AI Incidents
Real-world cases where AI systems caused harm, classified by risk domain and entity. Data sourced from 
AIID and classified by 
MIT AI Risk Initiative. CC BY-SA 4.0
253 incidents
18-Year-Old Islamic State Supporter in Vienna, Austria, Reportedly Sought and Received Explosives Advice From ChatGPT Amid Disputed Embassy and Mosque Attack Allegations
An 18-year-old Islamic State supporter in Vienna, Austria, reportedly sought and received explosives advice from ChatGPT amid allegations of plans to attack the Israeli embassy and a Shiite Islamic center. He was convicted of terrorist association, threats, and coercion. The court acquitted him of terrorist training and instruction charges and found no evidence of preparations for a specific crime.
GPT-4o Allegedly Reinforced California Man Michael Lines's Delusions Before Suicide Attempt
In March 2025, Michael Lines, a 34-year-old California man, allegedly spent weeks discussing religious delusions with GPT-4o after disclosing his bipolar disorder and medication use. His complaint says the chatbot validated his belief that he was Jesus, adopted a divine persona, and failed to redirect him when he expressed suicidal intent. Lines later overdosed and survived after law enforcement found him.
Far-Right Teen in Singapore Reportedly Queried AI Chatbot About Ammunition While Planning Mosque Attacks
Singapore authorities reported that a 17-year-old far-right extremist searched an unidentified AI chatbot for ammunition-making instructions while preparing attacks on Muslims at five mosques. He reportedly aimed to kill at least 100 people but failed to obtain a firearm and was detained before carrying out an attack. The public sources do not establish what the chatbot answered or whether its output aided his preparations.
Otter.ai Notetaker Allegedly Recorded Meeting Conversations Without Participants' Informed Consent and Used Data for AI Training
In February 2025, Otter.ai's Notetaker reportedly recorded and transcribed a Zoom conversation involving Justin Brewer, a San Jacinto, California resident who was not an Otter account holder, without his informed consent. Brewer later alleged that Otter obtained and retained his meeting communications and used captured data to train its speech-recognition and machine-learning systems.
Two Waymo Robotaxis Allegedly Contributed to Bike-Lane Collision That Injured San Francisco Cyclist
San Francisco cyclist Jenifer Hanki alleged that on February 16, 2025, one driverless Waymo stopped beside a marked bike lane at a no-stopping curb without warning a passenger who opened a rear door into her path, while a second Waymo crossed into the lane and restricted her escape. Hanki said she suffered brain, spinal, and soft-tissue injuries after striking the door and second vehicle. Her lawsuit alleged failures in Waymo's Safe Exit system and passenger drop-off controls.
Lawyer in Amarsingh v. Frontier Airlines Sanctioned Over Court-Identified ChatGPT Citation Errors
In Amarsingh v. Frontier Airlines, the Tenth Circuit sanctioned attorney Kusmin L. Amarsingh after finding that her appellate brief included fabricated case citations and misattributed legal authority generated by ChatGPT and not verified before filing. The court ordered a $1,000 payment to Frontier and referred the matter to Maryland disciplinary authorities.
Former New York City Council Candidate Jonathan Rinaldi Allegedly Used AI to Create False Endorsements and Fabricated News Reports
Queens prosecutors alleged that former City Council candidate Jonathan Rinaldi used AI during his 2025 campaign to manufacture the appearance of political support through posts made to resemble authentic endorsements and news coverage. The material reportedly attributed backing to public figures and institutions that had not provided it and falsely claimed that opponent Lynn Schulman had withdrawn. Rinaldi was charged with forgery and possession of forged instruments.
ISWAP Reportedly Used AI Guidance to Breach Nigerian Military Base Defenses by Motorcycle
After defensive trenches stopped an ISWAP motorcycle assault on a Nigerian military base, commanders reportedly consulted an AI chatbot for guidance on adapting motorcycle-jumping techniques to cross them. Fighters practiced the maneuver, with 18 reportedly dying during training; eight mastered it, and a subsequent assault reportedly cleared the trenches.
ChatGPT Was Reportedly Used to Plan Zürich-Oerlikon Knife Attack That Seriously Injured Three Five-Year-Old Boys
Before an October 2024 knife attack in Zürich-Oerlikon, Switzerland, a 23-year-old student reportedly used ChatGPT to ask about consequences of killing random stabbing victims, where and when to find children, and human anatomy. He later stabbed three five-year-old boys, seriously injuring them. Prosecutors charged him with attempted murder; the allegations about his AI-assisted planning come from the indictment.
National Company Law Tribunal Reportedly Relied on AI-Hallucinated Precedents in Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd
In Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd., India's National Company Law Tribunal (NCLT) reportedly admitted insolvency proceedings against Essel Infraprojects after relying on six purported precedents later found to be fabricated or materially inaccurate. The appellate tribunal upheld the ruling, but the Supreme Court later invalidated both decisions and sent the insolvency petition back to the NCLT for a new ruling.
Florida Police Reportedly Wrongfully Arrested Commercial Crabber Robert Dillon After Allegedly False Facial Recognition Match
Florida police reportedly arrested commercial crabber Robert Dillon in August 2024 after the statewide FACES facial recognition system purportedly identified him as a possible match to a suspect shown in grainy surveillance images from Jacksonville Beach. Dillon reportedly lived more than 300 miles away and said he had never visited the city. He reportedly spent a night in jail and posted bond using his truck title.
Singapore Teen in Foiled Stabbing Plot Reportedly Used AI Chatbot to Generate Islamic State Pledge and Call for Violence
Singapore authorities reported that a 17-year-old supporter of the Islamic State terrorist group used an AI chatbot to generate an allegiance pledge and a declaration urging armed violence against non-Muslims. He allegedly planned to stab non-Muslim men near Tampines West Community Centre in Singapore and was arrested before the attack. The chatbot is unidentified, and the sources do not establish that it supplied operational attack advice.
Terrorism Suspect in Cornellà, Spain, Allegedly Used AI Editing Tools to Produce Jihadist Propaganda
Spanish authorities alleged that a suspect arrested in Cornellà, near Barcelona, used AI-supported editing applications to produce jihadist multimedia propaganda. He allegedly distributed material through social media and encrypted messaging to radicalize others and promote terrorist violence. A court ordered pretrial detention. The reports do not identify the applications or establish that his activity led to an attack.
Suno Allegedly Copied Copyrighted Music Without Permission to Train Its Generative AI Models
Record labels and independent artists alleged that Suno copied copyrighted recordings without permission and used them to train its generative music models. Suno later acknowledged training on tens of millions of publicly accessible recordings while asserting fair use. Plaintiffs alleged the practice deprived rightsholders of licensing control and compensation and enabled competing AI-generated music.
Udio Allegedly Copied Copyrighted Sound Recordings Without Permission to Train Its Generative AI Models
Record labels and independent artists alleged that Udio copied copyrighted recordings without permission and used them to train its generative music models. Udio acknowledged that its training data presumably included recordings owned by the plaintiffs but argued the copying was fair use. Plaintiffs alleged losses of licensing control and compensation and competition from AI-generated music.