intelligence Group,
August 29, 2022; Jessica McBride, ``Ethan Miller, Bend Safeway Shooter:
5 Fast Facts You Need to Know,'' Heavy, August 29, 2022. (https://
heavy.com/news/ethan-miller-bend-safeway-shooter).
---------------------------------------------------------------------------
Another example of composite violent extremism that fits within the
ambiguous category is Nikolas Cruz. On February 14, 2018, Cruz carried
out a shooting at Marjory Stoneman Douglas High in Parkland, Florida,
killing 14 students and 3 staff.\17\ Cruz's worldview appears to have
been influenced by a blend of prejudices, including racism and
antisemitism (he posted repeatedly about his hatred for black people,
Jews, Latinos, and Asians), Nazism and white supremacy (he had
swastikas carved into his gun magazines and content on his phone
referencing the KKK), Satanism (his backpack and photos found on his
cell phone depicted the Satanist reference ``666''), and a general
interest in mass violence.\18\ Cruz read extensively about acts of mass
violence, including the 2016 jihadist shooting at the Pulse nightclub
in Orlando, the 2014 incel killings in Isla Vista, the 2012 Aurora
movie theater shooting, and the 1999 Columbine attack.\19\ The sum of
Cruz's expressed sentiments does not point to a coherent ideology.
Instead, his worldview appears to center around an amalgamation of
prejudices and a general interest in violence.
---------------------------------------------------------------------------
\17\ Paul Murphy, ``Exclusive: Group Chat Messages Show School
Shooter Obsessed with Race, Violence and Guns,'' CNN, February 18,
2018. (https://www.cnn.com/2018/02/16/us/exclusive-school-shooter-
instagram-group).
\18\ Ibid.; Law & Crime Network, Twitter, July 6, 2022. (https://
twitter.com/LawCrimeNetwork/status/1544716552641122308); Marjory
Stoneman Douglas Public Safety Commission, ``Cruz Cell Phone Content
and Internet Searches,'' School Shooters.info, November 8, 2018.
(https://schoolshooters.info/sites/default/files/Cruz-Cell-Phone-
Content-and-Internet-Searches.pdf).
\19\ Broward Sheriff's Office, ``Case Supplemental Report,'' School
Shooters.info, August 9, 2018. (https://schoolshooters.info/sites/
default/files/Broward_Sheriffs_Office_Documents.pdf); Marjory Stoneman
Douglas Public Safety Commission, ``Cruz Cell Phone Content and
Internet Searches,'' School Shooters.info, November 8, 2018. (https://
schoolshooters.info/sites/default/files/Cruz-Cell-Phone-Content-and-
Internet-Searches.pdf); Terry Spencer, ``Parkland School Shooter's
Swastika Carvings Are Focus of U.S. Court Fight,'' The Times of Israel
(Israel), July 7, 2022. (https://www.timesofisrael.com/parkland-school-
shooters-swastika-carvings-are-focus-of-court-fight); Gregory Richter
and Ariana Richter, ``The Incel Killer and the Threat to the Campus
Community,'' Security Magazine, March 12, 2019. (https://
www.securitymagazine.com/articles/89962-the-incel-killer-and-the-
threat-to-the-campus-community).
---------------------------------------------------------------------------
Violent extremists who fit the ambiguous category can be identified
by their expression (through their social media profiles, writings, and
other sources) of an amalgamation of disparate prejudices, ideas, or
grievances without a clearly discernible ideology. They often exhibit
elements of an ideology but only inconsistently and interspersed with a
variety of other beliefs or grievances, and their belief systems lack
structure and consistency. Cruz, for example, lacked ideological
consistency, piecing together disparate ideas--a determination
supported by a court psychologist's later findings.
Mixed
This category applies to violent extremists whose worldview appears
to be influenced by multiple distinct and discernible ideologies
alongside other prejudices, grievances, or subcultures. This includes
individuals who adhere to multiple discernible ideologies or a
combination of traditional and functional ideologies. Three individuals
across 3 countries represent examples of the mixed category of
composite violent extremism.
In April 2021, French authorities arrested 18-year-old Leila B. for
plotting a terrorist attack targeting a church in Montpellier on Easter
weekend.\20\ During a search of her residence, authorities found
material for constructing explosive devices and a journal filled with
sketches of symbols associated with jihadism (e.g., a depiction of an
ISIS member holding a decapitated head) and neo-Nazism (e.g., swastikas
and depictions of Nazi soldiers), alongside other evidence that she was
connected on-line with both ideological movements.\21\ She also
exhibited a fixation on mass violence: She was obsessed with gore,
Columbine, and serial killers.\22\
---------------------------------------------------------------------------
\20\ ``Suspect Arrested Over Easter Linked to French Church Plot,''
Associated Press, April 8, 2021. (https://apnews.com/article/arrests-
terrorism-europe-france-02d562667268b2f0158df8714- fe423eb); Jeremie
Pham-Le, `` `Je Voulais Mettre Cette Bombe dans l'Eglise': Revelations
sur L'Adolescente de Beziers qui Projetait un Attentat [`I wanted to
put this bomb in the church': revelations about the teenager from
Beziers who planned an attack],'' Le Parisien (France), April 17, 2021.
(https://www.leparisien.fr/faits-divers/je-voulais-mettre-cette-bombe-
dans-leglise-revelations-sur-ladolescente-de-beziers-qui-projetait-un-
attentat-17-04-2021-TH7TMLMZB5DXX- BQHSK7V7PTD54.php).
\21\ ``Enquete pour `Association de Malfaiteurs Terroriste' a
Beziers : La Suspecte Mise en Examen et Ecrouee [Investigation for
`terrorist criminal association' in Beziers: the suspect indicted and
imprisoned],'' Le Monde (France), April 8, 2021. (https://
www.lemonde.fr/societe/article/2021/04/08/enquete-pour-association-de-
malfaiteurs-terroriste-a-beziers-des-elements-de-radicalisation-
retrouves-chez-l-une-des-suspectes_6076012_3224.html); Pham-Le, `` `Je
Voulais Mettre Cette Bombe dans l'Eglise,' ''; Ronan Folgoas and
Jeremie Pham-Le, `` `Je Veux Faire Pire Que Columbine': le Projetee
Tuerie de Masse d'un Adorateur d'Hitler Dejoue par la DGSI [`I want to
do worse than Columbine': the mass murder plan of a Hitler worshiper
foiled by the DGSI],'' Le Parisien (France), October 2, 2021. (https://
www.leparisien.fr/faits-divers/je-veux-faire-pire-que-columbine-le-
projet-de-tuerie-de-masse-dun-normand-adorateur-dhitler-dejoue-par-la-
dgsi-02-10-2021-KPUTTZCGGV- B5HKN4PYGAUWFT5M.php).
\22\ Pham-Le, `` `Je Voulais Mettre Cette Bombe dans l'Eglise.' ''
Leila B.'s outlook appears to be based on a mix of ideologies blended
with a general fixation on mass violence.
---------------------------------------------------------------------------
Andrea Cavalleri, arrested by Italian authorities in January 2021,
also exhibited an interest in violent subcultures, writing about the
``pleasure'' he would feel carrying out a school shooting.\23\ He was
arrested for allegedly establishing and serving as the leader of a neo-
Nazi organization ``with the aim of recruiting volunteers and planning
extreme and violent acts for subversive purposes . . . inspired by the
American supremacist group Atomwaffen Division and the Nazi Waffen-
SS.''\24\ Beyond his evident adherence to neo-Nazi beliefs, Cavalleri
was also a self-declared incel and stated his desire to carry out an
attack motivated by this identity. At one point he wrote to a friend:
``We will be the first Italian incels to take action.''\25\ Cavalleri's
self-identification as an incel and his neo-Nazi ideology, along with
attributing a desire to commit violence based on both, places him in
the mixed category.
---------------------------------------------------------------------------
\23\ Paolo Frosina, ``Il Delirio del Suprematista di Savona:
`Hitler Come Cristo Guida Luce'. La Missione di una Guerra la Razziale
`Contro Negri E Degenerati' [The delirium of the supremacist from
Savona: `Hitler as Christ guides light'. The mission of a racial war
`against negroes and degenerates'],'' Il Fatto Quotidiano (Italy),
January 23, 2021. (https://www.ilfattoquotidiano.it/2021/01/23/il-
delirio-del-suprematista-di-savona-hitler-come-cristo-guida-luce-la-
missione-di-una-guerra-la-razziale-contro-negri-e-degenerati/6075257).
\24\ Ministry of the Interior of Italy, Polizia di Stato, Press
Release, ``Terrorismo: Arrestato 22enne Suprematista e Negazionista
[Terrorism: 22-year-old arrested for supremacist propaganda],'' January
22, 2021. (https://www.italpress.com/terrorismo-arrestato-un-22enne-
per-propaganda-suprematista).
\25\ Francesco Marone, ``Black Sun: A Case of Radicalization
Between Neo-Nazism and Incel Ideology,'' Italian Institute for
International Political Studies (Italy), January 27, 2021. (https://
www.ispion-line.it/en/pubblicazione/black-sun-case-radicalization-
between-neo-nazism-and-incel-ideology-29063).
---------------------------------------------------------------------------
Finally, in October 2014, Zale Thompson attacked a group of NYPD
officers with a hatchet, wounding 2 of them before being killed by
police. The NYPD described the sentiments that Thompson expressed on
social media as ``anti-Western, anti-government, and in some cases
anti-White,'' and the FBI said he sought ``inspiration from foreign
terrorist sources like ISIS, but there is also evidence he was focused
on black separatist ideology.''\26\ Our analysis indicates that his
outlook was influenced by two distinct and easily discernible
ideologies: jihadism (he viewed extensive ISIS and al-Qaeda propaganda
and posted on social media about jihadism) and black separatism (he had
loose connections to black nationalist groups and advocated for black
revolt).\27\ Drawing heavily from these two ideological frameworks
places him in the mixed category.
---------------------------------------------------------------------------
\26\ ``Transcript: Mayor de Blasio and Commissioner Bratton Provide
an Update on the Assault of Two Police Officers,'' NYC.gov, October 24,
2014. (https://www1.nyc.gov/office-of-the-mayor/news/897-14/transcript-
mayor-de-blasio-commissioner-bratton-provide-update-the-assault-two-
police); Jonathan Dienst, ``Hatchet Attack on NYPD Officers Was `Act of
Terror': FBI Director,'' NBC New York, November 17, 2014. (https://
www.nbcnewyork.com/news/local/hatchet-attack-nypd-officers-act-of-
terror-fbi-director-james-comey/844450).
\27\ Alexanda Klausner, Kieran Corcoran, David Martosko, and Sophie
Jane Evans, ``Armed and Radicalized: Ranting `Self-Proclaimed Convert'
New York Hatchet Attacker was a `Terrorist', Say Police,'' Daily Mail
(UK), October 24, 2014. (https://www.dailymail.co.uk/news/article-
2806731/Was-terror-attack-Police-probe-extremist-links-on-line-rants-
New-York-hatchet-attacker-emerge-shot-dead-attack-group-cops-
Queens.html); ``Terror Connection Not Ruled Out in Hatchet Attack,
Police Say,'' Fox News, November 21, 2015. (https://www.foxnews.com/us/
terror-connection-not-ruled-out-in-hatchet-attack-police-say); ``Black
Identity Extremists Likely Motivated to Target Law Enforcement
Officers,'' FBI, August 3, 2017. (https://privacysos.org/wp-content/
uploads/2017/10/FBI-BlackIdentityExtremists.pdf); Michael Schwirtz and
William K. Rashbaum, ``Attacker With Hatchet is Said to Have Grown
Radical on His Own,'' The New York Times, October 24, 2014. (https://
www.nytimes.com/2014/10/25/nyregion/man-who-attacked-police-with-
hatchet-ranted-about-us-officials-say.html); Priscilla DeGregory, Kevin
Sheehan, and Kirstan Conley, ``Black Panther Hails Ax Attack on Cops,''
New York Post, October 27, 2014. (https://nypost.com/2014/10/27/new-
black-panther-group-hails-ax-attacker).
---------------------------------------------------------------------------
Violent extremists in this category can be identified by their
expressions of multiple distinct and easily discernible ideologies
through their behaviors (e.g., school behavioral records or arrests),
writings (e.g., social media posts, manifestos), or personal belongings
(e.g., books, flags). Leila B.'s inclusion in mixed was primarily due
to her behaviors and active communication with members of the
Atomwaffen group as well as ISIS. Another marker that an individual
fits this category includes an expressed desire to attack based on
different beliefs, such as Cavalleri stating he would like to be an
incel attacker while also planning violence with neo-Nazis. Unlike
Cavalleri, Thompson did not explicitly self-identify with both
ideologies that we attributed to him, but his social media profile and
writings exhibited adherence to multiple distinct ideologies.
Fused
This category applies to violent extremists whose worldview is
based on one clearly discernible ideology, but who appear to fuse this
core ideology with other distinct prejudices or grievances. In cases
that fit this category, there are clear indications of a core ideology
but the presence of other sentiments complicates what might otherwise
be clean bucketing (e.g., the individual's on-line footprint points to
a single ideology but also contains references to other distinct
sentiments).
In 2019, authorities in the United Kingdom arrested Jack Reed for
planning a terrorist attack.\28\ He reportedly wrote a manifesto with a
list of targets to attack, including schools, pubs, council buildings,
post offices, and a synagogue.\29\ Reed fundamentally embraced neo-Nazi
ideology. His journal contained Nazi symbols and admiration for Hitler
and he initially came to authorities' attention when he expressed
support for the British neo-Nazi group National Action.\30\ However,
Reed's core neo-Nazi outlook appeared to be infused with Satanism (he
described his Satanic beliefs on an on-line forum, calling himself an
``immoral individual,'' and also had references in his journal to the
esoteric Satanist group Order of Nine Angles) and mass violence (he
reportedly expressed admiration for murderers Ian Brady and Charles
Manson, and repeatedly visited websites related to the Columbine
attack).\31\ Though it is clear that Reed fits the neo-Nazi label, his
interests in Satanism and mass violence complicate the picture--
possibly explaining why some of the targets on his list did not appear
to have connections to his neo-Nazi beliefs.
---------------------------------------------------------------------------
\28\ Lizzie Dearden, ``One of UK's Youngest Terror Plotters Named
After Losing Anonymity Battle,'' The Independent (UK), January 11,
2021. (https://www.the-independent.com/news/uk/crime/neo-nazi-terror-
plot-durham-jack-reed-b1785650.html)
\29\ Faye Brown, ``Terrifying Drawings Found in Bedroom of Neo-
Nazi, 16, Convicted of Terrorism,'' Metro (UK), November 20, 2019.
(https://metro.co.uk/2019/11/20/terrifying-drawings-found-bedroom-neo-
nazi-16-convicted-terrorism-11189596).
\30\ Ibid.
\31\ Daniel De Simone, ``Durham Teen Neo-Nazi Became `Living Dead,'
'' BBC (UK), November 22, 2019, (https://www.bbc.com/news/uk-england-
tyne-50397477); Brown, ``Terrifying Drawings Found in Bedroom of Neo-
Nazi.''
---------------------------------------------------------------------------
Scott Beierle carried out a November 2018 shooting at a yoga studio
in Tallahassee, Florida, killing 2 women and injuring 5 before killing
himself. The U.S. Secret Service's National Threat Assessment Center
used Beierle as a case study on misogynistic extremism, citing notes he
left before the attack, his history of sexual harassment, and the
content of the music he produced as evidence that incel and
misogynistic beliefs fueled his worldview.\32\ In other words, extreme
misogyny was Beierle's core ideology, as it was fundamental to his
identity and worldview. However, Beierle also exhibited racism and
white supremacist prejudices. The National Threat Assessment Center
reported that he ``openly admired Hitler and Aryan Nations'' and that
``other members of on-line social networks referred to him as a Nazi.''
A few of Beierle's song lyrics and descriptions also reveal racist and
white supremacist sentiments (one titled ``To Arms!'' calls for people
to take up arms to defend the homeland from immigrants).\33\ While
Beierle exhibited extreme misogyny as a core ideology, categorizing him
solely as such would be inaccurate.
---------------------------------------------------------------------------
\32\ National Threat Assessment Center, ``Hot Yoga Tallahassee: A
Case Study of Misogynistic Extremism'' (Washington: U.S. Secret
Service, March 2021). (https://www.secretservice.gov/sites/default/
files/reports/2022-03/NTAC%20Case%20Study%20-
%20Hot%20Yoga%20Tallahassee_0.pdf).
\33\ See, for example, Scott Beierle, ``American Whore,'' archived
November 4, 2018. (https://web.archive.org/web/20181104002705mp_/
https://pathofdefiance.com/american-whore); Scott Beierle, ``I'm
Dreaming . . . ,'' archived November 4, 2018. (https://web.archive.org/
web/20181104002542mp_/https://pathofdefiance.com/4-i-m-dreaming); Scott
Beierle, ``American Burden,'' archived November 4, 2018. (https://
web.archive.org/web/20181104001746mp__/https://pathofdefiance.com/8-
american-burden); Scott Beierle, ``To Arms!'' archived November 4,
2018. (https://web.archive.org/web/20181104000923/https://
pathofdefiance.com/to-arms).
---------------------------------------------------------------------------
Deciding whether a case of violent extremism fits within the fused
category depends on whether the extremists demonstrate a primary
fixation on one ideological framework or belief that dictates their
behaviors, writings, or personal belongings while still exhibiting some
level of adherence to other sentiments. The fused category, as opposed
to ambiguous or mixed, is for individuals whose worldview is primarily
centered around one idea. For Beierle, his history of sexual
harassment, attack target selection, and lyrics evidenced that extreme
misogyny was central to his worldview even though he also expressed
racist sentiments. In the case of Reed, his writings and admiration for
Nazism were clearly central but he nonetheless also exhibited some
behavior, writing, or personal belongings that suggests an interest or
less central belief in Satanism. This category primarily exists to
allow for nuance in evaluating attackers who are more complex upon
closer examination than they appear on the surface.
toward explanations of composite violent extremism
The CoVE framework helps to both identify and disaggregate various
acts of violent extremism that have challenged traditional categories
in recent years. The apparently increased frequency of such incidents
has gained attention from scholars and practitioners, prompting debate
and the introduction of new policy frameworks in multiple countries.
Why is this occurring?
The Information Environment and On-line Space
I believe that the information environment and on-line space is
critical and relevant to the present hearing. Much of the current
literature on the phenomenon we are referring to as CoVE highlights the
importance of the information environment. As Jakob Guhl, Moustafa
Ayad, and Julia Ebner note, multiple ideological trends have been
``converging into ideologically elastic on-line subcultures.''\34\
Cynthia Miller-Idriss and Brian Hughes argue that ``material
infrastructure enables the muddling of ideological rationales. . . .
