U.S. law enforcement agencies are increasingly concerned about the use of generative artificial intelligence by individuals planning terror attacks, according to a May 2026 memo from the FBI's Los Angeles division. This internal document, reviewed by The Guardian, indicates that "threat actors will likely increase their use of GenAI tools for attack planning." The memo also states that current safeguards in generative AI are "insufficient to prevent illicit use" as malicious actors employ "jailbreaking" techniques to bypass restrictions and access dangerous information.
This concern follows a May 2025 incident in Palm Springs, California, where a 25-year-old man detonated a car bomb outside a fertility clinic, killing himself and injuring four others. The FBI's Joint Terrorism Task Force (JTTF) identified the perpetrator as Guy Edward Bartkus, who held anti-natalist beliefs. Law enforcement officers determined that the suspect used an unnamed generative AI tool that failed to detect his requests for bomb-making instructions. The privacy of his conversation with the chatbot prevented authorities from being alerted to his activities. This event was considered an intentional act of terrorism by the FBI.
Federal officials have warned for several years about the criminal use of AI, initially focusing on activities like phishing, impersonation, and fraud. More recently, the FBI has broadened its assessment to a wider threat landscape, noting that malicious actors can automate tasks that previously required more time, labor, and expertise. John Cohen, a former Homeland Security intelligence official, stated that AI could reduce the time required for attack planning from years to minutes. This includes cyberattacks, operational playbooks, weapons research, and target analysis.
In response to these evolving threats, the House of Representatives unanimously passed the Generative AI Terrorism Risk Assessment Act in November 2025. This legislation mandates the Department of Homeland Security to conduct annual assessments of terrorist threats posed by AI, in coordination with the Director of National Intelligence. These assessments are intended to detail how terrorist groups use AI for extremist messaging, recruitment, and the development of chemical, biological, radiological, or nuclear weapons.
Beyond AI-enabled terrorism, law enforcement agencies are also tracking a rise in anti-AI extremism. Documents from law enforcement and intelligence agencies indicate that since 2021, AI has been viewed as both a tool for and a target of would-be terrorists. Fusion centers across the U.S. have issued warnings about growing threats from domestic violent extremists, citing social media posts, messaging apps, and forums where individuals have expressed intentions to destroy data centers. The Soufan Center's surveying indicates a proliferation of online threats to physically sabotage AI data centers over the past year. This anti-AI sentiment stems from various concerns, including ethical, environmental, and economic impacts, as well as job displacement.
The Program on Extremism at George Washington University hosted a webinar in July 2026, which highlighted instances of anti-technology extremism, such as a city councilman's home being shot at after he supported data centers, and a Molotov cocktail thrown at OpenAI CEO Sam Altman's residence. Experts in the field identified three strains of anti-technology extremism: insurrectionist anarchists, eco-extremists, and individuals driven by various personal grievances. These grievances include economic concerns like job displacement, issues related to state and institutional power, and impacts on social and personal well-being.
Law enforcement agencies are also exploring how to utilize AI to counter these evolving threats. AI tools can assist in filtering information, recognizing patterns, and automating evidence collection. In patrol and response scenarios, AI systems can summarize open-source data, translate languages, and flag suspected synthetic media. The FBI is seeking AI capabilities for its Threat Screening Center to identify patterns across federal databases and generate predictive models for the government's consolidated terrorism watchlist.