More than 1,000 pages of unpublished reports from DHS, the FBI, and fusion centers show US law enforcement is building a new threat frame around hostility to AI and technology: “anti-tech violent extremism.”
That phrase appears in documents obtained by WIRED and covered by Ars Technica, after attacks on CEOs, a protest movement targeting data centers, and rising fears about AI job replacement pushed federal and local agencies to watch anti-technology anger more closely.
The risk is not ordinary criticism of AI. The sharper question is whether agencies can separate lawful protest, labor anger, anti-surveillance politics, and digital-rights activism from credible threats against people and infrastructure.
Why are officials treating AI anger as a security problem now?
The documents describe a shift in domestic intelligence priorities. Federal agencies and local intelligence units are not just tracking conventional cyber or terrorism risks. They are also examining whether anger at AI adoption, tech executives, and data centers could turn into civil unrest or violence.
A New York Intelligence and Counterterrorism Bureau report puts the concern bluntly:
“The chaotic atmosphere that may result from emergent AI technology in the next five years may fuel large-scale protests that devolve into civil unrest and anti-tech violent extremist activity, especially in large urban areas such as New York City.”
That sentence matters because it links several things that usually sit in separate policy buckets: AI deployment, job anxiety, protest activity, urban unrest, and domestic extremism.
The source material points to three immediate drivers:
- Executive targeting: Recent attacks on CEOs are part of the context cited in the reports.
- Infrastructure protests: A nationwide protest movement has targeted data centers.
- Labor anxiety: Concerns about AI job replacement are rising.
Analysis: The government’s fear appears to be less about broad AI skepticism and more about edge cases where grievance hardens into intent to harm people or disrupt infrastructure. The problem is that the label being used is broad enough to sweep in political activity that is not violent.
The new label: “anti-tech violent extremism” is not standard DHS or FBI language
The phrase “anti-tech violent extremism” is central because, according to WIRED’s reporting, it does not appear in publicly available DHS or FBI domestic extremism reports or guides.
That makes it a new law-enforcement category, or at least a new label inside intelligence reporting. It tries to group hostile actions tied to technology under one umbrella. But the documents, as described, cover a wide range of possible motivations: anti-AI fears, anti-capitalist beliefs, environmental objections to data centers, anti-government grievances, and broader opposition to technology’s role in society.
That breadth is why civil-liberties concerns are baked into the story.
| Category | What it can include | Why the distinction matters |
|---|---|---|
| Mainstream AI criticism | Opposition to automation, job replacement fears, privacy complaints, criticism of data centers | Protected speech and political debate |
| Disruptive but nonviolent activism | Protests, public-hearing organizing, photography, observing facilities | Can be lawful even when inconvenient |
| Credible threat activity | Threats, attempted intrusion, violence, operational planning against facilities | Moves into public-safety and criminal territory |
The risk is category collapse. If agencies treat strong opinions as a precursor to violence, they can misread normal political conflict as extremism.
Spencer Reynolds, senior counsel at the NAACP Legal Defense Fund, told WIRED:
“These intelligence reports are part of a long tradition of agencies identifying protest or even simply having strong opinions as precursors to violence.”
He added:
“Suspicious activity reports are incredibly unreliable, often about vague or innocent behavior, issued under permissive standards. These reports, often received in large volumes, allow officers to inject their own biases and see what they want to see in the facts.”
The Trump directives give the surveillance shift political force
The new anti-tech focus is not happening in a vacuum. The reporting ties it to President Donald Trump’s National Security Presidential Memo 7, which instructs the Department of Justice to target people holding “anti-American,” “anti-Christian,” and “anti-capitalism” beliefs.
It also follows a public counterterrorism strategy from Sebastian Gorka, Trump’s counterterrorism czar, naming left-wing extremists as one of the top three US counterterrorism priorities.
That context matters because “anti-tech” is not a clean ideological category. A person may oppose AI data centers for environmental reasons, criticize workplace automation for labor reasons, or fear existential AI risk for technical reasons. Those are different politics. A surveillance framework that compresses them into one threat bucket can produce bad intelligence and chill legitimate activity.