The infrastructure of digital communication technology, at both
engineering and design levels, makes motley ideological blends
increasingly common.''\35\ There is an intuitive logic to this, as
today's information environment--which broadly refers to the full
spectrum of actors and systems that produce, share, and use
information--is widely understood to play some role in reshaping
people's beliefs and behaviors. The information environment is
particularly impacted by on-line spaces (e.g., the internet and social
media), which are quickly becoming the primary means by which people
communicate and consume information. A 2021 Pew Research survey
revealed that 86 percent of American adults get their news from digital
devices, about half of whom read their news on social media.\36\ The
CoVE cases in the dataset had extensive on-line footprints and social
media activity (on platforms like Facebook and YouTube) related to
their beliefs, but not necessarily to a greater degree than the general
population.
---------------------------------------------------------------------------
\34\ Jakob Guhl, Moustafa Ayad, and Julia Ebner, ``From The Vicious
Cycle To Ideological Convergence,'' VoxPol (Ireland), January 26, 2022.
(https://www.voxpol.eu/from-the-vicious-cycle-to-ideological-
convergence).
\35\ Cynthia Miller-Idriss and Brian Hughes, ``Blurry Ideologies
and Strange Coalitions: The Evolving Landscape of Domestic Extremism,''
Lawfare, December 19, 2021. (https://www.lawfareblog.com/blurry-
ideologies-and-strange-coalitions-evolving-landscape-domestic-
extremism).
\36\ Elisa Shearer, ``More than Eight-In-Ten Americans Get News
from Digital Devices,'' Pew Research, January 12, 2021. (https://
www.pewresearch.org/short-reads/2021/01/12/more-than-eight-in-ten-
americans-get-news-from-digital-devices).
---------------------------------------------------------------------------
Social psychology research indicates that social media enables
individuals to strategically ``connect with like-minded others and
distance themselves from people with conflicting belief sets.''\37\
This would suggest that social media tends to be a place where people
confine themselves to narrow viewpoints, rather than being exposed to a
range of beliefs. But it is also true that people consume information
in the current information environment like drinking from a firehose.
The proliferation of news sites, social media platforms, and other on-
line information channels readily available to internet users makes it
easier than ever to passively consume and be shaped by vast amounts of
information from a multitude of on-line actors and communities
simultaneously. Though someone might join a certain on-line channel
intentionally--based on their preexisting interests and proclivities--
that person might just as easily stumble upon a forum or thread that
piques a new interest. In some cases, it is possible to observe in
retrospect how the on-line space creates an environment where
worldviews are formed in both intentional and haphazard ways that
produce idiosyncratic beliefs.
---------------------------------------------------------------------------
\37\ Adrian Luders, Alejandro Dinkelberg, and Michael Quayle,
``Becoming `Us' in Digital Spaces: How Online Users Creatively and
Strategically Exploit Social Media Affordances to Build up Social
Identity,'' Acta Psychologica, Volume 228, August 2022. (https://
www.sciencedirect.com/science/article/pii/
S0001691822001585?via%3Dihub).
---------------------------------------------------------------------------
Lindsay Souvannarath--who plotted with a co-conspirator in 2015 to
carry out a shooting at a mall in Halifax, Canada--described her
radicalization process in a podcast interview, noting that it happened
``by chance'' through on-line communities. By her own account,
Souvannarath became a National Socialist after she joined an on-line
art community and connected with an artist who happened to be a
National Socialist. Through this relationship, she gained exposure to
the broader neo-Nazi community and came to accept this belief
system.\38\ Similarly, she initially became obsessed with Columbine
because of on-line research she conducted for a short story she was
writing. She wanted the story to include a shooting, and her research
exposed her to Columbine subcultures, where she built friendships and
gained exposure to the writings of Eric Harris and Dylan Klebold.\39\
---------------------------------------------------------------------------
\38\ Jordan Bonaparte, Interview with Lindsay Souvannarath, ``The
Story of Lindsay Souvannarath_Life Before Choosing Death,'' Night Time,
February 7, 2019. (https://www.nighttimepodcast.com/episodes/lindsay-
souvannarath-2).
\39\ Ibid.
---------------------------------------------------------------------------
These on-line influences culminated in an attack plan that
exhibited elements of both Nazi ideology and a fixation with Columbine.
Souvannarath and her co-conspirator, whom she met on-line, targeted a
mall in ``a protest against capitalism, against consumerism, against
greed'' in the vein of their National Socialist beliefs, and planned to
end the attack ``just like Columbine'' by shooting themselves on the
count of 3.\40\ Fortunately, the attack was thwarted by the Canadian
Border Services Agency when Souvannarath attempted to cross from the
United States into Canada to carry out the plot.\41\
---------------------------------------------------------------------------
\40\ Jordan Bonaparte, Interview with Lindsay Souvannarath, ``The
Story of Lindsay Souvannarath_3_Lindsay, James, and the Valentine's Day
Massacre,'' Night Time, February 15, 2019. (https://
www.nighttimepodcast.com/episodes/lindsay-souvannarath-3).
\41\ The Canadian Border Services Agency detained Souvannarath
based on an anonymous tip and lack of return ticket and luggage. See:
Mack Lamoureux, ``The Woman Who Plotted a Valentine's Mass Murder
Shares How the Internet Radicalized Her,'' Vice, February 21, 2019.
(https://www.vice.com/en/article/eve54j/the-woman-who-plotted-a-
valentines-mass-murder-shares-how-the-internet-radicalized-her).
---------------------------------------------------------------------------
Lindsay Souvannarath's case highlights how individuals can be drawn
into certain beliefs based on the networks and subcultures they choose
to engage with on-line, but people can also be shaped by the
information they consume exogenous to their own actions. The
information environment--social media and news media in particular--is
rife with information intentionally produced and disseminated to subtly
influence people's beliefs and behaviors without their knowledge.
Foreign adversaries--both state and non-state--are known to manipulate
the information environment to exacerbate and exploit political,
ideological, and other divides in American society. Adversaries benefit
from advancing any narrative that challenges the status quo while
driving further polarization. This type of ``hostile social
manipulation'' or ``virtual societal warfare'' may generally lend
itself to individuals being influenced by an amalgamation of inputs and
narratives designed to stoke chaos and a sense of urgency to act.\42\
---------------------------------------------------------------------------
\42\ Michael Mazarr, Abigail Casey, Alyssa Demus, Scott W. Harold,
Luke J. Matthews, Nathan Beauchamp-Mustafaga, and James Sladden,
Hostile Social Manipulation: Present Realities and Emerging Trends
(Santa Monica: RAND Corporation, 2019). (https://www.rand.org/pubs/
research_reports/RR2713.html); Michael J. Mazarr, Ryan Bauer, Abigail
Casey, Sarah Heintz, and Luke J. Matthews, The Emerging Risk of Virtual
Societal Warfare: Social Manipulation in a Changing Information
Environment (Santa Monica: RAND Corporation, 2019). (https://
www.rand.org/pubs/research_reports/RR2714.html).
---------------------------------------------------------------------------
Violence & Nihilism
Individuals primarily oriented toward violence could be attaching
themselves to a range of beliefs that provide ideological frameworks or
justifications for violence. Indeed, many of the cases we observed
exhibited distinct interests in mass violence. Many extremists in the
dataset extensively glorified mass violence and mass attackers, most
commonly Eric Harris and Dylan Klebold (perpetrators of the 1999
Columbine shooting), Elliot Rodger (perpetrator of 2014 Isla Vista
shootings), Timothy McVeigh (perpetrator of the 1995 Oklahoma City
bombing), and Dylann Roof (perpetrator of the 2015 Charleston church
shooting). At times, the relationship between an interest in violence
and an ideology was symbiotic. Leila B., for example, told
investigators that she adhered to jihadism and neo-Nazism to
``justify'' her ``fascination with violent death.''\43\ Still, this
explanation raises further questions about whether and why individuals
primarily oriented toward violence are a new or rising phenomenon, and
whether this is novel to CoVE cases.
---------------------------------------------------------------------------
\43\ Jeremie Pham-Le, `` `Je Voulais Mettre Cette Bombe dans
l'Eglise' : Revelations sur l'Adolescente de Beziers qui Projetait un
Attentat [`I wanted to put this bomb in the church': revelations about
the teenager from Beziers who planned an attack],'' Le Parisien
(France), April 17, 2021, (https://www.leparisien.fr/faits-divers/je-
voulais-mettre-cette-bombe-dans-leglise-revelations-sur-ladolescente-
de-beziers-qui-projetait-un-attentat-17-04-2021-TH7TMLMZB5DXX-
BQHSK7V7PTD54.php).
---------------------------------------------------------------------------
It is also possible that within the broader violent extremism
ecosystem, violence and nihilism are becoming more central than
ideology. Today's violent extremists may be coalescing more around
opposition to the current system by adopting any violent anti-status
quo belief, and less around specific desired ideological outcomes. And,
with the proliferation of existential threats and cynicism about the
ability of any extant political system to deal with them, the reasons
for the growth in nihilism are clear enough. From this vantage point,
destroying the current system is of foremost importance, while
determining what will replace it may be secondary or even irrelevant.
Driven by a sense of urgency for change, violent extremists may be
drawn to a range of belief systems that present perceived possibilities
of success.
This explanation requires more study, but one case that illustrates
the point is the Order of Nine Angles (O9A) and Ethan Melzer. O9A has a
complex and often deliberately obscurantist belief system, but it can
be understood as advocating for the destruction of Western society by
any means necessary, encouraging adherents to bolster or even
collaborate with movements like jihadism and neo-Nazism.\44\ Melzer, a
self-proclaimed O9A adherent, was involved in neo-Nazi channels on-line
while plotting what he believed would be a jihadist attack against a
U.S. military convoy in Turkey. He believed that this attack would draw
the United States into another prolonged conflict and thus contribute
to the collapse of the current U.S. political and social system.
Further, a coalescence around nihilism and anti-status quo aesthetics
could also explain why some factions of the white supremacist movement
have adopted Satanism, which on its surface has little overlap with
white supremacism.
---------------------------------------------------------------------------
\44\ Daveed Gartenstein-Ross and Emelie Chace-Donahue, ``The Order
of Nine Angles: Cosmology, Practice & Movement'' (forthcoming in
Studies in Conflict & Terrorism).
---------------------------------------------------------------------------
Decentralization & Ideological Fragmentation
The on-line space is also contributing to the formation of
decentralized extremist movements and networks, which may in turn
decrease the ideological singularity and purity that comes with
centralized command in an organized off-line group.\45\ In the context
of ever-evolving on-line movements and subcultures, it could be
increasingly difficult for groups to maintain control or influence over
ideology. Again, this is apparent in the white supremacist movement,
where groups like Atomwaffen and National Action became fractured over
the adoption of subcultures like Satanism that spread through the
movement on-line.\46\
---------------------------------------------------------------------------
\45\ Cynthia Miller-Idriss and Brian Hughes, ``Blurry Ideologies
and Strange Coalitions: The Evolving Landscape of Domestic Extremism,''
Lawfare, December 19, 2021. (https://www.lawfareblog.com/blurry-
ideologies-and-strange-coalitions-evolving-landscape-domestic-
extremism).
\46\ Kelly Weill, ``Satanism Drama is Tearing Apart the Murderous
Neo-Nazi Group Atomwaffen,'' The Daily Beast, March 21, 2018, (https://
www.thedailybeast.com/satanism-drama-is-tearing-apart-the-murderous-
neo-nazi-group-atomwaffen); ``Neo-Nazis Denounce Occultists Associated
with Terror Plot, Favor Optics-Friendly Groups,'' SITE Intelligence
Group, June 25, 2020; Lizzie Dearden, ``Student Who Founded Neo-Nazi
Terrorist Groups Convicted of Terror Offences,'' The Independent (UK),
July 11, 2021. (https://www.independent.co.uk/news/uk/crime/andrew-
dymock-srn-sonnenkrieg-nazi-b1864321.html).
---------------------------------------------------------------------------
Google's tech incubator Jigsaw, which conducts research on issues
related to violent extremism, touched on this move away from formal
groups to decentralized on-line networks in the February 2021 issue of
its magazine The Current. Based on interviews with former extremists,
the magazine discusses how on-line networks enable individuals to join
the white supremacist movement without exclusively adopting the
ideology. Jigsaw theorizes that ``the internet lowers barriers for
those curious about a supremacist idea to anonymously learn about it,
lurk in supremacist spaces on-line, and eventually interact with others
as part of loose, informal networks.'' One result is that this
``enables supremacists to pick and choose which aspects of supremacist
ideology resonate and engage selectively with those ideals . . .
supremacists no longer have to find a group with which they fit; there
is less friction to joining the distributed movement because they can
retain idiosyncratic beliefs.''\47\
---------------------------------------------------------------------------
\48\ ``Global, Connected and Decentralized,'' The Current 2 (2021).
(https://medium.com/jigsaw/global-connected-and-decentralized-
41e496b44daf).
---------------------------------------------------------------------------
Terrorism experts have also discussed the importance of this shift
and its impact on ideology and beliefs. In July 2020, Colin Clarke and
Bruce Hoffman wrote an analysis of the ``next American terrorist,''
emphasizing increasing decentralization and ideological fragmentation.
Writing that ``bureaucratic organizations with hierarchical leadership
structures and clearly-defined objectives have been supplanted by
loosely networked movements with amorphous goals that exist across the
ideological spectrum,'' they suggest that ``a confluence of ideological
affinities is more powerful in inspiring and provoking violence than
the hierarchical terrorist organizational structures of the past.''\48\
Ideological fragmentation seems like a natural extension of this
decentralization and may play some role in how individuals come to
adopt composite beliefs.
---------------------------------------------------------------------------
\48\ Bruce Hoffman, ``The Next American Terrorist,'' Cipher Brief,
July 2, 2020. (https://www.thecipherbrief.com/article/united-states/
the-next-american-terrorist).
---------------------------------------------------------------------------
implications for prevention
The rise of composite violent extremism (CoVE) presents a unique
set of challenges for terrorism prevention efforts. Traditional
counterterrorism strategies were largely designed to address violent
extremism rooted in well-defined ideological movements. Prevention
frameworks were built on the premise that extremist beliefs and
pathways to radicalization followed a structured, identifiable
trajectory. CoVE disrupts these assumptions.
Prevention efforts can be broadly categorized into 3 levels:
1. Publicly-Focused Initiatives.--Awareness campaigns, media
literacy programs, and efforts to promote critical thinking and
community resilience.
2. Community-Based Approaches.--Collaboration with local leaders
and organizations to monitor risk factors and conduct early
interventions.
3. Targeted Interventions.--Focused efforts on individuals at risk
of carrying out violence, including deradicalization,
mentorship programs, and law enforcement action.
While these categories remain useful, CoVE presents distinct
challenges that require adaptation:
1. Distinguishing Genuine Threats.--A core difficulty in countering
CoVE is separating real threats from noise in an information
space saturated with irony, trolling, and shock value content.
Short-form media--including tweets, memes, TikToks, and YouTube
shorts--has reshaped how extremists express their beliefs.
While traditional extremists often articulate a coherent
ideological framework, composite extremists may express
disjointed and contradictory beliefs that nonetheless reinforce
their intent to commit violence.
Practitioners must recognize that, despite their contradictions,
composite violent extremists do have belief systems. Their
worldviews, while fragmented, still have patterns, boundaries,
and contours that can be analyzed. A critical challenge is
determining when an individual's extreme expressions, even if
incoherent, signal an intent to act. The failure to address
this ambiguity can result in real threats being dismissed as
unserious or performative.
2. Avoiding Counterproductive Engagement.--Prevention measures
could backfire if they trigger unexpected grievances hidden in
the complex worldviews of people at risk of succumbing to
composite violent extremists. In targeted interventions, this
uncertainty makes selecting a credible intervener--someone seen
as trustworthy, competent, or impartial--more complex than it
is for traditional ideological terrorists.
In an age of polarization, institutional efforts to inoculate
against extremism (e.g., messaging campaigns) can themselves
provoke grievances. Similar issues crop up in direct
communication with composite violent extremists, such as one-
on-one interventions. Misrepresenting a composite worldview
with an imprecise blanket term like ``white supremacism'' could
potentially undermine the trust necessary for effective
deradicalization. After all, one aspect of building trust in
the deradicalization context is providing individuals with the
confidence that important aspects of their former extremist
worldview are understood by the intervenors with whom they
interact. The potential for misunderstandings in this regard is
even greater in the context of CoVE.
3. Tracking the Evolution of Fringe Beliefs.--Many ideologies that
today seem well-defined began as loose collections of
grievances and subcultures. The incel (involuntary celibate)
movement, for example, evolved from disparate on-line spaces
into a structured belief system with its own terminology,
grievances, and moral framework. Understanding when composite
worldviews harden into new ideological movements is important
for anticipating emerging threats.
Practitioners should track when composite violent extremists begin
coalescing around shared narratives and self-reinforcing
communities. While some CoVE cases may remain fragmented and
individualized, others may lay the foundation for the next
generation of extremist movements. Recognizing these patterns
early is thus vital.
fine-tuning prevention strategies
The challenges presented by CoVE do not necessitate a wholesale
reinvention of terrorism prevention. The existing toolkit remains
valuable, but fine-tuning is necessary to account for the ideological
fluidity, digital dynamics, and radicalization pathways unique to CoVE.
Prevention must remain adaptive, nuanced, and aware of the changing
nature of extremism. Addressing composite violent extremism requires
refining intervention strategies, ensuring that risk assessments
account for ideological ambiguity, and enhancing efforts to track how
radical beliefs take shape in the digital era. By doing so, we can
build a more resilient system capable of identifying threats earlier,
engaging at-risk individuals effectively, and preventing future
attacks.
Mr. Pfluger. Thank you very much.
The Chair now recognizes Dr. Zelin for his opening
statement.
STATEMENT OF AARON ZELIN, PH.D., SENIOR RESEARCH FELLOW, THE
WASHINGTON INSTITUTE FOR NEAR EAST POLICY
Mr. Zelin. Thank you, Chairman Pfluger, it's a great honor
to do a hearing with you once again, Ranking Member Magaziner,
and the distinguished Members of the subcommittee for giving me
this opportunity to testify today on how terrorists use the
internet today.
This is an important topic in light of the recent New
Orleans attack. On that day, Shamsud-Din Jabbar was killed in a
shootout with police after driving his Ford pickup truck with
an Islamic State flag through a New Year's crowd on Bourbon
Street, killing 14 and injuring 57. Prior to that attack, last
October, he recorded a video of the French Quarter using his
Meta smart glasses. Jabbar also expressed support for IS in
videos posted to Facebook, researched the December 2016 car
attack at the Christmas market in Germany, and pledged
allegiance to the group before his deadly attack.
The following week, the Islamic State wrote an editorial in
its weekly newsletter titled ``We Were There,'' where it
gloated over Jabbar's attack. It highlighted the group's
influence and incitement capabilities and the piece also
boasted that the perpetrator used American technology,
referring to the Meta smart glasses. The piece concluded by
once again urging Muslims in both Europe and the United States
to carry out more terrorist attacks, highlighting how IS uses
one attack to push yet another.
