US cracked down on anti-tech extremists amid AI security fears
US agencies step up surveillance of anti-tech extremist groups linked to AI unrest since 2024.
Why it matters: Legal professionals must track these government moves impacting AI regulation, civil liberties, and national security risks.
- Department of Homeland Security and FBI expanded monitoring of 'anti-technology extremist' groups amid AI-related protests (leaked 1,000+ page reports).
- White House Office of Science and Technology Policy accused China of industrial-scale AI theft campaigns using proxies and software exploits.
- Pentagon officially labeled AI company Anthropic a supply chain security risk, restricting federal contracts for its chatbot Claude.
- Anthropic reported over 16 million fraudulent exchanges targeting its systems from Chinese AI firms, raising security alarms.
The U.S. Department of Homeland Security (DHS) and Federal Bureau of Investigation (FBI) have increased surveillance of groups classified as 'anti-technology extremists,' citing concerns about potential violent protests and disruptions targeting AI projects and related infrastructure. This escalation was revealed in a cache of over 1,000 pages of leaked documents from DHS, FBI, and joint intelligence centers, dated 2023–2024, showing surveillance expansion linked to rising AI policy conflicts.
These groups oppose certain AI data centers and technology deployments and are monitored due to fears of escalating civil unrest. The documents, detailed by Bloomberg Law, emphasize that this monitoring includes tracking online communications, protest plans, and coordination efforts. Legal experts warn of potential civil liberties concerns as peaceful opposition risks mischaracterization under these surveillance measures.
Simultaneously, the White House Office of Science and Technology Policy (OSTP) publicly highlighted national security threats from foreign actors. OSTP Director Michael Kratsios detailed allegations of "deliberate, industrial-scale campaigns" by China involving tens of thousands of proxy users and software 'jailbreaking' techniques—methods that bypass AI model security to steal proprietary data and gain unauthorized access. This is documented in OSTP reports summarized by Reuters.
Moreover, the Department of Defense designated the AI company Anthropic a supply chain risk in late 2024, prohibiting its use in government contracts. Anthropic's chatbot service Claude was reportedly targeted by Chinese AI competitors executing over 16 million fraudulent attempts to exploit and infiltrate its system, as reported by The Wall Street Journal.
These actions reflect the broader July 2025 White House AI Action Plan, which targets adversarial exploitation of AI innovations, alongside the "Cyber Strategy for America" emphasizing public-private collaboration in cybersecurity and technological safeguarding.
For legal professionals engaged in AI regulation, national security law, and civil liberties, these developments underscore the growing complexity of AI policy enforcement. Surveillance expansions and national security restrictions will likely affect AI companies’ operations, governmental procurement, and opposition rights.
By the numbers:
- 1,000+ pages — leaked reports documenting expanded monitoring of anti-tech extremist groups
- 16 million — fraudulent AI interactions Anthropic reported from Chinese operators
- 24,000+ — accounts used as proxies in Chinese campaigns to steal U.S. AI technology