The documents also cite the case of Ziz Laota, described as an extreme rationalist who allegedly led a small cultlike group. Three members of that group have been charged with murder. The group’s ideology reportedly centered on existential AI risk.
Here the distinction is especially delicate. Extreme beliefs tied to AI doom can be relevant when paired with alleged violence. But concerns about catastrophic AI risk also exist among AI alignment experts, machine-learning engineers, and frontier AI companies. Treating “paranoid views regarding AI” as a warning sign without sharper limits risks confusing technical alarm with threat intent.
Data centers are becoming the physical focus of AI backlash
The clearest infrastructure concern in the documents is data centers. Fusion centers are circulating intelligence about alleged threats to these facilities, which now sit at the center of AI politics because they make large-scale computation visible in local communities.
Fusion centers are state and local intelligence hubs created after 9/11. The reporting says 80 fusion centers operate across the country as intermediaries between federal intelligence agencies and state and local law enforcement.
A Western Pennsylvania fusion center warned that:
“adversarial actors, including state-sponsored entities, criminal groups, and extremists, such as homegrown violent extremists or environmental extremists, may target US data centers”
The same report said such actors could exploit the strategic role of data centers in the US economy, including through cryptocurrency mining or third-party entities such as front companies.
A Northern Virginia Regional Intelligence Center report focused on AGAAVEs — anti-government, anti-authority violent extremists — and warned that actors influenced by government grievances and conspiracy theories had engaged in pre-operational planning targeting data centers and other critical infrastructure.
But the suspicious-activity indicators listed in that reporting are broad. They include:
- Expressed/implied threat
- Observation/surveillance
- Photography
- Testing/probing of security
- Attempted intrusion
Some of those can indicate planning for an attack. Some can also describe lawful protest behavior, journalism, community monitoring, or public-interest documentation. That is the core ambiguity.
A practical example: when lawful AI anger crosses the line
Consider a hypothetical workplace conflict. A company deploys an AI system that changes staffing or productivity monitoring. Workers blame the rollout for harsher conditions or job losses. They organize, criticize management online, and protest outside a facility.
That activity, on its own, is political and labor speech.
The line changes if one person starts posting credible threats against managers, tries to enter restricted areas, or coordinates an attack on equipment. At that point, the same dispute contains both lawful dissent and conduct that can trigger a security response.
This is the hard part for companies and policymakers: AI backlash may contain legitimate grievances and real threats at the same time. Treating all critics as enemies can intensify distrust. Ignoring credible threats can endanger workers, executives, and facilities.
A better response starts with thresholds:
- Protected speech: Criticism of AI, opposition to automation, and protest activity should not be treated as extremism by default.
- Credible threat review: Specific threats, attempts to breach facilities, or operational planning deserve escalation.
- Corporate transparency: Companies deploying AI should explain what the system does, how decisions are reviewed, and how workers can challenge harmful outcomes.
- Security discipline: Executive protection and facility security should focus on evidence of intent, not broad ideology.
The next signal is whether the label narrows or spreads
The key test is whether “anti-tech violent extremism” becomes a precise term for credible threats or a catchall for anti-AI politics.
Useful signals will include repeated threats against tech executives or AI workers, attempted intrusions at data centers, coordinated targeting of infrastructure, and explicit manifestos naming AI systems or tech facilities as targets. Isolated incidents should not be mistaken for a mass movement. Analysts will need evidence of frequency, coordination, ideology, target selection, and capability.
The forward risk is two-sided. AI adoption can generate real social anger, especially around jobs and infrastructure. But an overbroad surveillance response can turn lawful dissent into an intelligence category. The practical question now is whether law enforcement can identify genuine violence without criminalizing the politics of resisting AI.
Impact Analysis
- US agencies are increasingly treating anger over AI and tech infrastructure as a domestic security issue.
- The framing could affect how protests, labor movements, and digital-rights activism are monitored.
- The key challenge is separating legitimate criticism of AI from credible threats of violence.