Jabbar's attack also falls in line with the Islamic State's
instructional attack planning. In particular, in mid-November
2016, in IS English language magazine at the time called
Rumiyah, it released an article titled ``Just Terror Tactics,''
that discussed the use of pickup trucks, renting a vehicle if
you did not have one, targeting pedestrian streets, and using a
secondary weapon, among other things. In this case, Jabbar
placed 2 coolers with explosive devices at 2 other locations in
the city's French Quarter, and thankfully they didn't explode,
also.
Finally, the article tells prospective attackers to find an
appropriate way for announcing one's allegiance to the
caliphate, and in this way we saw this with the Facebook post,
as well as him using the ISIS flag on the car during the
attack. This all shows that while this attack occurred in 2025,
in many ways examples and guidance literature from almost a
decade ago now that remains accessible on-line has a long shelf
life, and not enough is still being done to make sure that
potential attackers do not have access to such content on-line.
Beyond traditional propaganda efforts to recruit and
incite, there are more recent innovations related to crypto,
live streaming and AI. Regarding crypto, based on data
collected for the Washington Institute's Islamic State
Worldwide Activity Map, since 2015, there have been at least 36
arrest cases globally related to jihadi use of
cryptocurrencies, with 13 of them happening in 2024 alone,
illustrating a huge uptick in the past year. It's likely that
there are other cases as well that we just don't know about
because it didn't come within the judicial system.
This is not theoretical either. Last year in April, the FBI
arrested an 18-year-old plotting an attack on churches in
Idaho, and as part of his plot, he attempted to donate $11,000
to the Islamic State using the Monero cryptocurrency, which
they promote in one of their English language magazines
nowadays.
Beyond traditional social media platforms, there's also
been a rise in the use of TikTok by IS supporters. Based on
data from the Islamic State Worldwide Activity Map, there have
been 15 arrest cases since 2023 that have shown that the
arrestees have been involved in one way or another with IS
propaganda on TikTok. Just last week, for example, in
Minneapolis, a man was arrested by the FBI and charged for
attempting to provide material support to IS in Somalia, but as
part of the investigation he also actually praised the
perpetrator of the IS-inspired attack in New Orleans on TikTok
itself.
As for live streaming, it could have been much more of a
psychological effect nationally if the New Orleans attacker had
live streamed his attack on his Meta glasses through Facebook
instead of only using it for reconnaissance purposes. We've
seen past attacks in Europe from IS supporters using live
streaming either during or after the attack, so it's not
something theoretical.
Last, there are worries that generative AI could be
exploited by terrorists. Thus far, at least amongst jihadis,
beyond propaganda, there has not been much evidence that it's
used in terms of--for deadly effects, but as Daveed mentions,
it's definitely a possibility in the future. There have been
incoherent conversations by IS supporters on-line only in last
2 weeks or so about the Chinese application DeepSeek and how
they could potentially exploit it. It's, of course, too soon to
see how that might evolve, but it's definitely troubling.
Beyond the specifics of how terrorists might exploit
technology, policies related to technology and platforms also
have a role in providing space, making it very difficult to
use. In recent years, due to controversies related to
censorship within the West politically, there have been a
backlash related to moderating content on-line, even if it's
extremist in nature. This should not be overblown, however,
since jihadi use is still not as wide-spread on mainstream
platforms as it was a decade ago, but in a relative sense,
there is greater space for jihadis to exploit these platforms
than in the past 2 or 3 years.
I know time is short, so I'll just briefly mention that the
U.S. Government should urge technology companies and social
media platforms to redouble their efforts at content moderation
related to the jihadi movement and go beyond just Arabic and
English to nascent spaces of greater exploitation in recent
years, with languages in Africa and Central Asia in particular.
Although there have been recent calls to cut funding and jobs
across the U.S. Government, cutting ones related to tracking
on-line jihadi recruitment and attack plotting could undermine
future security and lead to greater risks at home and abroad.
No one, of course, wants to see yet another successful attack
like the one we recently saw in New Orleans. Thank you.
[The prepared statement of Mr. Zelin follows:]
Prepared Statement of Aaron Zelin
March 4, 2025
Thank you Mister Chairman and Members of the committee for giving
me the opportunity to testify today on how terrorists use the internet
and on-line networks for recruitment and radicalization. This is an
important topic in light of the recent New Year's Eve attack in New
Orleans. On that day, Shamsud-Din Jabbar was killed in a shootout with
police after driving a Ford pickup truck with an Islamic State (IS)
flag through a New Year's crowd on Bourbon Street, killing 14 and
injuring 57. Prior to the attack, he visited New Orleans twice, on
October 31 and November 10 last year. During his October trip, he
recorded a video of the French Quarter using smart glasses from Meta.
Jabbar also expressed support for IS in videos posted to Facebook,
researched the December 2016 car attack at a Christmas market in
Germany, and pledged allegiance to the group shortly before the
attack.\1\
---------------------------------------------------------------------------
\1\ ``IS-Inspired Attacker Killed After Driving Through Crowd in
New Orleans,'' Islamic State Worldwide Activity Map, January 1, 2025,
https://www.washingtoninstitute.org/islamicstateinteractivemap/#view/
4029.
---------------------------------------------------------------------------
The following week, the Islamic State wrote an editorial in its
weekly newsletter called al-Naba, titled ``We Were There!,'' where it
gloated over Jabbar's attack.\2\ In the editorial, it highlighted the
group's influence and incitement capabilities. The piece also boasted
that the perpetrator used American technology, referring to the Meta
smart glasses he employed to conduct reconnaissance. The piece
concluded by once again urging Muslims in Europe and the United States
to carry out more terrorist attacks. Highlighting how IS uses one
attack to push for a new one and creates a virtuous cycle from its
perspective. That is why it is not surprising that Jabbar himself,
prior to his own attack, researched a prior IS attack: the December
2016 Christmas Market car ramming attack in Berlin that killed 12 and
injured 48.\3\
---------------------------------------------------------------------------
\2\ The Islamic State, ``al-Naba' Newsletter Issue #477,'' January
9, 2025, https://jihadology.net/2025/01/09/new-issue-of-the-islamic-
states-newsletter-al-naba-477.
\3\ ``Attack on Berlin Christmas Market,'' Islamic State Worldwide
Activity Map, December 20, 2016, https://www.washingtoninstitute.org/
islamicstateinteractivemap/#view/989.
---------------------------------------------------------------------------
Jabbar's attack also falls in line with Islamic State instructional
attack planning. In particular, in mid-November 2016, in IS's English
language magazine at the time called Rumiyah, it released an article
titled ``Just Terror Tactics'' that provided guidance on the best way
to kill as many of their enemies as possible.\4\ It says ``the type of
vehicle most appropriate for such an operation is a large load-bearing
truck.'' Jabbar did this with the pickup truck. The article also
highlights that if one doesn't have the wealth, it suggests renting a
vehicle, again Jabbar did this as well. Further, it suggests specific
targets, with one of them being pedestrian-congested streets, again
something Jabbar followed. The article also suggested that an attacker
can use a secondary weapon. In this case, Jabbar placed 2 coolers with
explosive devices at 2 other locations in the city's French Quarter.
Finally, the article tells prospective attackers to find ``an
appropriate way . . . for announcing one's allegiance to the
Caliphate.'' Not only did Jabbar do this with his pledge of allegiance
on Facebook, but also by raising the Islamic State flag on the truck he
used during the attack. This all shows that while this attack occurred
in 2025, in many ways examples and guidance literature from almost a
decade ago that remains accessible on-line has a long shelf life and
that not enough is still being done to make sure that potential
attackers do not have access to such content on-line.
---------------------------------------------------------------------------
\4\ The Islamic State, ``Rome Magazine Issue #3,'' November 11,
2016, https://jihadology.net/2016/11/11/new-release-of-the-islamic-
states-magazine-rome-3.
---------------------------------------------------------------------------
background
Unfortunately, none of this is new. Since the commercial internet
came about, jihadis have been there in parallel. The first known jihadi
presence on the internet can be traced back to 1991, with the Islamic
Media Center (IMC). Al-Qaeda's official debut dates to February 2000,
with the creation of maalemaljihad.com. This was followed in March 2001
by alneda.com, which was active through mid-July 2002.\5\ In the summer
of 2001, al-Qaeda created a media arm, al-Sahab Media Production
Establishment, and released its first video, ``The Destruction of the
American Destroyer [USS] Cole.'' Several other websites at the time
were not directly connected with al-Qaeda, but sympathized with its
jihadi worldview, including Azzam Publications, al-Tibyan Publications
(which had one of the earliest jihadi-leaning, English-language,
interactive forums), and Sawt al-Qawqaz.\6\
---------------------------------------------------------------------------
\5\ Abdel Bari Atwan, The Secret History of al-Qaeda, London: Saqi
Books, 2006, pp. 127; Patrick Di Justo, ``How Al-Qaida Site Was
Hijacked,'' Wired Online, August 10, 2002, http://www.wired.com/
culture/lifestyle/news/2002/08/54455.
\6\ For more see: Hanna Rogin, ``Al-Qaeda's on-line media
strategies--From Abu Reuter to Irhabi 007,'' Norwegian Defence Research
Establishment (FFI), January 12, 2007, http://rapporter.ffi.no/
rapporter/2007/02729.pdf.
---------------------------------------------------------------------------
Since then, the jihadi movement has taken advantage of new on-line
technologies at every turn to spread its message, recruit individuals
to fight abroad, incite or help plan attacks, and raise money. For
example, the onset of interactive forums in the mid-2000's, concurrent
with the rise of Abu Mus'ab al-Zarqawi and the Iraq jihad, shattered
the elitist nature of jihadi communications. Web forums still offered
administrators (who were often directly connected with al-Qaeda)
extensive influence over what was posted because they could delete
threads or ban members. But individual forum members, not directly
connected to al-Qaeda, could not only view what was posted by
administrators, but also comment and post their own content as well.\7\
Mustafa Setmariam Nasir, better known by his nom de guerre Abu Mus'ab
al-Suri, called for producing jihadi media in languages other than
Arabic, including English, and devising messages that appealed more to
the masses. The popularization of the on-line jihadi movement empowered
organizations dedicated to translating material, most of which was
still produced in Arabic. The Global Islamic Media Front (GIMF),
established in August 2004, was a key innovator in this regard, and
could trace its roots all the way back to June 2001.\8\
---------------------------------------------------------------------------
\7\ Gordon Corera, ``Al-Qaeda's 007,'' The Times, January 16, 2008,
https://www.thetimes.com/article/al-qaedas-007-c2sx2r5bdgc.
\8\ Rogin, ``Al-Qaeda's on-line media strategies,'' pp. 56.
---------------------------------------------------------------------------
After this, Web 2.0 innovations and the creation of social media
platforms (blogging, Facebook, YouTube, and Twitter) flattened control
over the production of on-line jihadi media. Social media platforms
enabled global jihadi entrepreneurs to share news items, original
articles and essays, tribute videos, and anashid (Islamically
sanctioned a cappella music). The newer technologies, at that time,
lowered the bar for participation, making the involvement of low-level
or non-jihadis in on-line conversation a new feature of the global
jihadi movement. Those so inclined could talk about jihad all day on
the Web, even if they were geographically dispersed. This was not
possible beforehand.\9\
---------------------------------------------------------------------------
\9\ Aaron Y. Zelin, ``The State of Global Jihad Online,'' New
America Foundation, February 2013, https://www.newamerica.org/future-
security/policy-papers/the-state-of-global-jihad-online.
---------------------------------------------------------------------------
As a consequence, when the Islamic State eclipsed al-Qaeda in the
mid-2010's they used this to deadly effect.\10\ To further their
message they would create innocuous hashtag aggregators yet only post
their propaganda, create hashtag targeting bots, have multiple backup
accounts in case they were taken down, and build the Fajr al-Basha'ir
app.\11\ The latter let members and supporters connect to it and
thereby whenever the app tweeted something it would automatically be
posted by those that signed up for it onto their own account. The
breadth of IS's on-line Twitter campaign was unprecedented. However,
this would not last, due to massive complaints by countries reeling
from the Islamic State's on-the-ground successes in Iraq and Syria and
tens of thousands of foreign fighters being recruited to join their
ranks.\12\
---------------------------------------------------------------------------
\10\ Aaron Y. Zelin, ``Picture Or It Didn't Happen: A Snapshot of
the Islamic State's Official Media Output,'' Perspectives on Terrorism,
Volume 9, Number 4, August 2015, https://www.jstor.org/stable/
26297417?seq=1.
\11\ J.M. Berger, ``How ISIS Games Twitter,'' The Atlantic, June
16, 2014, https://www.theatlantic.com/international/archive/2014/06/
isis-iraq-twitter-social-media-strategy/372856.
\12\ Michael Isikoff, ``Twitter under pressure to act more
aggressively against terrorists,'' February 18, 2015, https://
www.yahoo.com/news/amphtml/politics/twitter-under-pressure-to-act-more-
aggressively-114435221601.html.
---------------------------------------------------------------------------
That led Twitter (and all the other main technology companies) to
originally establish their Trust and Safety teams. This led to a
crackdown on IS networks in 2015 by going after IP addresses and those
within their follower networks.\13\ This helped also establish the
Global Internet Forum to Counter Terrorism (GIFCT), which established a
consortium of Western technology companies to work together to take
down terrorist content by sharing digital fingerprints (or hashes) of
different types of content (pictures, audio, videos, etc.).\14\ As a
consequence, IS and the jihadi movement shifted to the encrypted
messaging application Telegram in August/September 2015, which also had
a broadcast feature that allowed anyone to follow official IS and other
groups' channels.\15\ Telegram never had the same utility as Twitter
since it couldn't just reach anyone randomly as Twitter had, one had to
know where to go ahead of time to find the content. This takedown cycle
eventually happened again on Telegram when Europol convinced the
company to go after jihadi accounts, which led to a huge purge in
November 2019.\16\ As a consequence, both IS and AQ networks
established their own decentralized forums using block chain technology
on RocketChat to make it more difficult for content to be taken down.
To this day, both RocketChat forums remain on-line.\17\
---------------------------------------------------------------------------
\13\ ``An update on our efforts to combat violent extremism,''
Twitter, August 18, 2016, https://blog.x.com/en_us/a/2016/an-update-on-
our-efforts-to-combat-violent-extremism.
\14\ ``Global Internet Forum to Counter Terrorism: an update on our
efforts to use technology, support smaller companies and fund research
to fight terrorism on-line,'' June 18, 2018, https://gifct.org/2018/06/
18/global-internet-forum-to-counter-terrorism-an-update-on-our-efforts-
to-use-technology-support-smaller-companies-and-fund-research-to-fight-
terrorism-on-line.
\15\ ``IS exploits Telegram mobile app to spread propaganda,'' BBC
News, October 7, 2015, https://www.bbc.com/news/technology-34466287.
\16\ ``Europol and Telegram take on terrorist propaganda on-line''
Europol, November 22, 2019, https://www.europol.europa.eu/media-press/
newsroom/news/europol-and-telegram-take-terrorist-propaganda-on-line.
\17\ Peter King, ``Analysis: Islamic State messaging on RocketChat
still on-line after 7 months,'' BBC Monitoring, August 9, 2019, https:/
/monitoring.bbc.co.uk/product/c200zjhz; ``Veteran jihadist outlet uses
RocketChat for al-Qaeda propaganda,'' BBC Monitoring, December 6, 2019,
https://monitoring.bbc.co.uk/product/c201akwr.
---------------------------------------------------------------------------
harnessing the internet today
While al-Qaeda still operates its RocketChat forum, the Islamic
State's on-line ecosystem and infrastructure is far more diverse and
sophisticated than only its own RocketChat forum. In addition to that,
IS also established its own Cloud-based archive of its historical
propaganda called Obedient Supporters.\18\ Moreover, in recent years,
it has also established a number of traditional websites. In some ways,
a return to the beginning of the internet in the 1990's and early
2000's since it was much harder to operate on mainstream social media
platforms. To make these websites more difficult to take down, they
jump domain names often using different country codes to hide in plain
sight. To make it even more complicated, they have also developed
mirrored versions of these websites on the Dark Web, which can only be
accessed using a Tor browser and an Onion router link. Each website
also provides a specific purpose to try and break up where all the
content is so as to disperse it to make their on-line network more
resilient over time. They do this by having a repository website called
Fahras al-Ansar, which is based in South Africa right now at least,
that shares the last links for each site that are currently available
on-line on the surface web and on the Dark Web.\19\ The main website
that has been most active recently is called Sah al-Wagha, which shares
the latest IS attack claims of responsibility, videos, and weekly
newsletter al-Naba.\20\
---------------------------------------------------------------------------
\18\ Miron Lakomy, ``In the digital trenches: Mapping the structure
and evolution of the Islamic State's information ecosystem (2023-
2024),'' Media, War & Conflict, August 24, 2024, https://
journals.sagepub.com/doi/10.1177/17506352241274554.
\19\ Ibid.
\20\ ``Briefing: New archive of official IS material appears on-
line,'' BBC Monitoring, October 28, 2024, https://monitoring.bbc.co.uk/
product/b0002o3c.
---------------------------------------------------------------------------
The latter website is where one would find official media content
from IS, which comes directly from its Central Media Diwan
(Administration), which today includes al-Furqan Media, Amaq News
Agency, Provincial Media Centers, among other lesser-used outlets
today. Other websites and the RocketChat forum also disseminate
``unofficial'' auxiliary propaganda. These are not created by the
Central Media Diwan, but by members of the group in their own capacity
and work with on-line supporters of the group. These include media
outlets such as al-Batar Media, al-Saqri Media, al-Dir'a al-Sunni
Media, Sirat al-Khilafah, al-`Adilat Media, etc. The next layer under
this is IS's translation collective, called Fursan al-Tarjuma, which
helps disseminate all of its official propaganda into dozens of
languages.\21\ Again, this is ``unofficial'' insofar as the Central
Media Diwan is not involved, but rather members of the group in their
own capacity and on-line supporters take part in it. These outlets
today include Halammu (English), Nida al-Haqq (Urdu), al-Aza'im
(Pashto, Uzbek, Tajik, Farsi), Markaz al-Nur (French), al-Tamkin
(Bengali, Indonesian), Arshad (Russian), al-Bahiriyah (Hausa), al-
Basha'ir (Dhivehi), Maydan (Turkish), Rastara (Kurdish), Fakhr al-Ummah
(Albanian), Sawt al-Andalus (Spanish), and al-Qital (Hindi).
Illustrating the breadth of languages IS has access to for spreading
its message, ideology, and incitement all across the world. At the
height of IS's territorial control in Iraq and Syria a decade ago, they
translated their content into even more languages, illustrating that
even if it is relatively weaker today, it is still quite resilient and
there are enough people interested in its worldview to assist in these
endeavors.
---------------------------------------------------------------------------
\21\ Lucas Webber and Daniele Garofalo, ``Fursan al-Tarjuma Carries
the Torch of Islamic State's Media Jihad,'' Global Network on Extremism
and Technology, June 5, 2023, https://gnet-research.org/2023/06/05/
fursan-al-tarjuma-carries-the-torch-of-islamic-states-media-jihad.
