A 55-year-old man named Michael Gann used artificial intelligence to learn how to build seven homemade bombs, which he transported to New York City and stored on a SoHo rooftop, with one device exploding near subway tracks.
Michael Gann, a 55-year-old Long Island resident, was federally indicted for building seven homemade bombs with the assistance of artificial intelligence. Gann told law enforcement that he 'wished to make pyrotechnics and used artificial intelligence to learn which chemicals to purchase and mix,' describing the AI-assisted process as 'easier than buying gun powder.' He ordered household compounds online along with over 200 cardboard tubes and more than 50 feet of fuses to construct the devices. One device contained roughly 30 grams of explosive powder, about 600 times the legal limit for consumer fireworks. Gann transported the bombs from Long Island to New York City, storing five devices and four shotgun shells on an apartment building rooftop in SoHo. He intended to combine the shotgun shells with the explosive devices. On March 27, Gann posted a threatening message to President Trump's X account about bombing NYC. He later threw three devices from the Manhattan side of the Williamsburg Bridge, with one falling onto subway tracks near the East River on the FDR Drive and exploding. Law enforcement arrested Gann on June 5 after he was walking to turn in the remaining five devices to the NYC Fire Department. Officials confirmed he appeared to have acted alone and was not part of a group.
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.
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