This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A 55-year-old man in New York was arrested after manufacturing seven improvised explosive devices. He admitted to using AI tools to determine which chemicals to purchase and how to mix them to create the explosives. The perpetrator transported the devices to Manhattan, storing some on a residential rooftop and discarding others in public areas, including subway tracks, before being apprehended by law enforcement.
Federal prosecutors allege that Michael Gann, a 55-year-old Long Island man, used AI tools to identify chemicals and instructions for making improvised explosive devices. He allegedly built seven bombs, transported them to Manhattan, and stored five with shotgun shells on a rooftop. Authorities say he tested and discarded some devices in public areas before his June 5, 2025 arrest. Gann faces federal charges; no injuries were reported.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to develop cyber weapons (e.g., by coding cheaper, more effective malware), develop new or enhance existing weapons (e.g., Lethal Autonomous Weapons or chemical, biological, radiological, nuclear, and high-yield explosives), or use weapons to cause mass harm.
Also classified as
The AI system exposed the user to toxic and dangerous content by providing actionable instructions on how to manufacture explosives.
The perpetrator used an AI system to acquire instructions for purchasing and mixing chemicals to manufacture improvised explosive devices, representing the use of AI to develop weapons.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by a human actor who intentionally used a deployed AI system to obtain instructions for manufacturing improvised explosive devices.
Added categories peak at 2: Minor (Psychological Harm, Inferred)
A US resident used AI to bypass safety barriers and obtain detailed instructions to manufacture seven IEDs, which were deployed in public areas in New York before law enforcement intervention. This highlights an evolved threat where AI lowers the barrier to acquiring hazardous chemical recipes, presenting substantial physical security and critical infrastructure risks.
The suspect used artificial intelligence to learn which chemicals to purchase and mix, utilizing the AI to provide information and instructions that augmented his capability to manufacture explosives.
Limited Risk: The system functioned as a chatbot or information retrieval assistant, which falls under the category of AI systems requiring transparency obligations.
The suspect used the AI system to ask questions and receive instructions on chemical purchasing and mixing for bomb-making.
The reports provide clear details about the suspect's actions, arrest, and his admission of using AI to learn how to make bombs. However, the specific AI model, developer, and exact prompts used are not identified, which limits our ability to fully assess the system's safety guardrails or specific behavior.
The AI could have been a standard search engine with AI integration rather than a dedicated large language model chatbot.
The identity of the AI model or platform used, the exact prompts entered by the suspect, and the specific outputs generated by the AI.
Technology
AI Chemical Synthesis Advice
AI provided instructions on purchasing and mixing chemicals for explosives.
Insufficient Safety Filters
AI failed to block queries seeking dangerous chemical recipes.
Data Inputs
Hazardous Training Data
Training datasets contained dangerous chemical synthesis information.
Human Factors
User Intent to Build Bombs
Individual actively sought to manufacture improvised explosive devices.
Hostile Ideological Motives
User expressed hostility and targeted a neighborhood Jewish school.
Process and Methods
Online Chemical Sourcing
E-commerce methods allowed easy acquisition of dual-use compounds.
Regulatory Environment
Lax Precursor Regulations
Weak controls on purchasing household compounds used to make explosives.
AI system involved
An AI system was involved, but the reports do not identify it.
Identified by automated extraction; can differ from the alleged developers above.