---------------------------------------------------------------------------
emerging tech: crypto, live streaming, and ai
Beyond traditional propaganda efforts to recruit and incite, due to
the greater reach and ease of use of cryptocurrencies, there has been a
huge uptick in its use by jihadi groups, its supporters, and those
involved in attacks abroad. For example, on an official level,
beginning in December 2023, the Islamic State's Khurasan Province
(ISKP), based out of Afghanistan and Pakistan, began promoting its own
wallet for the Monero cryptocurrency to help fund its efforts locally
and also as a conduit to pay for external operations abroad. These
promotions for ISKP's Monero wallet first appeared in its official
English-language magazine the Voice of Khurasan.\22\ The utility of a
cryptocurrency like Monero is that it is more difficult to track the
movement of money through the wallet like other cryptocurrencies. Thus
making it more secure, which in of itself for normal activity, on the
surface, would be fine, but for a terrorist group, not ideal. Since
then, ISKP has promoted a number of different wallets in subsequent
issues of its Voice of Khurasan magazine. This is not theoretical
either, back in April 2024, the FBI arrested 18-year-old Alexander
Scott Mercurio who pledged allegiance to the Islamic State and plotted
to attack churches in Coeur d'Alene, Idaho. As part of this plot, he
confided in an undercover agent that he wanted to donate his money
($11,000) ahead of this attack to ISKP using Monero.\23\ A similar case
in the United Kingdom occurred as well with an even larger donation
attempt of 16,000.\24\ Based on data collected for my
Islamic State Worldwide Activity map, since 2015, there have been 36
arrest cases globally related to jihadi use of cryptocurrency, with 13
of them happening in 2024 alone, illustrating a huge uptick in only the
past year. And these cases are only known cases that have come through
judicial manners, not those that were not detected.
---------------------------------------------------------------------------
\22\ The Islamic State's Wilayat Khurasan, ``Voice of Khurasan
Magazine Issue #31,'' al-Aza'im Media, December 22, 2023, https://
jihadology.net/2023/12/22/new-magazine-issue-from-the-islamic-states-
wilayat-khurasan-voice-of-khurasan-31.
\23\ ``Idaho Teen Arrested For Plotting Church Attacks,'' Islamic
State Worldwide Activity Map, April 6, 2024, https://
www.washingtoninstitute.org/islamicstateinteractivemap/#view/3228.
\24\ ``Luton Man Arrested for Sending Cryptocurrency to ISKP,''
Islamic State Worldwide Activity Map, March 13, 2024, https://
www.washingtoninstitute.org/islamicstateinteractivemap/#view/4191.
---------------------------------------------------------------------------
Beyond traditional American social media platforms, there has also
been a rise in use of TikTok by supporters of IS. This is not
surprising since it has become ubiquitous amongst Gen Z individuals.
While the trend likely began earlier, based on data from my Islamic
State Worldwide Activity map, there have been 15 arrest cases that
began in 2023 that have shown the arrestee to have been involved in one
way or another with IS propaganda on TikTok, either sharing it
themselves or watching it. Just last week, Minneapolis resident
Abdisatar Ahmed Hassan was arrested and charged for attempting to
provide material support to the Islamic State. As part of the
investigation, he actually praised the perpetrator of the IS-inspired
attack in New Orleans on TikTok on January 1, 2025.\25\ Highlighting
how one attack could inspire others to either plot their own attack or
to try and travel abroad to fight with IS in a warzone, as Hassan
attempted to do with IS in Somalia.
---------------------------------------------------------------------------
\25\ ``Minneapolis Man Arrested for Attempting to Provide Material
Support to ISIS,'' Department of Justice, February 28, 2025, https://
www.justice.gov/usao-mn/pr/minneapolis-man-arrested-attempting-provide-
material-support-isis.
---------------------------------------------------------------------------
There are also other technological-related issues that terrorists
could attempt to exploit, including live-streaming their attack. It
could have had much more of a psychological effect nationally, if the
New Orleans attacker had live-streamed his attack on his Meta glasses
through Facebook instead of only using it for reconnaissance purposes.
However, such use is not unprecedented. Back in June 2016, Islamic
State attacker Larossi Abballa, broadcast the aftermath of his attack
on a French police captain and his partner, in real time on Facebook
Live and remained on-line for almost 12 minutes.\26\ A copy of the
recording was downloaded and later reposted via IS's official media
outlet Amaq News Agency, since it was taken down from Facebook 11 hours
after its recording.
---------------------------------------------------------------------------
\26\ Caitlin Dewey and Sarah Parnass, ``For the first time, an
alleged terrorist has broadcast a confession in real time on Facebook
Live,'' Washington Post, June 14, 2016, https://www.washingtonpost.com/
news/the-intersect/wp/2016/06/14/for-the-first-time-an-alleged-
terrorist-has-broadcast-a-confession-in-real-time-on-facebook-live.
---------------------------------------------------------------------------
In addition to that, there are also worries that Generative
Artificial Intelligence (AI) could be exploited by terrorists. However,
thus far, at least among jihadis, there has not been much evidence that
they have used it yet to deadly ends. However, that does not
necessarily mean they couldn't in the future. For now, it has mainly
been used by followers to create on-line graphics to promote the same
kind of content they would have previously using Photoshop. Even if
this appears to have less stakes, it does lower the bar for those to
make content and interact more deeply with the ideology and worldview
since one doesn't need the same level of skill set as someone that
knows how to use Photoshop, something which takes training and a lot
more time. However, there have been inchoate conversations by IS
supporters on Rocketchat beginning a few weeks ago, related to the
Chinese application Deepseek and how they could potentially exploit it.
It is too soon, however, to know how that might evolve.\27\
---------------------------------------------------------------------------
\27\ ``Users on pro-IS chat group begin discussing DeepSeek,'' BBC
Monitoring, February 19, 2025, https://monitoring.bbc.co.uk/product/
b0003e3h.
---------------------------------------------------------------------------
moderation backsliding
Beyond the specifics of how terrorists might exploit technology,
policies related to technologies and platforms also have a role in
providing space or making it very difficult to use. As noted above,
beginning in 2015, there was a much greater focus by major technology
companies to moderate their platforms so that terrorists couldn't
exploit them. While the moderation was in no way perfect vis-a-vis
jihadis on-line, overall, it curbed usage on mainstream platforms at a
good enough rate that it was not noticeably affecting random people as
it easily had in the 2013-15 time period. However, in recent years, due
to controversies related to alleged censorship within the West
politically, there has been a backlash to moderating content even if it
is extremist in nature. As a consequence, beginning with Elon Musk's
purchase of Twitter, which is now X, the level of content moderation in
general, and related to the jihadi movement specifically has
backslid.\28\ It should not be overblown, however, since it is not as
widespread as it was in 2015 prior to the offensive policies against
jihadis, but in a relative sense, there is greater space for jihadis to
exploit platforms in recent years.\29\ It should be noted that this is
not just an issue with X, but has Mark Zuckerberg's Meta social media
platforms as well, including Facebook, Instagram, and WhatsApp as well
as the aforementioned TikTok. Yet unlike a decade ago, much of these
relative capabilities by IS networks online are by supporters of the
group and not at the official level.\30\
---------------------------------------------------------------------------
\28\ ``Musk admitted to firing 80 percent of Trust and Safety
Engineers at Twitter,'' January 12, 2024, https://sarajevotimes.com/
musk-admitted-to-firing-80-of-trust-and-safety-engineers-at-twitter;
``Musk's Twitter has dissolved its Trust and Safety Council,''
Associated Press, December 12, 2022, https://www.npr.org/2022/12/12/
1142399312/twitter-trust-and-safety-council-elon-musk.
\29\ Moustafa Ayad, ``Islamic State Supporters on Twitter: How is
`New' Twitter Handling an Old Problem?,'' Global Network on Extremism
and Technology, November 18, 2022, https://gnet-research.org/2022/11/
18/islamic-state-supporters-on-twitter-how-is-new-twitter-handling-an-
old-problem.
\30\ Moustafa Ayad, ``Teenage Terrorists and the Digital Ecosystem
of the Islamic State,'' CTC Sentinel, Volume 18, Issue 2, February
2025, https://ctc.westpoint.edu/teenage-terrorists-and-the-digital-
ecosystem-of-the-islamic-state.
---------------------------------------------------------------------------
recommendations
The U.S. Government should urge technology companies and
social media platforms to redouble their efforts at content
moderation related to the jihadi movement. In particular,
beyond only Arabic and English content, these platforms need to
beef up their moderation in languages that are increasingly
used as the center of gravity of the jihadi movement on-line
such as multiple languages in Africa and Central Asia for
example.
Although there have been recent calls to cut funding and
jobs across the U.S. Government, cutting ones related to
tracking on-line jihadi recruitment and attack plotting could
undermine future security and lead to greater risks at home and
abroad. At a time, when greater resources are put toward power
competition and less resources are given to counterterrorism,
eliminating even more resources in this field could provide
more opportunities for adversarial jihadis and entrepreneurial
supporters of the movement to take advantage and attack the
homeland more easily.
While there have been many discussions about the utility of
using AI in terrorism investigations and content moderation on-
line, it still does not replace human expertise and contextual
clues on these issues whether within tech companies or the U.S.
Government.
The U.S. Government should also urge GIFCT to establish a
whitelist for researchers that work on these sensitive issues
so that their accounts do not get mistakenly taken down while
actual terrorist accounts are targeted. This has been a problem
in the past and should be resolved.\31\
---------------------------------------------------------------------------
\31\ Aaron Y. Zelin, `` `Highly nuanced policy is very difficult to
apply at scale': Examining researcher account and content takedowns on-
line,'' Policy & Internet, Volume 15, Issue 4, December 2023, https://
on-linelibrary.wiley.com/doi/full/10.1002/poi3.374.
Mr. Pfluger. Thank you, Dr. Zelin.
The Chair now recognizes Mr. Flesch for his opening
statement.
STATEMENT OF DANIEL FLESCH, SENIOR POLICY ANALYST, MIDDLE EAST
AND NORTH AFRICA, ALLISON CENTER FOR NATIONAL SECURITY, THE
HERITAGE FOUNDATION
Mr. Flesch. Good afternoon, Chairman Pfluger and Ranking
Member Magaziner and Members of the subcommittee. Thank you for
the opportunity to testify on this important topic today.
In addition to my work in the Middle East at the Heritage
Foundation, I also manage Heritage's National Task Force to
Combat Antisemitism, a coalition of organizations committed to
defeating the current anti-American wave of antisemitism in the
United States.
From my experience, including, sir, as you mentioned,
serving in the IDF, I saw how terror organizations engaged in a
war against Israel are also recruiting and radicalizing
Americans. Put another way, there's a direct connection between
the war for survival Israel is fighting in the Middle East and
the one America, whether we are aware of it or not, is fighting
at home.
The threat today must be understood in the context of
Hamas' October 7th attack. Within hours of Hamas executing the
deadliest day for the Jewish people since the Holocaust, many
across this country took to the streets against the Jewish
state, our ally, and in solidarity with a U.S.-designated
terrorist organization. Seventeen months later, they have not
relented. Just last week, a mob of students wearing masks and
keffiyehs and shouting Free Palestine at Barnard College in New
York City became violent, leading to several arrests. Among
many things, October 7th revealed the extent to which a certain
population within our country is already or inclined to become
radicalized.
It is telling that Hamas' invasion of Israel and the
massacre and abduction of civilians, including Americans, was a
spark that prompted this population to reveal itself. Consider
the now-commonplace call for an intifada. Those who threaten to
globalize the intifada make clear that their goal is not to
affect U.S. policy toward Israel or the Middle East, but to
destroy the United States through violent political revolution
that features a Palestinian-inspired strategy of indiscriminate
attacks on civilians to include the murdering of children.
These intifada-supporting organizations and individuals and the
useful idiots that march or encamp alongside them, presents the
most fertile ground for recruitment by terrorist groups. This
population is most susceptible to the propaganda of Hamas and
other extremist factions within the Palestinian and broader
Arab world, who depict the graphic violence of the October 7th
attack and the ensuing war as a form of martyrdom. They share
videos and images on social media platforms like X, Facebook,
and Instagram, and to rally support, spread them on encrypted
messaging apps like Telegram.
Extremist groups and supporters turn these platforms and
apps into a radicalizing propaganda tool. They post unverified
and highly disturbing content that lacks context or factual
checks. This is designed to elicit a raw emotional reaction,
which is best to reach younger generations. Organizations like
Students for Justice in Palestine, SJP, and others rebroadcast
and disseminate these messages and images. They invite their
social media account followers to various trainings and to join
more private communication channels such as Signal, where they
may conduct their actual planning and coordination. In framing
the October 7th violence, for example, as a noble act in the
struggle against an occupying force, Hamas linked in other
extreme channels cross-pollinate ideas from different groups,
including those agitating against Israel, the Jews, the United
States, and the West.
Last summer, the director of national intelligence revealed
that the Islamic Republic of Iran was also ``posing as
activists on-line seeking to encourage protests and even
providing financial support to protesters.'' The combined
effect of these efforts is to radicalize adherence to engage in
violent criminal activity. Consider just some recent examples.
In November, FBI and local police found rifles, ammunition, an
explosive device in the home of 2 Palestinian-American sisters,
who are students at George Mason University and leaders in
their campus' SJP chapter. They had previously defaced the
student center with threats of a student intifada. On New
Year's Day, as we've discussed, hours after an ISIS-inspired
terrorist killed 14 civilians in a deadly car attack in New
Orleans, pro-Palestinian protesters marched in Times Square
chanting for an intifada. One of the organizers of the Barnard
College protest from last week, Columbia SJP, proudly and
publicly proclaims on its X profile, ``Long live the student
intifada and glory to our martyrs.''
Terror organizations are actively radicalizing and
recruiting an impressionable population for their nefarious
means. Efforts should be taken to detect, deter, and defeat
their networks, including their organizational supporters and
foot soldiers in America. Israel defeated the Palestinian
second intifada by simultaneously playing offense and defense.
We need such a dual approach today, one that sees law
enforcement be proactive against this threat in a civil society
that creates its own barrier, beyond which today's anti-
American form of antisemitism is no longer tolerated. If we
fail to act now, it is only a matter of time before we may see
an intifada on American streets. Thank you.
[The prepared statement of Mr. Flesch follows:]
Prepared Statement of Daniel Flesch
March 4, 2025
My name is Daniel Flesch. I am the senior policy analyst for the
Middle East and North Africa at The Heritage Foundation. The views I
express in this testimony are my own and should not be construed as
representing any official position of The Heritage Foundation.
The role of the digital space in how terrorists recruit and
radicalize adherents is an important and particularly timely one
because we are seeing it play out in real time across this country.
At The Heritage Foundation, I focus on U.S. policy toward Israel
and the region. I also manage Heritage's National Task Force to Combat
Antisemitism (NTFCA), which is a coalition of organizations committed
to combating the current wave of antisemitism in the United States,
which is a particular anti-American form of antisemitism.
These 2 policy orientations enable me to see commonalities between
how the terrorist organizations fighting Israel are also recruiting and
radicalizing Americans. Put another way, I see the direct connection
between the war Israel is fighting for survival in the Middle East and
the one America, whether we are aware of it or not, is fighting at
home.
This is the genesis behind the establishment of the NTFCA and the
drafting of Project Esther: a National Strategy to Combat
Antisemitism.\1\ Project Esther recognizes that today's form of
antisemitism that has exploded across the country in the wake of the
October 7th attack on Israel by the virulently anti-Israel, anti-
Zionist, and anti-American ``pro-Palestinian movement'' is part of a
global Hamas Support Network (HSN) that is trying to compel the U.S.
Government to abandon its long-standing support for Israel. Supported
by activists and funders dedicated to the destruction of capitalism and
democracy, the HSN benefits from the support and training of America's
overseas enemies and, within the United States, receives the support of
a vast network of activists and funders. This network has a much more
ambitious, insidious goal--the destruction of capitalism and democracy.
As their ends align, the HSN and its nihilist supporters indoctrinate
the gullible into supporting Hamas and hating Israel to create the
street mayhem that serves their ends.
---------------------------------------------------------------------------
\1\ Project Esther: A National Strategy to Combat Antisemitism/The
Heritage Foundation.
---------------------------------------------------------------------------
In this ecosystem, Hamas and other terrorist organizations and
countries are able to recruit and radicalize a willing population on-
line with a level of ease and openness that cannot be understated.
It is easy because there are a plethora of social media and digital
communication tools--including Instagram, TikTok, Telegram, Signal and
others--that organizations and individuals can use to engage with each
other.
It is open because many of these organizations do not hide who they
are and communicate publicly about their purpose, activities, and
intent. Though organizations often use multiple messaging platforms
that feature various levels of security, much of their material is
intended for public consumption. They are intentional about
broadcasting their presence to attract adherents or sympathizers.
The current and immediate threat of malign actors--both foreign and
domestic--recruiting and radicalizing within the United States must be
understood in our post-October 7th environment.
The October 7th invasion and massacre of southern Israel--in which
Hamas, Palestinian Islamic Jihad and Palestinian civilians massacred
over 1,200 people and abducted over 250 others back into Gaza,
including Americans--catalyzed, or revealed, a unique radicalization
that has occurred in America, and across the West more broadly.
In many ways, October 7th was the spark that caused the tinder to
burst into a giant, all-consuming flame.
Within hours of the massacre unfolding in Israel, many across this
country took to the streets in support of Hamas, which the U.S. State
Department designates as a terrorist organization and had just
perpetrated the deadliest day for the Jewish people since the
Holocaust.
Under the guise of protesting our ally's legitimate response to its
people being butchered, mutilated, raped and kidnapped, we saw
Americans waving Palestinian and Hamas flags and shouting slogans such
as ``we don't want 2 States, we want 1948'' and ``from the river to the
sea, Palestine will be free.''
These and other evocations demonstrate that rather than seek
accommodation and a just solution to the war, these protestors view
themselves as active participants and are advocating for the
destruction of Israel, the world's only Jewish state and a U.S. ally.
And their calls for violence have only increased, including calling
to ``globalize the intifada'' and threatening ``there is only one
solution: intifada, revolution''.
In calling for an ``intifada'', the protestors refer to 2 periods
of sustained violent Palestinian revolt against Israel. The first
intifada (1987-93) ended with the onset of the Oslo peace process, as
Israelis believed Palestinian violence stemmed from their desire for
independence. The second intifada (2000-05) began with the failure of
Oslo as Israelis learned that rather than seek their own state, the
Palestinians sought to destroy the Jewish state. And they would seek to
achieve this through any means necessary, including targeting
civilians.\2\
---------------------------------------------------------------------------
\2\ Is Intifada Coming to America?/RealClearPolitics.
---------------------------------------------------------------------------
Prior to Oct. 7, the second intifada was the most traumatic period
in Israeli history. In its first full year, Palestinian suicide bombers
targeted buses, pizza shops, nightclubs, and other ``soft'' targets,
killing over 100 civilians, including Americans. In a country where it
is commonplace for children to take public transportation to school,
parents could no longer trust their kids would return home alive. By
the time it ended, over 1,000 Israelis and 2,700 Palestinians were
dead, with thousands more injured.
This is what those calling to ``globalize the intifada'' mean to
import to American streets. Their goal is not to affect U.S. policy
toward Israel or the Middle East, but to destroy the United States
through political revolution that features a Palestinian-inspired
strategy of indiscriminate attacks against civilians, to include the
murdering of children.
Those agitating for an intifada have used Hamas' invasion and
massacre as a pretext to engage in further violence against Jews,
including blocking Jewish congregants from entering synagogues while
shouting ``long live intifada,'' stabbing a Jewish man while shouting
``free Palestine,'' fatally hitting a Jewish counter-protestor in an
altercation, and many other incidents.\3\
---------------------------------------------------------------------------
\3\ Ibid.
---------------------------------------------------------------------------
Nowhere has discrimination, threats, and violence against Jews and
Americans been more concentrated than on university campuses. According
to the ADL, in the first 2 months of the Fall 2023 semester, 73 percent
of Jewish students had witnessed or experienced antisemitism.\4\ In
November of last year, a Congressional investigation into antisemitism
on college campuses found that ``our most prominent American
universities refused to crack down on antisemitism'', including those
``students who engaged in antisemitic behavior, encampments, and
intimidating tactics such as campus checkpoints and tax-exempt
organizations that enabled and funded violent campus protests.''\5\
---------------------------------------------------------------------------
\4\ The Alarming Surge of Antisemitism on College Campuses/No
Tolerance for Antisemitism.
\5\ edworkforce.house.gov/uploadedfiles/
10.30.24_committee_on_education_and_the_work-
force_republican_staff_report_antisemitism_on_college_campuses_exposed.p
df.
---------------------------------------------------------------------------
While the past 16 months have featured the greatest display of
overt antisemitism, anti-Zionism, and anti-Israel agitation this
country has ever seen, it is all a means to disrupting, degrading, and
destroying the fabric of American society and Western civilization.
Many of those who engage in this antisemitic and criminal activity
may best be described as ``useful idiots'' that are supporting the
progressive cause du jour; in many cases, they cannot tell you from
which river to which sea ``Palestine will be free.''
Yet more than the presence of these useful idiots who wave flags or
wear the apparel or other paraphernalia of U.S.-designated foreign
terrorist organizations, or march alongside or pitch a tent at an
encampment of those who support the same, it is the pervasiveness of
those who call for an ``intifada'' that demonstrates there is already a
sizable contingent of radicalized organizations and individuals in this
country, which presents the most fertile ground for recruitment.
To engage this willing population, Hamas and other extremist
factions within the Palestinian and broader Arab world use the graphic
violence of the October 7th attack and ensuing war as a form of
``martyrdom'' to rally support. These groups portray individuals
involved in such attacks as heroes, with the violence being framed as a
noble act in the struggle against an occupying force. By emphasizing
the idealized heroic nature of violence, this framing targets younger
generations.
Graphic videos and images are shared across social media platforms
like X, Facebook, and Instagram, and spread on encrypted messaging apps
like Telegram. Extremist groups and supporters post unverified and
highly disturbing content that lack context or factual checks, ensuring
that raw emotional reactions take precedence over thoughtful analysis,
effectively turning these platforms and apps into a propaganda tool to
fuel anger and hate.
These messages and images are then rebroadcast and disseminated by
organizations in America, such as Students for Justice in Palestine
(SJP), Within Our Lifetime (WOL), Jewish Voices for Peace (JVP), and
others.
These groups invite their account followers to trainings and to
join more private communication channels, such as Signal, where they
may conduct their actual planning and coordination.
Telegram is also used by Hamas-linked and other extreme channels to
cross-pollinate ideas from different groups, including those agitating
against Israel, against Jews, against colonialism, against imperialism,
against the United States, against the West, etc.
Throughout all this, efforts are made to remain evasive. Account
owners and members often change their names. They use coded language to
demonstrate alignment across organizations. The intent is not to hide
but to maintain a veneer of plausible deniability.
This is the ease and openness with which terrorist groups and U.S.-
based organizations that advocate for an ``intifada'' on American
streets coordinate, plan, and execute their activities.
And they are not alone in this regard. In July of last year, the
Director of National Intelligence Avril Haines released a statement
noting that ``Iranian government actors have sought to
opportunistically take advantage of on-going protests regarding the war
in Gaza.'' The statement continued to note that the intelligence
community has ``observed actors tied to Iran's government posing as
activists on-line, seeking to encourage protests, and even providing
financial support to protestors.''\6\
---------------------------------------------------------------------------
\6\ Statement from Director of National Intelligence Avril Haines
on Recent Iranian Influence Efforts.
---------------------------------------------------------------------------
Hamas has also used public communication channels to disseminate
talking points in the United States and radicalize susceptible
Americans. In August 2024, the Palestine Chronicle, a news outlet run
by the People Media Project, a U.S.-based, tax-exempt nonprofit,
published an article \7\ about Khaled Meshaal, a Hamas leader, urging
university students to protest.
---------------------------------------------------------------------------
\7\ `Stop this Criminal Aggression'--Meshaal Urges Students to
Resume Campus Protests--Palestine Chronicle.
---------------------------------------------------------------------------
According to a lawsuit filed on February 21, 2025, one of the
writers for the Palestine Chronicle was Abdallah Aljamal,\8\ who held 3
Israelis--Almog Meir Jan, Andrey Kozlov, and Shlomi Ziv--hostage after
their abduction from the Nova music festival on October 7, 2023.
Aljamal, the lawsuit alleges, is a member of Hamas.
---------------------------------------------------------------------------
\8\ Gaza captor told hostages that Hamas collaborates with U.S.
campus protesters, lawsuit alleges/The Times of Israel.
---------------------------------------------------------------------------
The combined effect of this evasive and public digital
communication is to radicalize adherents to engage in violent, criminal
activity.
In November, local police and the FBI raided the home of 2
Palestinian-American sisters at George Mason University (GMU) in
Virginia and found rifles, ammunition, and an explosive device, along
with signs that read ``Death to the Jews'' and ``Death to America'' and
the flags of Hamas and Hezbollah, both U.S.-designated terror
organizations. The sisters are leaders in their campus's SJP chapter
and previously participated in an act of vandalism when they defaced
the student center with threats of a ``student intifada.''
In December, the FBI arrested an Egyptian national and GMU student
for plotting an attack on the Israeli consulate in New York City
because the ``building represented the `Yahud,' '' Arabic for ``Jew.''
Abdullah Ezzeldin Taha Mohamed Hassan had used X to communicate with
the FBI.\9\
---------------------------------------------------------------------------
\9\ Concerns Raised About Anti-Israel Sentiment at University.
---------------------------------------------------------------------------
On New Years day, hours after an ISIS-inspired terrorist killed 15
civilians in a deadly car attack in New Orleans, pro-Palestinian
protesters marched in Times Square in New York City chanting for an
intifada.
Particularly on college campuses, many foreign students are
engaging in this antisemitic and anti-American rhetoric and activity,
including demonstrating for U.S.-designated terror organizations, in
violation of their foreign student visa.
In recent news, a Georgetown graduate student was discovered to
have direct ties to Hamas. A Committee for Accuracy in Middle East
Reporting and Analysis (CAMERA) report found that Mapheze Saleh had
previously worked for ``the Ministry of Foreign Affairs in Gaza''--
which is Hamas--and her father is Ahmed Yousef, ``senior adviser to the
Hamas Foreign Ministry.''\10\ According to CAMERA, Saleh's ``social
media profiles have glorified Hamas and its acts of terror'' and
referred to America as ``the plague.''
---------------------------------------------------------------------------
\10\ https://www.msn.com/en-us/society-culture-and-history/human-
rights/university-whose-students-invited-terrorist-to-campus-also-
enrolled-former-hamas-official/ar-AA1z05St.
---------------------------------------------------------------------------
The revelation of Saleh's ties came amid a probe into an event the
Georgetown Law SJP planned for Ribhi Karajah, who spent 3.5 years in
prison for his involvement in the U.S.-designated terror group Popular
Front for the Liberation of Palestine's August 2019 roadside bombing
that killed a young Israeli woman, Rina Schnerb. Karajah has also
promoted PFLP material on his social media accounts.\11\
---------------------------------------------------------------------------
\11\ Georgetown pressured to cancel event with convicted member of
PFLP terror group/Fox News.
---------------------------------------------------------------------------
In response to the presence of and wide-spread support for U.S.-
designated terror organizations on campus from foreign students, the
Trump administration's Executive Order to remove those students who
provide material support to terror organizations and thereby pose a
specific threat of violence is a welcome step to protect Jewish and
non-Jewish students on campus. The threat is not to the free speech
rights of a few on campus but to the safety, protection, and well-being
of all students.
More broadly, efforts should be taken to detect, deter, and defeat
the terror organizations' networks, including their organizational
supporters and foot soldiers, in America. Those organizations that
support calls for an ``intifada'', provide material support for or
receive benefits from U.S.-designated terror groups should be
designated as hate groups or domestic terrorists and have their tax-
exempt status revoked.
In addition, local, State and Federal law enforcement should share
credible and actionable intelligence, including declassifying if and
when appropriate, of intent to commit criminal acts.
Terror organizations have found the soft underbelly of American
society and will radicalize and recruit an impressionable population
for their nefarious means. The ability to counter their efforts exists,
it just requires marshalling the requisite capabilities and attention
to stem the tide of antisemitism and save America.
Mr. Pfluger. Thank you, Mr. Flesch.
The Chair now recognizes Dr. Braddock for his opening
statement.
STATEMENT OF KURT BRADDOCK, PH.D., ASSISTANT PROFESSOR, PUBLIC
COMMUNICATION, AMERICAN UNIVERSITY
Mr. Braddock. Thank you, Chairman Pfluger. Thank you,
Ranking Member Magaziner, and distinguished Members of the
subcommittee.
I'd first like to thank you for the invitation to appear
before this subcommittee. Issues related to recruitment
radicalization via the internet have pervaded the study of
violent extremism for decades. It's heartening to know that
legislators remain cognizant of these complex processes and are
determined to undermine them when and where possible.
I'd also like to note that any statements I make in my
testimony don't necessarily reflect the position of my
institution or any research centers with which I'm affiliated.
The testimony I provide here is based solely on my 20 years of
findings as a researcher of communication, media
radicalization, and terrorism.
I formally studied the communicative tactics of domestic
left-wing and right-wing extremist groups, foreign terrorist
entities of nearly every political, religious ideology, lone-
actor terrorists, and everything in between. Incidentally, it's
the time that I've spent studying diverse kinds of terrorist
actors that helped me to understand radicalization processes in
a systematic manner.
The first and most difficult fact to confront is that the
processes by which people come to support the use of terrorism
are as varied as the individuals that experience them. As such,
no single hearing can be sufficient to comprehensively describe
these issues, but I'm glad to hear from Chairman Pfluger that
this is only the start of a bipartisan conversation about them.
Nevertheless, after 20 years of observation, interviews,
controlled experiments, and data analysis, I've noticed some
patterns among not only the cases of radicalization of
terrorism I've studied, but also the context in which they have
occurred. It's these commonalities on which I'll focus. But
rather than use my initial time to outline the elements of
radicalization processes as I've come to understand them, I'd
like to offer 4 key themes, lessons learned, and
recommendations in my spoken testimony. Of course, I'm happy to
elaborate on any of these.
First, I feel it's a mistake to treat quote, on-line
radicalization as exclusively distinct from off-line
radicalization. In many cases, individuals who are recruited by
extremist groups undergo radicalization via their behaviors on-
line are often found to be gauging in parallel behaviors off-
line. Moreover, the psychological processes that people undergo
when they radicalize on-line are similar to what they would
experience off-line. Still, there are some technological
features of the internet in general, and social media in
particular, that can affect these processes. While it can be
valuable to be cognizant of these features and their effects,
as well as how they can be leveraged by extremist groups, on-
line and off-line radicalization are heavily intertwined and
should be understood as such.
Second, disinformation perpetuated via social media is a
key driver of beliefs and attitudes that make users susceptible
and vulnerable to recruitment and radicalization by extremist
groups. To a large extent, radicalization to violence via
straightforward propaganda disseminated by a cohesive, well-
defined extremist group is limited, at least in the United
States. Instead, most cases in which an individual radicalizes
the violence seems to be due to a process whereby that
individual engages with content on social media or other
internet-based platforms, both as a function of their own
information seeking and algorithmic suggestions that's
consistent with their beliefs and attitudes. Consistent with
what Daveed said, this often leads to what's been informally
called ``salad bar radicalizations,'' where individuals can
pick and choose from different ideologies.
Third, the revenue models of many social media platforms
are designed such that the platform's executives and
administrators are disincentivized from stemming the flow of
disinformation and other content that might promote
radicalization to violence. Social media platforms are capable
of building models of users that help them know what additional
content will keep them engaged and thereby valuable to
advertisers. As such, users, particularly those who are
interested in politics, tend to remain engaged with this
ideological content that arouses anger and fear about some out
group. This in turn increases polarization and animosity among
those who ideologically disagree. The longer eyeballs stay
fixed on content, the more valuable ad space is, and content
that promotes anger at perceived enemies is very effective at
keeping those eyeballs where they're most lucrative.
Finally, fourth, addressing on-line recruitment
radicalization requires a multipronged approach that's both
reactive and proactive. Research shows that content moderation
effectively stems the flow of disinformation and propaganda on
social media platforms, exposing fewer viewers to them.
However, I would also advocate for digital literacy initiatives
that inform users, especially young users, not only about the
kinds of content they might encounter, but the strategies used
by malicious actors to persuade them. There are several studies
in both the United States and in Europe to show that pre-
bunking or preemptively undermining this kind of content helps
build resilience to its persuasive effects.
With that, I thank you again for the opportunity to appear
before this subcommittee and I look forward to your questions.
Thank you very much.
[The prepared statement of Mr. Braddock follows:]
Prepared Statement of Kurt Braddock
To Chairman Pfluger, Ranking Member Magaziner, and the
distinguished Members of the subcommittee, I would first like to thank
you for the invitation to appear before you. Issues related to
recruitment and radicalization via the internet have pervaded the study
of violent extremism for decades. It is heartening to know that
legislators remain cognizant of these complex processes and determined
to undermine them, when and where possible.
I would also like to note that any claims or statements I make in
my written or spoken testimonies do not necessarily reflect the
position of American University, the School of Communication, or any
research center with which I am affiliated. The testimony I provide
here is based on 20 years of my findings as a researcher of
radicalization and terrorism.
I have studied domestic left-wing and right-wing extremist groups,
foreign terrorist entities of nearly every political or religious
ideology, lone-actor terrorists, and everything in between.
Incidentally, it is the time that I have spent studying diverse kinds
of terrorist actors that has helped me to understand radicalization
processes in a systematic manner. The first, and most difficult fact to
confront is that the social and psychological processes by which
individuals come to support the use of terrorism are as varied as the
individuals that experience them. As such, no single hearing would be
sufficient to comprehensively describe these issues.
Nevertheless, after 20 years of observation, interviews, controlled
experiments, and data analysis, I have noticed some patterns among not
only cases of radicalization and terrorism, but also the contexts in
which they have occurred. It is these commonalities on which I will
focus my testimony. Rather than attempt to offer every detail related
to why radicalization occurs (which can be discussed in some detail
during the hearing itself, should members wish to discuss it), I will
be focusing my testimony along 3 key themes that I feel would be of
interest to the committee:
(1) Psychological radicalization to violence and its facilitation
on-line,
(2) Social media as a communicative mechanism for fostering
radicalization, and
(3) Responding to the threat of on-line recruitment and
radicalization by malicious actors.
1. radicalization to violence: psychological processes
Before delving into the radicalization process as it plays out in
on-line spaces, conceptual disagreements about the nature of
radicalization--among both researchers and security practitioners--
require that we utilize a working definition. For the purposes of this
testimony, I define radicalization as a social and psychological
process by which an individual comes to adopt beliefs and attitudes
that are consistent with an extremist ideology. It is important to note
that radicalization, per se, does not automatically result in violent
behavior on the part of the radicalized. In fact, the vast number of
individuals who undergo radicalization never support or engage in
violent activity. Radicalization of beliefs and attitudes may render a
person a greater risk for engaging in terrorism, but this is by no
means a forgone conclusion.\1\
---------------------------------------------------------------------------
\1\ Ghayda Hassan, Sebastien Brouilette-Alarie, Seraphin Alava,
Divina Frau-Meigs, Lysiane Lavoie, Arber Fetiu, Wynnpaul Varela, et al.
``Exposure to Extremist Online Content Could Lead to Violent
Radicalization: A Systematic Review of Empirical Evidence,''
International Journal of Developmental Science 12 (2018): 71-88.
---------------------------------------------------------------------------
An extended form of radicalization--called radicalization to
violence \2\--depicts this process. Radicalization to violence involves
not only a change in beliefs and attitudes such that they are
consistent with those of violent extremists, but also the added
intention (and possible opportunity) to carry out a violent attack
against civilian targets. To reiterate: not all who undergo
radicalization turn to violence, but it does serve as a risk factor.
---------------------------------------------------------------------------
\2\ John Horgan, The Psychology of Terrorism, 2d ed. (Oxon, UK:
Routledge, 2014).
---------------------------------------------------------------------------
Given this distinction, and under the assumption that the primary
concern of this subcommittee is to prevent violence against American
citizens rather than deplorable (yet perfectly legal) beliefs and
attitudes, my testimony concerns radicalization to violence--that is,
radicalization of behavior--rather than radicalization of beliefs and
attitudes.\3\
---------------------------------------------------------------------------
\3\ See, for example, Clark McCauley and Sophia Moskalenko,
``Understanding Political Radicalization: The Two-Pyramid Model,''
American Psychologist 72(3) (2017), pp. 205-216.
---------------------------------------------------------------------------
Given this, please allow me to turn to psychological mechanisms by
which radicalization can occur. As noted above, radicalization
processes are various, and are a product of several individual-, group-
, and societal-level factors. Still, decades of research on political
violence reveals several social and psychological processes that may be
affected by the messages with which an individual engages. These
processes include (but are not limited to) self-deindividuation, other-
deindividuation, dehumanization, and demonization. Although these
processes are not unique to the on-line domain, the polarizing nature
of social media (see below) can facilitate and catalyze these processes
at scale.
First, self-deindividuation \4\ is a psychological process by which
a person comes to believe that the importance of their identity as a
member of some group has superseded their identity as an individual.
That is, they see themselves as part of something bigger or more
important than themselves, and are therefore willing to make individual
sacrifices to their own well-being (including personal safety) to
support the group of which they are a part. In the context of
radicalization to violence, individuals who self-deindividuate while
engaging with extremists on-line may come to believe themselves to be
part of an important social movement, their membership in which is the
central part of their identity. This may render them more ready to
engage in violence on behalf of that movement.
---------------------------------------------------------------------------
\4\ Originally in Max Taylor, The Terrorist (Lincoln, NE: Potomac
Books, 1988); for a summary of these processes, see Chapter 5 in
Elyamine Settoul and Thierry Balzacq. Radicalization in Theory and
Practice: Understanding Religious Violence in Western Europe. (Ann
Arbor, MI: University of Michigan Press, 2022).
---------------------------------------------------------------------------
Other-deindividuation relates to a process by which an individual
comes to perceive members of some outgroup (e.g., those depicted as
enemies) as lacking individual traits. Instead of perceiving those
people as individual human beings, they are instead perceived as a
homogenous mass of ``them.'' By characterizing an outgroup in this way,
extremists can psychologically prepare their recruitment or
radicalization targets to harm them, should the need arise.
Related to this, dehumanization \5\ is the psychological process by
which an individual comes to perceive members of the outgroup to be
non-human, thereby suggesting they are not worthy of the respect
bestowed upon fellow humans. Often, this is prompted by speakers'
characterization of the outgroup as being animals, vermin, or some
other organism worthy of derision and hate. When an outgroup is
discussed in these terms over time, audiences come to lose their
perceptions of the outgroup's humanity, which likewise facilitates
their willingness to harm them, should the group require it.
---------------------------------------------------------------------------
\5\ See Nour S. Kteily and Alexander P. Landry, ``Dehumanization:
Trends, Insights, and Challenges,'' Trends in Cognitive Sciences 26,
no. 3 (2022): 222-240.
---------------------------------------------------------------------------
Finally, demonization \6\ relates to a psychological process
whereby an individual comes to perceive others as embodying evil. When
an outgroup is characterized as evil, particularly if an individual
believes that their ingroup is charged with defending some
constituency, it becomes easier for that individual to commit violence
against that group.
---------------------------------------------------------------------------
\6\ See Roger Giner-Sorolla, Bernhard Leidner, and Emanuele
Castano, ``Dehumanization, Demonization, and Morality Shifting,'' in
Michael A. Hogg and Danielle L. Blaylock, eds., Extremism and the
Psychology of Uncertainty (Hoboken, NJ: Wiley, 2011), Chapter 10.
---------------------------------------------------------------------------
Self-deindividuation, other-deindividuation, dehumanization, and
demonization are not the only psychological processes that can occur
when an individual undergoes radicalization to violence, but many cases
of terrorist violence indicate that they are relatively common among
the violent.
Given that the purpose of this hearing is to understand recruitment
and radicalization in the on-line space, however, it is useful to
consider the features of the internet generally, and social media
specifically, that facilitate these processes.
2. social media as a mechanism for fostering radicalization to violence
Understanding the psychology of radicalization to violence requires
the understanding that a consideration of ``on-line'' versus
``offline'' radicalization is a false dichotomy.\7\ By distinguishing
processes that occur in an on-line environment from those that occur
away from a computer screen incorrectly suggests that these phenomena
are distinct. In truth, individual trajectories toward terrorism are
often driven by activities that take place over time in both the real-
world and on-line domains.
---------------------------------------------------------------------------
\7\ Joe Whittaker, ``Rethinking Online Radicalization,''
Perspectives on Terrorism 16, no. 4 (2022), 27-40.
---------------------------------------------------------------------------
Still, there are some phenomena that are unique to the on-line
sphere generally (and social media specifically) that lend themselves
to our understanding of how on-line engagement with malicious actors
and problematic content contributes to the radicalization process. In
this section, I outline some of these processes and phenomena, with a
specific focus on social media, how its revenue streams are structured,
and how on-line engagement and the commodification of attention combine
to form the perfect storm for radicalization to violence when malicious
actors engage with vulnerable audiences.
The primary means by which large social media platforms generate
revenue is through advertising.\8\ To generate income, these large
companies (e.g., Meta, Twitter/X) allow advertisers to appear among the
posts to which users are engaged, and when those advertisements appeal
to a user, they may pay closer attention to that ad (measured in clicks
or amount of time viewing the ad) or purchase the product or service
being offered. To maximize the value of an advertisement, the social
media platforms develop models of their users based on previous on-line
engagement, thereby allowing them to promote advertisements that will
be most appealing. In this way, advertising space on social media
platforms is valuable to the degree that the platform can attract views
and engagement from its users.
---------------------------------------------------------------------------
\8\ Amanda Raffoul, Zachary J. Ward, Monique Santoso, Jill R.
Kavanaugh, and Bryn Austin, ``Social Media Platforms Generate Billions
in Dollars in Revenue from U.S. Youth: Findings from a Simulated
Revenue Model,'' PLOS One 18, no. 12 (2023), e0295337.
---------------------------------------------------------------------------
Because attention has been effectively commodified, the social
media platforms are financially incentivized to prioritize and feature
content that is likely to arouse thoughts and emotions that promote
engagement. In many cases, this content takes the form of messaging
with which the user has previously engaged, news that will evoke
engaging feelings like anger \9\ or otherwise subconsciously persuade
the user to keep interacting with the platform.\10\ Moreover, the
algorithms that determine the basis upon which users are recommended
additional content are designed to keep them in ideological echo
chambers in which the messages to which they are exposed grow
increasingly extreme and no dissenting voices can ever be heard.\11\
---------------------------------------------------------------------------
\9\ For example, see Jacquelien van Stekelenburg, ``Radicalization
and Violent Emotions,'' American Political Science Association Politics
Symposium (2017).
\10\ For a discussion related to subconscious advertising, see
Anne-Sophie Bayle-Tourtoulou and Michel Badoc, The Neuro-Consumer:
Adapting Marketing and Communication Strategies for the Subconscious,
Instinctive, and Irrational Consumer's Brain (London: Routledge, 2020).
\11\ Michael Wolfowicz, David Weisburd, and Badi Hasisi,
``Examining the Interactive Effects of the Filter Bubble and the Echo
Chamber on Radicalization,'' Journal of Experimental Criminology 19
(2023): 119-141.
---------------------------------------------------------------------------
When this content is political or ideological--contexts in which
disinformation is abound--users can develop increasingly extreme
beliefs and attitudes about the use of violence against perceived
enemies on the basis of false perceptions and imagined grievances. In
this way, the cultivation of echo chambers in the on-line space due to
the revenue structures of social media platforms builds to a ``perfect
storm'' of engagement, isolation, and anger that can lead to the
aforementioned psychological processes (i.e., deindividuation,
dehumanization, demonization), thereby increasing risk for
radicalization to violence.
3. responding to the threat of on-line recruitment and radicalization
to violence
Although the on-line space serves to facilitate several processes
associated with radicalization to violence, there is an abundance of
research on steps that can be taken to mitigate the likelihood that the
on-line space (particularly social media) can be leveraged by
extremists to recruit and radicalize target audiences.
First, several studies have demonstrated that the responsible
moderation of some content, primarily in the form of content takedowns
and user bands, can reduce the impact of malicious content.\12\ These
studies have collectively demonstrated that when social media platforms
work with experts in extremist messaging, media, and psychology to
identify content that poses a risk for audience radicalization, the
content is not distributed as widely and the malicious actors are less
likely to reach their target audiences. That said, the prevalence of
malicious content on-line suggests that exclusive reliance on content
moderation would not be sufficient for reducing the efficacy of
recruitment and radicalization efforts. Instead, moderation should be
considered only a tool in the overall toolkit of platform
administrators.
---------------------------------------------------------------------------
\12\ For a current review of these practices, see latest articles
in Studies in Conflict and Terrorism, especially Maura Conway and
Stuart Macdonald, ``Introduction to the Special Issue: The
Practicalities and Complexities of (Regulating) Online Terrorist
Content Moderation,'' Studies in Conflict & Terrorism (2025, on-line).
See also Heather Wolbers, Christopher Dowling, Timothy Cubitt, and
Chante Kuhn, ``Understanding and Preventing Internet-Facilitated
Radicalisation,'' Australian Institute of Criminology: Trends and
Issues in Crime and Criminal Justice, no. 673.
---------------------------------------------------------------------------
Rather than rely solely on a reactive approach like content
moderation, there is also research to suggest that prophylactic
strategies that seek to increase audience resistance to extremist
content would be particularly effective. Specifically, media literacy
initiatives \13\ designed to teach audiences--particularly young
audiences--about how malicious actors may develop content designed to
lead them to violence could be particularly useful. Given the
increasingly young age at which many children are becoming digitally
literate, it would behoove interested parties to consider media
literacy campaigns as early as is reasonable.
---------------------------------------------------------------------------
\13\ For an early synopsis, see Jan-Jaap van Eerten, Bertjan
Doosje, Elly Konjin, Beatrice de Graaf, and Marielle de Goede,
Developing a Social Media Response to Radicalization: The Role of
Counter-Narratives in Prevention of Radicalization and De-
Radicalization (Amsterdam, NL: Colophon), 108.
---------------------------------------------------------------------------
Finally, there exists a specific counter-persuasion strategy in
which users are exposed to weakened versions of the extremist messages
to which they will later be exposed when they are on-line.\14\ Several
decades of research have shown that when audiences are told about the
content they will encounter (or the strategies that malicious actors
may use to distribute that content), and are provided with counter-
arguments against it, they are substantially less likely to be
persuaded by it.\15\ This approach would be particularly fruitful for
dealing with propaganda and disinformation produce with generative
artificial intelligence,\16\ which can be particularly difficult to
identify and resist without proper training.
---------------------------------------------------------------------------
\14\ Josh Compton and Kurt Braddock, ``Inoculation Theory and
Conspiracy, Radicalization, and Violent Extremism,'' in Sergei A.
Samoilenko and Solon Simmons, eds., The Handbook of Social and
Political Conflict (Hoboken, NJ: John Wiley and Sons, 2025).
\15\ For experimental evidence in the realm of violent extremism,
see Kurt Braddock, ``Vaccinating against Hate: Using Attitudinal
Inoculation to Confer Resistance to Persuasion by Extremist
Propaganda,'' Terrorism and Political Violence 34, no. 2 (2022): 240-
262.
\16\ Stephane J. Beale and Lewys Brace, ``AI Extremism:
Technologies, Tactics, Actors,'' VOX-Pol Report (2024). https://
dial.uclouvain.be/pr/boreal/object/boreal:291289. which can be
particularly difficult to identify and resist without proper training.
---------------------------------------------------------------------------
The written testimony I have provided above represents only a small
drop in the bucket of our collective knowledge related to the on-line
sphere, radicalization to violence, and the increasingly complex ways
that extremist groups are targeting vulnerable audiences. However, as
extremists develop increasingly sophisticated methods for recruiting
and radicalizing audiences to violence, so too must we develop
increasingly sophisticated methods for undermining them.
Note that additional scholastic references for the concepts
described above are available upon request. I would also be happy to
provide the committee with additional information concerning
radicalization to violence more generally and the role of the internet
in facilitating it.
Mr. Pfluger. Thank you all for your opening statements.
Members will be recognized in order of seniority. We will
also have Members that are cycling in and out. An additional
round of questioning may be called after all Members have been
recognized, and I will now recognize myself for 5 minutes of
questioning.
Dr. Gartenstein-Ross, I mean, the big question here is
basically the First Amendment and how we prevent attacks. How
we prevent, you know, these terrorist attacks from happening,
as I have heard your testimonies and read your written
statements. So how do we differentiate between the
Constitutionally-protected speech and then also the speech that
may advocate for violence and terrorism? How do we protect the
First Amendment and thwart that violence?
Mr. Gartenstein-Ross. Thank you, Chairman Pfluger. That's a
great question. One case that I recommend people to be familiar
with is the Sami Al-Hussayen case. It occurred in Idaho shortly
after the 9/11 attacks. He was prosecuted for terrorist
propaganda. Ultimately, the jury found him not guilty based on
First Amendment grounds. After that case, there have been very
few just propaganda prosecutions.
I think Ranking Member Magaziner put it right when he
talked about how there's no right to incite people to attack,
there's no right to recruit people to terrorist groups. Foreign
terrorists don't have a right. But I think when you get into
that kind of gray area, I expect to see very few prosecutions.
Instead, it's when there's incitement to imminent lawlessness,
which is the First Amendment standard.
Mr. Pfluger. Thank you. Dr. Zelin, you mentioned something
in your spoken testimony about Jabbar and his posting. So I
want to follow up on this question. Should we have known? Could
we have known what could have been done? Did his speech
actually violate the protected speech or was he still in that
protected speech place?
Mr. Zelin. Thank you, Chairman. It's definitely a
complicated issue, but based off of Facebook's own policies,
through their dangerous organizations unit within Facebook,
they have their own sort of FTO list. If you provide support to
that group, which this individual did by pledging allegiance to
the caliph, that would be considered contravening, you know,
their own policies. Again, these are private companies. This
isn't related to speech about politics per se, but within the
context of a private company that sets its own policies. Of
course, there isn't such a thing as necessarily 100 percent
security or successes in these issues. But if there are more
proactive measures with more people involved in trying to
follow these issues, I think it's definitely a possibility that
something could be stopped, for sure.
Mr. Pfluger. You talked about content moderation, and is it
your belief that is done at the platform level?
Mr. Zelin. Yes, I mean, of course, Congress could
potentially regulate to make sure that they actually follow a
certain way of doing it, but there have been methodologies that
have been laid out by the Global Internet Forum to counter
terrorism, which is essentially a consortium of key American
tech companies such as Google, X, Facebook, Microsoft, et
cetera, in terms of how they find the digital footprint or
fingerprint of these issues, to take this content down. But the
terrorists, of course, are evolving themselves as well. So it's
key to try and stay on top of it, because as we take content
down or ways of taking content, they try and then go around it.
For example, instead of saying, like, ``the Islamic State'' or
whatever, they'll put like dots in between each letter or an
underscore so that it trips up the content moderation, at least
in terms of, you know, written types of stuff.
Mr. Pfluger. Thank you. Mr. Flesch, you talked about George
Mason University. I was made aware that there was an incident
on Georgetown's campus where a convicted terrorist named Ribhi
Karajah, who was found to be a member of the PFLP, was invited
to speak and allowed to speak at, you know, a student-led
organization or on campus there. So you have got somebody that
was involved in a bombing in Israel that was part of the PLO
and comes onto one of the campuses.
Just what needs to happen here? What is the response, the
university's responsibility and then, conversely, what is the
responsibility of either Federal or State and local law
enforcement?
Mr. Flesch. Certainly. Thank you, Mr. Chairman. I think the
first thing is to recognize that while we all talk a lot about
Hamas, I focus on them in my testimony. The fact is there are a
number of terrorist organizations in the, I believe,
Palestinian population that has presence here. PFLP, the
Popular Front for Liberation of Palestine, goes back 50-plus
years of all the other U.S.-designated terrorist organizations.
So recognizing that the threat is not just for host Hamas,
Hezbollah, these other big names we hear about, there's a
number of other organizations that also are designated U.S.
terrorist organizations. So that's No. 1.
No. 2 is reaching out to and ensuring that student groups
on campus are aware of this and those that do invite members of
these organizations, those have been convicted for terrorism
and this individual was involved in the killing of a 19-year-
old Israeli a few years ago, that they should not be involved--
invited to campus whatsoever. I think from a Federal
perspective you just have to start looking into the
universities that provide for these student groups, provide for
these organizations, and if they give platforms to terrorist-
supporting individuals or terrorists themselves, they should,
you know, withdraw Federal funding from them.
Mr. Pfluger. I think it was shameful that the university
allowed this individual to even announce such of a meeting.
So my time has expired. I now recognize the Ranking Member.
Mr. Magaziner. Thank you. I would like to start by focusing
on the responsibility of the social media and messaging
platforms or what their responsibility should be. I think
starting with the framework that we have laid out that,
hopefully, everyone here agrees with, which is that there is a
First Amendment right to have different political views, even
extreme political views. But there is not a First Amendment
right to plot violence or incite violence. Like if we can agree
with that as a premise, when you have speech that is inciting
violence, plotting violence, what does effective content
moderation look like and how can we ensure some level of
standardization across the platforms to make sure they're
actually doing it? You know, we can start here with Dr. Zelin
and then Dr. Braddock or anyone else.
Mr. Zelin. Thank you, Ranking Member Magaziner. In terms of
ways of doing this, there have been a few methodologies that
have tried out. One is related to having a digital fingerprint
for different textual, audio, pictorial, and video content, so
that if that particular piece of content shows up on any
particular platform, it'll be wiped out right away. Going
further than that, there's also a way of going after IP
addresses of people who have posted this type of content to
make sure that they're not allowed to create secondary accounts
because this is one of the things that has happened before.
The bigger question, though, beyond just content in English
language, is that other languages have lagged far behind on
these fronts when we're talking about issues like ISIS, where
Arabic and English has gotten better, but when we get into
Africa, Central Asia, and elsewhere, it's way worse. There's a
free-for-all in some ways.
Mr. Magaziner. Thank you. Dr. Braddock.
Mr. Braddock. I would just echo what Dr. Zelin just said
about the moderation of content and automating it, looking for
visual content, looking for written content. But something that
I think that we haven't missed it, but something that can be
emphasized is it's not just content that's taken down. Dr.
Zelin mentioned the banning of IP addresses, but banning users
who spread this kind of information, I think that needs to be
part of the tool kit as well because there's research that
shows that as these users are banned from certain platforms,
their reach grows smaller and smaller. That's what we want from
these mass propagandists.
Mr. Magaziner. So this is a helpful use of sort-of best
practices. But when we think about what the role of Congress
should be, I mean, you know, I personally feel that Congress
does have a role to ensure that a baseline of best practices
are followed across all platforms. Again, to me, this is not a
First Amendment issue because we are not saying that we are
moderating political beliefs or anything like that. It is
specifically incitement of violence. There is precedent here
because Congress has essentially banned nonconsensual
pornography and other types of content. There is no successful
First Amendment challenge to that.
So, you know, again, what would Congressional action look
like potentially in order to ensure that there is at least a
baseline of appropriate moderation occurring across all
platforms? Because I agree some platforms are doing it, they
are doing it well, doing it somewhat well, others aren't even
trying. So what is our role here?
Mr. Zelin. I mean, one way would be to potentially regulate
to make sure that all platforms that work in the United States
follow certain types of, you know, methodologies, I guess you
can say, so that when certain companies change, who's in charge
of them or, you know, policies change for whatever reason, that
there's a certain standard that everybody lives up to, to make
sure that violent extremists, terrorists, what have you are not
able to take advantage.
Mr. Magaziner. Yes. We have been talking a lot about social
media platforms and I think we all, you know, we think about
TikTok and X and Facebook, et cetera. But my understanding is
that on-line gaming platforms are becoming ripe ground for
radicalization as well. Sort-of the messaging and communication
features on gaming platforms, people are recruited and sort-of
groomed sometimes for extremism. Would anyone like to comment
on gaming platforms specifically and what you are seeing?
Mr. Braddock. I can comment briefly. My understanding is it
started really, in some of these games, it starts in the
interactive chat where people are spreading epithets or trying
to recruit almost passively. But there are certain--I mean
there's social media gaming-type platforms, like Discord, where
people can meet and talk and plan and exchange different kinds
of pictures, text, anything they want in the same way they
could on another social media platform. So it's almost like
traditional social media and gaming social media, so to speak.
There's this intersection where they're exploiting the freedom
they have there.
Mr. Magaziner. Yes. I am low on time, but real quick, pre-
bunking, can you just explain a little bit more what that is
and so forth?
Mr. Braddock. Yes. So there's a lot of research out of
Cambridge that shows that when you address a person, if we know
they're going to be persuaded by a certain kind of content, if
we make them aware of that content before they're exposed to it
and let them know that they're susceptible to persuasion,
people, especially in the United States and the United Kingdom,
they don't like knowing that they're susceptible to persuasion.
So they put their defenses up and they're more likely to
recognize it when they encounter it. The research coming out of
Cambridge has shown that you can gamify this, you can turn it
into a game. People get very, very good at recognizing when
malicious actors, whether they be foreign or domestic, are
trying to persuade them.
Mr. Magaziner. Thank you. That is helpful.
I yield back.
Mr. Pfluger. The gentleman's time has expired.
The Chair now recognizes the gentlelady from Georgia, Ms.
Greene.
Ms. Greene. Thank you, Mr. Chairman. Thank you for being
here today, our witnesses.
As we know, the politically-weaponized FBI and DOJ under
the Biden administration went after peaceful pro-lifers, J
Sixers, and Catholics attending Latin Mass, parents trying to
hold their school boards accountable, and many more
conservatives. They refused to prosecute heinous sex crimes
against children and actual terrorists who wanted death to
America and chanted it in our streets.
In October 2020, 3 rioters from the Jewish Voice for Peace
and IfNotNow, both violent pro-Islamic groups that support
Hamas, invaded the Capitol complex. As a matter of fact, they
came in this building right here and they took over the Cannon
Rotunda. Their actions, they completely shut down the
elevators, the stairwells, the hallways, they blocked exits,
and the police officers were assaulted. They took over and
stopped hearings and they stopped Congress, same way January
6th did, although they were not prosecuted and they were not
held accountable the same way protesters on January 6th were.
Now, I bring this up today because the riot organizers that
held this riot here in the Cannon Building not only
communicated through a chat that appears to be Signal, similar
to the ones--similar to different messaging programs you have
spoken about, but the title of the chat was Global Intifada. So
if you see--it is hard for you to see, I know from here, but
this picture right here is a screenshot of one of their cell
phones as they were communicating with each other. You can see
in the background their signs. It was taken over in the Cannon
Rotunda, but right here it says Global Intifada. They were an
organized group and they had organized this pro-Hamas, pro-
terrorist riot here in the Cannon Building.
Mr. Flesch, in your testimony, you talk about globalizing
the intifada. What is the goal for globalizing intifada not
only world-wide, but in the streets of America?
Mr. Flesch. Yes, ma'am. Well, the goal of people calling to
globalizing intifada is to bring a Palestinian-inspired
strategy of violent political revolution targeting civilians,
including children, and essentially to overthrow the United
States and to defeat the fabric of society of the United
States.
Ms. Greene. So their stated goal is to overthrow the United
States. They organized and they call themselves Global Intifada
and they came in this building in 2023. That is shocking.
If you look at the names of the members in the group chat,
you will find the lead attorney for the Southern Poverty Law
Center, the SPLC, Katrina Bleckley, right here. Katrina
Bleckley is part of the Global Intifada, lead attorney for the
Southern Poverty Law Center. This is the same law center that
provided nearly 40 pages of testimony to the same January 6th
committee that was prosecuting people for protesting the 2020
election. This is also the same law center that was cited in
the FBI memo that labeled traditional Catholics as radical
extremists. This is the same law center which has nearly $770
million in assets, by the way, and was given over $8,000 by the
Department of State in 2021 for speaking engagements with Lecia
Brooks.
So we are here talking about on-line radicalization, but we
are also through this, this right here is proof for every
single American that this radicalization that is happening on-
line is radicalization that has happened right here in our
Government that stopped Congress.
I would also like to highlight the threat of on-line
radicalization and how the threat within our own homeland has
been exasperated by more than 12 million illegals the Democrats
allowed to pour into our communities under Biden, with 2
million of them being known gotaways who we have no clue where
they are from, where they are now, and what their motives are
inside our homeland.
Dr. Gartenstein-Ross, can you briefly outline, I know I am
out of time here, the key methods foreign terrorists groups use
to radicalize individuals on-line to not only join their ranks,
but carry out attacks?
Thank you, Mr. Chairman.
Mr. Gartenstein-Ross. Thank you.
Mr. Pfluger. You have about 30 seconds.
Mr. Gartenstein-Ross. Sure. To do it very quickly, as Dr.
Braddock said, there often is not a great gulf between what's
done on-line and off-line. On-line, generally speaking, methods
include looking at people who seem susceptible to propaganda,
forging a relationship with them. They sort-of act as a social
movement in that way, getting to know people who they want to
recruit. I've written about what's called the virtual plotter
model that was--that ISIS used a lot, and I believe still does
in a number of different ways, which uses encryption often to
be with attackers right up until the moment that they attack.
Thank you.
Mr. Pfluger. The gentlelady's time has expired.
The Chair now recognizes the gentleman from California, Mr.
Correa.
Mr. Correa. Thank you, Mr. Chairman.
I think one thing that unites all of us across the aisle is
our interest and our goal of protecting our citizens from
terrorist attacks, domestic or foreign terrorists as it may be.
That job, I would say, is like finding a needle in a haystack.
You need intel, you need to work with our foreign partners.
That is why we have FISA and our ability to gather intelligence
to figure it out.
You know, General Kelly, former Secretary of Homeland
Security, used to say that border security does not end or
begin at the border, but it is rather a compilation of a lot of
work, a lot of intel work with our partners across the seas and
other places. At the end of the day, you have U.S. Government
agencies that put this stuff together and decide, try to
decide, try to anticipate these attacks. FBI, CIA, other
agencies, along with State and locals that try to figure this
out. I would say a lot of these agents are patriots. They work
Government where they could be making a whole lot more money
private sector, but they are working for the Government because
they believe in the mission.
I am concerned over these firings that I see FBI and CIA
agents. What is it going to do for morale for those who are
left? What is it going to do for those that just got fired?
There is a story that Chinese and Russian operatives are trying
to recruit some of these fired individuals to work for them.
But at the end of the day, I want to ask you, gentlemen, each
of you, these firings, is this good for our national security?
Mr. Gartenstein-Ross, is this good or bad morale? You are
losing a whole lot of intelligence experience. Please, opine.
Mr. Gartenstein-Ross. Sir, thank you for the question. I'd
be very happy to talk to you about this. I'm not here to
testify on it and the reason specifically----
Mr. Correa. Thank you very much. Dr. Zelin.
Mr. Zelin. It's hard to know what specifically is being cut
with a bit more transparency, but there's definitely a case to
be made that less people focusing on these issues could lead to
greater risks in the future.
Mr. Correa. Look, you fire folks, even though that person
fired may be somebody who is new 2 years, 5 years, I mean, they
are going to look, the person that is left behind working on
this stuff is saying, what is it going to do for your morale? I
might be the next one. Do you think that would possibly affect
your effectiveness defending the homeland?
Mr. Zelin. It's definitely plausible. Obviously, what's
going on now is very new, so we don't have a test case to know
the specifics of what could happen. But less people focusing on
these threats on-line definitely could mean less eyes on
something that something then passes through.
Mr. Correa. Let me say that these folks have to have
defense. You can't score one point on us. They cannot. That
means American lives.
Mr. Flesch, what do you think? Firing good, bad, morale?
Mr. Flesch. I would echo what my colleagues have said, that
it's a little early to tell the effect of this policy, but
certainly I'm happy to have a conversation after the fact, but
focusing on the radicalizing element is why I've been brought
here to talk.
Mr. Correa. Well, I certainly hope that we don't get to the
day when we say it was a little bit too late. Not too early,
but not too late.
Mr. Braddock, thoughts?
Mr. Braddock. I'll say bad, but not so much--I agree with
Dr. Zelin about the fact that it's an issue of the number of
personnel, the number of eyes you have on this content and the
users. I also think it's a issue of institutional knowledge.
There are ways the terrorists communicate on-line that take a
long time to understand. Give you an example. I recently saw a
study that are showing that some foreign terrorist
organizations, they communicate via exclusively emojis. They
just use emojis and they're coded in different ways. Now, if
somebody has that institutional knowledge related to the use of
emojis----
Mr. Correa. So experience matters.
Mr. Braddock. Experience matters.
Mr. Correa. Firing people may not be the wisest thing to
do. I am not trying to get political here. I am just saying
that all of us have a vested interest in protecting the
homeland. That is why Homeland Security was created after 9/11,
when we had holes in the system, when we had silos that didn't
talk to each other. Nine-eleven could have been prevented
possibly, but that is why we are here. I am concerned that the
firing of these individuals will hurt us defending the homeland
in the long term.
Thank you gentlemen, for your testimony. Mr. Chairman, I'm
out of time. I yield.
Mr. Pfluger. The gentleman yields.
The Chair now recognizes the gentleman from Texas, Mr.
Luttrell.
Mr. Luttrell. Thank you, Mr. Chairman.
Good afternoon, gentlemen. I represent a small portion of
Harris County, Houston, Texas, and this pains me to say it is,
but Houston, Texas, is the worst city in the country for sex
trafficking. I have had multiple discussions, meetings, various
agencies, law enforcement, up and down the elected leadership,
not only in Harris County, but the State. How do you solve a--I
sit on the I-10/45 corridor. How do you solve a problem where
these--and just to be clear, cartels are now known as terrorist
organizations. So I am going to throw this on you guys. The
conversations that I have with these folks is like, how do we
solve this problem? How do you cut the head off the snake? How
do you get rid of these bad actors to save our babies? Because
they are going after our children. It always circles back to
the metaverse. It lives and breathes on-line. That is how they
communicate. That is how they recruit, and that is how they set
the hook. Then our children are moved into this horrible space
and become victims.
I want to hear from all 4 of you. How do we--and I think it
was Mr. Flesch who said regulate. How do we regulate the
system? I am sorry, might have been you, Doctor, that said we
regulate the space, right, regulate the system, which I am not
speaking for my colleagues, but I am sure they would absolutely
get on board with the fact that this is a problem we shouldn't
have when it comes to the protection of our babies. Right?
How do we fix this? If we regulate it, it is going to have
to come from the subject-matter experts, which would be you all
or those that control either the social media platforms, the
gaming systems. Because the best thing I can do for my children
is inform them of what happens out there. But when I talk to
the parents whose children have been taken, they never saw it
coming.
Dr. Ross, I will start with you, sir.
Mr. Gartenstein-Ross. Thank you, sir. It's a great
question, and I'll take it from the cartel's angle first and
foremost. That's something that I've done a fair amount of work
on in some of my projects for U.S. Customs and Border
Protection.
I'd move it a little bit away from the on-line space. The
way we've looked at cartels in our work for CBP is how do you
shift their incentives? Cartels are many things, but among
them, they're a business. I would look at the cartel's center
of gravity. I think that right now we're playing defense vis-a-
vis the cartels with respect to trying to stop them at the
borders. One way I would look at it is there's a number of
different vulnerabilities that they have. How do you shift
those incentives? Through causing some pain to them, but doing
so in a way that isn't too costly for us at the same time. I
think that that's very important because the flow of human
trafficking----
Mr. Luttrell. Where would that--great. I mean, that's--OK,
you are opening up--I don't know if I am going to say you open
up Pandora's box there. But if we are going to move them away
from sex trafficking, drug trafficking, and all that illegal
nefarious acts, we are not going to make them a legitimate
company. I mean, where do we push them?
Mr. Gartenstein-Ross. Yes, I don't think that trying to
make them a legitimate company is the way to go.
Mr. Luttrell. Yes.
Mr. Gartenstein-Ross. Obviously, there are some who argue
for that, but I think that there are different ways that we can
undertake points of pain vis-a-vis cartels. Obviously, there's
a whole conversation around cartels right now, designations,
possible use of military force, and the like. One thing that we
know is that cartels are fighting against two sets of enemies
in most places they operate. They're fighting, No. 1, against
the Mexican government, both Federal and State. No. 2, they're
also fighting against other cartels as well.
Oftentimes, you know, picking your enemy and leaving other
enemies alone can have a very serious effect on them. Anyway, I
don't want to advocate for something specific other than say
that I think there's a different way to look at it, that we
tend to look at defense at the border. I think there's a
secondary way to look at it, which is cartels have incentives,
they have vulnerabilities. If we look at the question as, how
is it that rather than doing what you could call Operation Same
the same, meaning we have the same operations over and over
mobilizing to the border, how is it that you shift their
incentives in a way that will have a lasting effect? I think
that it's a complex answer, but that's the way I would
formulate the question to hopefully get something that is more
lasting than what we have now.
Mr. Luttrell. Dr. Zelin, you got 21 seconds, sir. How we
doing?
Mr. Zelin. Thank you for the question, but it's not a
subject I personally focus on, so I don't have as much
experience on it.
Mr. Luttrell. Mr. Flesch, go ahead.
Mr. Flesch. I'd echo my colleague.
Mr. Luttrell. Then we can't run. Hey, hey, hey, hey. You
need to hear me say this. Don't run away from this. This is the
worst, worst thing you can possibly imagine when you see these
babies. OK? If you are not studying it and I'm not going--I
can't sit here on this dais and tell you to do that. OK?
Mr. Braddock. I'll take 10 seconds because I know we're out
of time. But----
Mr. Luttrell. You good?
Mr. Braddock. Am I good with that?
Mr. Luttrell. Yes.
Mr. Braddock. Yes. We had a child abducted near my home
town as well a couple of years ago, assumed to have been sold
into sex trafficking as well, so it's near and dear to me as
well.
My main point would be we've been focusing on content
moderation, the responsibility of the platforms to focus on
content. I also think that there can be more, the platforms can
do more to regulate what's going on in private messaging on
those platforms between people. Because that's a lot of the
ways that these individuals are reaching out to children is
they're engaging with them in these private messages on
Facebook, on Instagram, on whatever platform you choose. When
those personal relationships develop, that's how kids are
groomed and they're taken.
Mr. Luttrell. Thank you.
Mr. Braddock. So I want you to focus on that as well.
Mr. Luttrell. Thank you, Mr. Chairman.
Mr. Pfluger. The gentleman's time has expired.
The Chair now recognizes the gentleman from New York, Mr.
Goldman.
Mr. Goldman. Thank you very much, Mr. Chairman. I also want
to echo my support for having this hearing, especially because
I tried futilely and desperately last Congress to have a
hearing on a similar topic. In November 2023, I wrote a letter,
led a letter to Elon Musk on X, because in the 6 weeks
following October 7th attack that Mr. Flesch so powerfully
referenced, there was massive distribution of Hamas and other
terrorist propaganda videos and even some generated profit on
X. I got no response, written response. I guess that part of
the 75 percent cut in Twitter that Elon Musk did went to the
Government Affairs Office.
So then, in March, I wrote to Chairman Comer. I was on the
Oversight Committee asking for a hearing on this because it
continued. In fact, the secretary general of Hezbollah, Hassan
Nasrallah, actually wrote a tweet complaining about having his
blue checkmark taken down. So that was an issue that also
related to violating sanctions. No word from Chairman Comer.
Then we wrote a letter to the Speaker in August. No
response, no hearing. So I am glad we are having this and I
take the gesture of bipartisanship seriously.
Dr. Zelin, I want to come to you first because I noticed in
your opening statement, Dr. Braddock, you said the similar
things, that a lot of these problems have been exacerbated by a
reduction in content moderation. In fact, you said beginning
with Elon Musk's purchase of Twitter, the level of content
moderation in general has backslid. Of course, recently Mark
Zuckerberg announced that Meta was eliminating all content
moderation.
Certainly we are grappling with the First Amendment, but
there is no possible way to stop the on-line proliferation of
terrorism without some degree of content moderation. Is that
right, Dr. Zelin?
Mr. Zelin. Yes, thank you for the question. I agree. I
mean, obviously we have our own internal debates within the
United States, partisan debates, what have you. But when we're
talking about foreign terrorist organizations, we need to be
clear that they should not have a space on these platforms,
especially American-owned companies. The current state of
things has definitely backslid in recent years.
Mr. Goldman. Part of the reason that they should not is
that they are professional criminal organizations, to put it a
different way. Is that right?
Mr. Zelin. Yes. I mean, they do--you know, they're doing
things illegally according to the U.S. law for sure.
Mr. Goldman. So we don't want crimes to be committed on-
line on our American companies, right?
Mr. Zelin. That's correct.
Mr. Goldman. So this is where we run into a problem because
some believe that the free speech is an absolute right. You can
say anything you want to say. Of course, that would include
inciting violence, potentially committing crime, and also sex
trafficking, as my colleague aptly pointed out. So this is not
an academic exercise in discussing the First Amendment. This
has real-world significance that you all have identified.
Mr. Flesch, when I was walking over here after we just
voted, outside the Capitol, I saw a huge poster with a swastika
on it. I know you have spoken a lot about the rise of
antisemitism, anti-Israel sentiment post-October 7th, and I
agree with you about that and I condemn it wholeheartedly. But
we are also seeing a rise of neo-Nazi attitudes on-line. In
fact, some of them are truly perpetuated by the owner of X, who
has expressed support for AfD in Germany, which, I think you
will agree with me, is effectively a neo-Nazi organization.
I guess I would just ask you, you have addressed a lot of
the rise in the global intifada and I agree with you, but talk
about the risk of the on-line support and proliferation of
nuclear neo-Nazi views.
Mr. Flesch. Well, I think certainly neo-Nazis are
antisemites. They certainly present a risk to everyday
Americans. A lot of my focus has been on the current immediate
threat that the--and for short-hand, this Hamas supporting
network that we've been looking at Heritage presents to
Americans of all persuasions, not just Jews. So neo-Nazis, you
know, will certainly be a threat for--maybe endemic, but
they're not, as we estimate, the immediate threat that a lot of
these individuals I've identified in my testimony we see almost
every day who are attacking Jewish and non-Jewish students and
Americans across the country.
Mr. Goldman. Yes. Well, I mean, we can debate which is
worse. Certainly the white supremacist theory affects Jews as
well as non-Jews as well.
So I see my time is up. But I appreciate all of you for
being here and I look forward to addressing this issue
together.
Mr. Pfluger. The gentleman's time has expired.
The Chair now recognizes the gentleman from Arizona, Mr.
Crane.
Mr. Crane. Thank you all for coming to testify before the
Homeland Security, Counterterrorism and Intelligence
Subcommittee. The threat picture Americans face is constantly
changing, and with the development of AI and the use of the
internet, any small group can have a relatively high impact on
any given population.
Dr. Aaron Zelin, I know you have written about on-line
mobilization of foreign fighters. To counter that, the United
States has taken some steps through blacklisting some terror
groups from posting on-line. According to Adam Hadley, the
executive director of Tech Against Terrorism, who gave an
interview to Wired magazine, big tech platforms have worked
hard to create databases of known violent extremist content,
known as hashing databases, which are shared across platforms
to quickly and automatically remove such content from the
internet. Are any of the witnesses familiar with the industry
term ``hashing sites''?
Mr. Zelin. Yes, I am.
Mr. Crane. Or groups to create a censorship database to
prevent bad actors from spreading misinformation from terror
groups? My question is this, how would you recommend the
committee evaluate the balance between freedom of speech and
the need to protect the public from terror groups using AI to
conduct deepfakes?
Mr. Gartenstein-Ross. That's a good question. Thank you for
raising it. I would judge it in 2 ways.
The first one is whether the deepfake poses an immediate
threat. In some cases, deepfakes can be created, you know,
designed specifically to incite. For example, we've seen deadly
riots overseas based on Koran desecration. Something like that
would have a purpose of trying to incite. In cases like that, I
think that legislative efforts are important.
There are other areas that were referenced before, such as
deepfake pornography and the like, where I think, again, there
is a strong impetus for the legislature to step in in order to
stop harm for individuals. That's essentially where I would say
the line is drawn with respect to deepfakes.
I think there's other ways to deal with propaganda
deepfakes because ultimately there is a purpose for deepfakes.
Right? Deepfakes can be used for cinema and the like. There's a
reason why people create them. They're not just pure harm, but
harm to the individual, I think is the area where the
legislature can come in and should be able to have a strong
hand.
Mr. Crane. Thank you, Dr. David Gartenstein-Ross, extremist
groups are becoming more tech-savvy. Former Google Chief
Executive Officer Eric Schmidt highlighted his worry during the
February 13, 2025, hearing with the BBC. He said there is
extreme risk posed by AI development that could harm innocent
people. In his opinion, countries like Russia, North Korea, and
Iraq have ulterior motives and could misuse emerging
technologies for their own purposes.
My question, is the correct balance for the tech companies
and government that take into account the need for freedom of
these companies to develop AI and the government's interest in
protecting the public from deepfakes and AI-generated images
that stir panic?
Mr. Gartenstein-Ross. It's another great question. A lot of
these are difficult to answer because of the fact that it's a
balance. But I'd say that this is also understood in light of
an AI competition between U.S. companies and PRC-led companies.
Ultimately, most people looking at it, including Mr. Schmidt,
who you mentioned, believe that the country that's able to win
the AI race is also going to be the global leader. So I think
there's a lot invested. I'd also say, as I mentioned in my
opening statement, that right now, the safeguards that are put
in place, even by very responsible companies, are
extraordinarily easy to jailbreak and to get around.
Mr. Crane. Thank you.
Mr. Gartenstein-Ross. Thank you.
Mr. Crane. Thank you. Mr. Braddock, on March 24, 2023, you
were quoted in Business Insider saying, ``If there is violence
as a result of Trump's words, the former President will hide
under a blanket of plausible deniability, saying that he never
ordered anyone to become violent.'' Did you say that?
Mr. Braddock. Probably.
Mr. Crane. My question is this, you are extremely critical
of Republicans through what you call stochastic terrorism,
inciting terrorism through volatile rhetoric and what you deem
as extremist rhetoric. But when Democrats incite violence, you
seem to be a little bit silent. My colleague who just left, Dan
Goldman, said Trump is dangerous to democracy and that he has
to be eliminated. Did you speak out or criticize Mr. Goldman
when he said that? We couldn't find anything.
Mr. Braddock. I was--never, ever heard that quote.
Mr. Crane. OK. How about this?
Mr. Braddock. I can give you----
Mr. Crane. Hold on.
Mr. Braddock. I can give you examples of when I did speak
out.
Mr. Crane. Nor did you, when Maxine Waters said, tone down
the rhetoric, when she encountered protesters in Minnesota to
get more confrontational, encouraging Democrats to confront
Trump officials if they see them in public, did you say
anything about that one?
Mr. Braddock. I did.
Mr. Crane. You did? Can you provide that for the committee,
please?
Mr. Braddock. During conferences and stuff? Sure.
Mr. Crane. Yes. Can I have one more quick question? How
about this one? President Biden on October 23, 2024, who said
we got to lock him up or that Trump was a threat to democracy
right before numerous assassination attempts. Did you say
anything about that one?
Mr. Braddock. No. That wouldn't meet my threshold.
Mr. Crane. We got to lock him up, that doesn't reach your
threshold?
Mr. Braddock. No.
Mr. Crane. OK. Thank you. I yield back.
Mr. Pfluger. The gentleman's time has expired.
The Chair now recognizes the gentlelady from New Jersey,
Ms. Pou.
Ms. Pou. Thank you, Mr. Chairman and to our Ranking Member
Magaziner, for holding this very important hearing.
You know, while terrorism is by no means a new problem, the
internet gives extremists a broader reach for spreading
dangerous ideas. It also provides for those seeking to commit
violent attacks. I was proud to join my committee colleagues
last month on a visit to New Orleans to see first-hand the
scene of a devastating terrorist attack on New Year's Day. We
can and must do more to prevent tragic attacks like this. Doing
this demand, we address how the internet radicalizes people.
In my own State of New Jersey, authorities arrested a man
in November 2022 for using social media to share a manifesto
containing threats to synagogues. In the manifesto, he claimed
Jewish people were responsible for Muslim hatred. He drew
inspiration from the white supremacist attack at a Black church
in Charleston and stated a desire to attack a gay club.
Mr. Gartenstein-Ross, you have referred to this as a salad
bar extremism, where an attacker lacks a clear ideology or
cites numerous categories for extremist views to justify
violence. How does mixing of extremist ideas impact the ability
to accurately predict extremist behavior?
Mr. Gartenstein-Ross. Thank you. That's a great question.
To clarify my remarks, I refer to it as composite violent
extremism, which rather than salad bar. The reason why is I
think that oftentimes the internet and ideas and grievances
that are in the ether impact the individual much more than they
pick and choose. I think that it has a tremendous effect on
complicating the idea of when someone is radicalizing to pose a
threat.
If they're radicalizing with respect to a traditional
ideology, oftentimes there's a few different models, but to put
it simply, you can understand when they fulfill the threat
criterion that they seem to, you know, have grievance meet
ideological justification, and then you can look to whether
they're also taking preparation. When it's mix-and-match and
grievances and sentiments, I think it's much more complex and
that's kind-of the petri dish that we're all living through
currently.
Ms. Pou. In addition to learning how to prevent on-line
radicalization, sorry about that, I am interested in how we can
prevent people back from dangerous ideas. Again, to you, in
your written testimony, you said about terrorism prevention,
``The existing toolkit remains valuable, but fine-tuning is
necessary to account for the ideological fluidity, digital
dynamics, and the radicalization unique to composite violent
extremists.'' How exactly do we fine-tune and are there
opportunities for on-line counter messaging?
Mr. Gartenstein-Ross. I think there are a lot of
opportunities. Those which I see in the prevention sphere as
being most important are when you look at the individual who
poses a risk of carrying out an act of violence, looking at
whether there are needs that aren't being met. One of the
models, for example, is to look at whether there are mental
health needs, who an individual could be connected to.
Sometimes that provides an opportunity for interventions other
than efforts to take off-line or to use content moderation to
eliminate bad ideas because for as long as human history has
been around, there've been a profusion of bad ideas.
Legislators don't want to stop bad ideas, but what we can try
to stop is that connection to violence.
Ms. Pou. Thank you. Very quickly, Dr. Braddock, could you
please also provide us with your thoughts? I know we have like
30 seconds, but I would love to hear your thoughts on that real
quickly, please.
Mr. Braddock. Is there something you'd like me to focus on
in 30 seconds?
I mean, one of the things that we've been talking about
over the course of this hearing is the importance of content
moderation. The fact that the way that people pull their
ideologies together, the fact that they pull from such
disparate ideologies, makes it very, very difficult. But I
think that also speaks to the fact that one of the things we
mentioned earlier was the need of human eyes on these things.
We want to automate everything. We want to use AI to take care
of everything. But I think in situations like this, which are
so nuanced, I think we do need human eyes to make these sorts
of judgments. I'm not in a place to say which eyes those should
be, but humans can make better judgments about nuanced content
like this better than AI can, at least presently.
Ms. Pou. Thank you. The word ``radicalization'' was
difficult for me to say just moments ago, but thank you, Mr.
Chairman. I yield back.
Mr. Pfluger. The gentlelady's time has expired.
In lieu of a second round, what we will do is give each
side 5 minutes, and I will hold to the 5 minutes. We won't go
overboard. So after this 5-minute period, and I will yield to
my colleagues for their questions, I will let the Ranking
Member also yield to his colleagues if they so choose.
So at this point, we will yield 5 minutes to our side of
the aisle. I will start with Mr. Zelin. Just you mentioned a
couple of things on platforms, and I am kind-of interested on,
you know, how do you go about, on platforms like TikTok, on
platforms like Facebook, what does that training look like to
recognize? We have talked about the human in the loop here and
maybe go into a little bit of depth if, you know, you could
wave that magic wand. What does that look like?
Mr. Zelin. I haven't worked at these companies, so I'm not
quite sure how they specifically train people, but I would
imagine that they have different, you know, lessons related to
understanding specific ideologies, groups, looking at the
propaganda, and bringing in outsiders so that they understand
the context and then build up from there. Obviously, there will
be more senior versus junior people on these issues.
Then besides that, when we're talking outside of the human
element, there's, of course, the digital footprint, which, you
know, is the hashing methodology, but that sometimes has false
positives.
Mr. Pfluger. Thank you. I yield to the gentleman from
Texas.
Mr. Luttrell. Thank you, Mr. Chairman. Dr. Zelin, or it
might have been you, Mr. Flesch, talking about the language
barriers on the social media platforms and how we are
substantially behind. Is that a fair statement to say? Did I
say that correctly?
Mr. Zelin. Yes.
Mr. Luttrell. That is remarkable considering our ability to
aggregate data in 2025. How do we fix this problem? By the way,
which one--the emoji thing? I had not heard that yet. That is
horrible to hear that. But OK, go ahead, sir, sorry.
Mr. Zelin. I mean, part of it is just making sure that
you're hiring people with these language skills that also know
English, too, so they can work within the context of an
American company. Part of the issue, too, though, when you get
in that context is that usually might be working with people
that live in authoritarian regimes, so they might insert
people, also, and, therefore, degrade the possibility that
certain things are done in a correct manner. So it gets more
complicated once you get sort-of out of that Western mold of
things in some ways, even if it's something that's important
when we're talking about jihadis and the fact that they've
proliferated in different areas in the last 5 or 6 years.
Mr. Luttrell. In order to combat that, I mean, AI machine
learning, we got to be somewhere close.
Mr. Zelin. It's definitely something that should be used
and tried, but there are, of course, certain nuances that are
used in different languages as well that we might not even know
about because we're not native speakers. So that's why you also
need the human element, too.
Mr. Luttrell. OK, thank you.
Mr. Zelin. Yield to the gentleman from Arizona.
Mr. Crane. Thank you. I want to come back to you, Mr.
Braddock. Do you believe that President Trump is a stochastic
terrorist?
Mr. Braddock. I believe he has been in the past.
Mr. Crane. Yes? Why is that?
Mr. Braddock. Typically stochastic terrorism, if you want
to offer a definition, is where an individual utters things
that can be perceived by a reasonable person as a promotion of
violence, if not outright incitement.
Mr. Crane. So you think that----
Mr. Braddock. There have been things that he has said that
I think----
Mr. Crane. So you think that President Trump is trying to
incite violence?
Mr. Braddock. Whether he wants to or not is immaterial.
Stochastic terrorists typically don't need to be able to incite
it on purpose or not. It's their words that are important.
Mr. Crane. OK.
Mr. Braddock. If I could, on the record, I could say
stochastic terrorism and stochastic terrorist is a terrible
term. I've actually been trying to get rid of that term as
something that describes this phenomena. I call it malicious
provocation, because that has to relate to----
Mr. Crane. But you have used that word a lot, haven't you?
Mr. Braddock. Yes, it's what we use in the field.
Mr. Crane. OK. Do you think that our colleague, Congressman
Goldman, with the statement I just read you, do you think he's
one as well?
Mr. Braddock. If I could see the context in which he said
it.
Mr. Crane. I already read it to you.
Mr. Braddock. You read the statement, but I need the
context as well. Statements aren't----
Mr. Crane. ``Trump is so dangerous to democracy that he has
to be eliminated.'' I will read it for you again. Do you think
that that is stochastic terrorism?
Mr. Braddock. If you can tell me the context of when he
said it.
Mr. Crane. I can't tell you the context of which he said
it, but those are his words.
Mr. Braddock. Context is important when people say things.
Mr. Crane. Yes. So is partisanship. I think that that is
what we are seeing here today.
Mr. Braddock. Is that right?
Mr. Crane. Yes, exactly. I am going to tell you this right
now. This is why Americans have such a hard time with
censorship coming from guys like you. They call balls and
strikes only on one side of the aisle, and then they won't do
it on the other side.
Mr. Braddock. Would you like me to do it for the other side
of the aisle?
Mr. Crane. Yes, yes, go for it.
Mr. Braddock. Be happy to.
Mr. Crane. Yes. OK. Tell me who on the other side of the
aisle?
Mr. Braddock. We talked about Maxine Waters, which I have
said in public statements to threat analysts that counts as
stochastic terrorism----
Mr. Crane. Yep.
Mr. Braddock [continuing]. Because the context in which she
said it was already heightened aggression. So there's a bigger
chance that people are going to engage in violence.
Mr. Crane. So you don't think that anybody would take Dan
Goldman's statement that Trump is so dangerous to democracy
that he has to be eliminated, and you don't think that that has
the potential to incite violence? Are you serious?
Mr. Braddock. I don't make judgments based on statements I
haven't seen in the context in which they're said. Do you want
another one from the left to make it feel better?
Mr. Crane. Yes, please.
Mr. Braddock. Yes. When Chuck Schumer was on the steps of
the I think it was Supreme Court near Roe v. Wade and a very
energetic crowd and said these justices will reap a whirlwind.
Does it make you feel better?
Mr. Pfluger. The gentlemen's time has expired.
Mr. Crane. Well, it would make me feel better if you could
actually listen to the statement from Goldman and say the same
thing, but----
Mr. Pfluger. The gentleman's time has expired.
Mr. Crane [continuing]. Clearly you can't. All right, I
yield back.
Mr. Pfluger. The Chair now recognizes Mr. Magaziner.
Mr. Magaziner. All right, thank you. I thank the Chairman
again for a, I think, productive, enlightening, and mostly
bipartisan hearing today. I think this was actually a very
productive back-and-forth for the most part. So I thank you
guys, all of you, for your insights. I probably won't use my
full 5 minutes, but there are just two things I wanted to
follow up on real quick.
We had, I think, a very good conversation about where the
line should be in terms of free speech, you know, inciting
violence versus not, and about methods of content moderation
that we should encourage and potentially legislate. One thing
that we haven't delved into is the algorithms. So I wonder if
any of you have any insight into that. Should there be certain
standards or best practices that the social media companies
should adhere to in terms of algorithms that kind-of lead
people deeper down the rabbit hole to more and more extreme
content? Or is it not really a matter of policing the
algorithms, it is more just about the underlying content? I ask
any of you to weigh in.
Mr. Gartenstein-Ross. I would counsel the legislature
against trying to moderate or legislate for algorithms. There's
a lot of problems with algorithms. There have also been
attempts to legislate algorithms before. Specifically, both
Florida and Texas implemented bills known as anti-censorship
bills, which, among other things, would limit the amount of
times that social media companies can change their algorithms,
No. 1, and No. 2, demand algorithmic transparency.
The reason why I personally was against those bills is, No.
1, algorithms change constantly, and I think it's anti-free
enterprise to try to limit the number of times they can change
algorithms. No. 2, if they reveal what's going on with
algorithms and what they're flagging, then every single bad
actor knows exactly what those companies are looking for. So I
think to me right now, I think the safest thing is to stay away
from their algorithms.
Mr. Magaziner. OK. No, that makes sense. That is a helpful
answer.
Last issue, and this is more of a comment than anything
else, the importance of open-source data collection I think is
important. You know, when we talk about Signal's intelligence
and trying to find bad actors who commit acts of violence
before they can commit those acts, you know, we often are
thinking about covert methods. But as I understand it, with the
attacker in New Orleans and many others, I mean, oftentimes
people are posting stuff openly and in an environment where the
social media companies are not effectively regulated and some
aren't removing that content or not doing it in a timely way.
It's important to have, you know, Department of Homeland
Security I&A office, which does open-source collection.
National Counterterrorism Center has an open-source desk. So I
just want to make sure that we are emphasizing the importance
of those collection efforts as well when we are making
budgetary decisions and the like, because we need to be putting
more resources toward that, I think, not less.
You know, with the time I have remaining, I just ask any of
you, is there something that you were dying to say that we
didn't get to? If you want to take a final minute here. I see,
Dr. Zelin, go ahead.
Mr. Zelin. I'll just add to your comment that I think, you
know, working together amongst Government agencies is
important, but I also think Government to tech companies
important and between tech companies because I think all of us
in this room know that we all use more than just one social
media application or other, you know technology in our phones
or what have you all the time. So it's, you know, just as we
do, so do the terrorists in some ways. These are commercial
usages. So there needs to--even if when we're talking about the
digital footprint type of issue, that's one aspect of it. But
there also needs to be greater coordination.
One of the issues where we get to is when it's companies
outside of sort-of the American jurisdiction or they're
companies that are in adversarial countries, like TikTok with
China or DeepSeek and things along those lines, that makes it
that much more complicated. That's why, you know, talking about
bans and things like that is something that I think should
continue to be on the table.
Then one more point related to the algorithms, while I
agree with Daveed that it's definitely complicated from a legal
perspective, we should also recognize that if you follow
somebody on whichever platform, they'll start feeding you
recommendations. If you start following people that are
terrorists or extremists, they'll feed you people within that
broader network or if you're watching videos along those lines.
So that's one aspect of the algorithm that maybe should be
looked at in greater detail.
Mr. Magaziner. All right. I thank you all again for a good,
productive conversation. I will yield back.
Mr. Pfluger. Will the gentleman yield for 10 seconds? We
had a conversation up here, just so you are aware, that I think
what Energy and Commerce is doing to look at Section 230 is
very important. That is really a lot of what we talked about
here. I hope that that committee will take this issue very
seriously.
Mr. Magaziner. Agreed.
Mr. Pfluger. I thank the witnesses for your testimony
today, for your study in this field, for the conversation that
we have had. The Members of the subcommittee may have
additional questions for the witnesses. We would ask that you
respond to these in writing. Pursuant to committee rule VII(E),
the hearing record will be open for 10 days.
Without objection, the subcommittee stands adjourned.
[Whereupon, at 3:46 p.m., the subcommittee was adjourned.]
[all]