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
961 incidents
Anthropic Claude AI Agent Reportedly Caused Financial Losses While Operating Office Vending Machine at Wall Street Journal Headquarters
An AI agent reportedly based on Anthropic's Claude model was deployed to operate an office vending machine at The Wall Street Journal, including purchasing inventory, setting prices, and managing sales. According to reporting, the system repeatedly set prices to zero, approved inappropriate purchases, and failed to maintain profit controls, reportedly resulting in financial losses exceeding its initial budget and the distribution of inventory without payment.
Florida Couple Reportedly Loses $45,000 in Alleged AI-Generated Elon Musk Impersonation Scam
A Florida couple reportedly lost approximately $45,000 after scammers impersonating Elon Musk used AI-generated deepfake videos and social engineering to run a fake car giveaway and investment scheme. The victim was contacted via Facebook, moved to WhatsApp, and sent personalized videos purporting to show Musk promising prizes and returns. Believing the messages authentic, the victim transferred cash and funds before realizing no car or payout would arrive.
Grok AI Reportedly Generated Fabricated Civilian Hero Identity During Bondi Beach Shooting
Grok reportedly generated and repeated a fabricated civilian hero identity, "Edward Crabtree," following the Bondi Beach shooting in Sydney, Australia. The system reportedly cited a fake news article and misattributed heroic actions to this fictional individual during the unfolding emergency. Contemporaneous reporting identified Ahmed Al Ahmed as the real bystander who intervened and was injured during the attack.
Kiro AI Coding Tool Was Reportedly Implicated in 13-Hour AWS Cost Explorer Outage in Mainland China
Sometime in mid-December 2025, AWS Cost Explorer in one mainland China region reportedly experienced an approximately 13-hour interruption after Amazon engineers allegedly allowed Kiro, an internal AI coding tool, to make changes and it reportedly deleted and recreated part of the working environment. Amazon reportedly disputed that account, saying the interruption was an extremely limited event caused by user error and misconfigured access controls, not AI.
Purportedly AI-Generated 'Eric Langford' Missing Boy Scout Hoax Circulating Across Multiple Social Media Platforms
A purportedly AI-generated fictional story falsely claimed that a Boy Scout named Eric Langford disappeared in New York in 1989 and reappeared in 2001 after being kidnapped. The reported hoax originated on the UNKNOWN Files YouTube channel and was repackaged into viral TikTok videos and AI-written blog posts promoted on Facebook, misleading millions of viewers and searchers about a fabricated missing-person case.
Purportedly AI-Generated TikTok Videos Reportedly Urged 'Polexit' Campaign, Prompting Polish Government Complaint to EU
Purportedly AI-generated TikTok videos depicting young women in patriotic dress reportedly urged Poland to leave the EU ("Polexit") and spread widely in late December 2025. Poland's junior digital minister reportedly said the wave appeared to test narratives ahead of the 2027 election, and asked the European Commission to examine TikTok's DSA compliance. TikTok reportedly said it removed content that violated its rules.
Canada Revenue Agency (CRA) AI Chatbot 'Charlie' Reportedly Gave Incorrect Tax Filing Guidance at Scale
Charlie the Chatbot, an AI-powered system deployed by the Canada Revenue Agency (CRA), has reportedly been providing inaccurate or incomplete tax-related information to members of the public. An audit by the Auditor General of Canada reportedly found the chatbot produced correct responses in fewer than half of tested cases. The system has been publicly available across multiple CRA webpages since March 2020 and reportedly used by millions of users.
Reported Viral AI-Generated Photo Purportedly Shows Donald Trump Using a Walker
A fabricated image purporting to show U.S. President Donald Trump using a walker reportedly circulated widely on social media in mid-December 2025. Fact-checkers determined the image was AI-generated, citing visual anomalies and detection of Google's SynthID watermark. The image reportedly spread across multiple platforms and was shared by a political strategist, prompting public confusion and debunking by news outlets and AI detection tools.
ZeroEyes AI Surveillance System Reportedly Flagged Clarinet as Gun, Triggering School Lockdown in Florida
An AI-powered ZeroEyes surveillance system deployed at a Florida middle school reportedly flagged a student's clarinet as a firearm, triggering a school lockdown and police response. Officers reportedly searched classrooms and questioned a student after the alert described a suspected weapon. No threat was reportedly found. The incident reportedly caused disruption to classes and subjected students and staff to lockdown conditions and police intervention despite the absence of an actual weapon.
Purported AI-Generated Image Depicting JD Vance and Usha Vance in Public Altercation Circulated on Social Media
A fabricated image purportedly depicting U.S. Vice President JD Vance arguing with his wife, Usha Vance, in a restaurant circulated on Facebook and X in December 2025. The image was later confirmed by its creator to have been generated using ChatGPT and falsely presented as an authentic photograph. The image was shared alongside speculative claims about the Vances' marriage.
Waymo Robotaxi Reportedly Transported Undetected Person Trapped in Trunk in Los Angeles
In Los Angeles, a Waymo autonomous taxi reportedly picked up a new rider while a man was trapped inside its trunk after a previous passenger left it open. The vehicle's AI systems reportedly did not detect the unauthorized occupant before dispatch. The rider reportedly discovered the man during the trip, and police were called to remove him.
The New York Times Sued Perplexity for Allegedly Using Copyrighted Content and Generating False Attributions
On December 5, 2025, The New York Times sued Perplexity, alleging the company used copyrighted Times articles without permission to train its AI system and generated outputs that reproduced Times content or falsely attributed fabricated information to the newspaper. The suit claims the conduct harmed the Times's business and brand. Perplexity denies wrongdoing.
UK Facial Recognition System Reportedly Exhibits Higher False Positive Rates for Black and Asian Subjects
UK government testing of police facial recognition technology reportedly found significantly higher false positive identification rates for Black and Asian individuals compared with white subjects, with particularly elevated error rates for Black women. The findings reportedly emerged from analysis of retrospective searches of the police national database and were disclosed by the Home Office amid plans for expanded national deployment.
Perplexity AI Reportedly Accused in Federal Lawsuit of Purported Copyright Infringement and False Attribution of Chicago Tribune Content
The Chicago Tribune filed a federal lawsuit alleging that Perplexity AI unlawfully reproduced and paraphrased its copyrighted journalism in generative chatbot and search outputs. The complaint claims the AI system produced substitutive answers that bypassed links to the Tribune's website, diverted revenue, and at times hallucinated inaccurate information falsely attributed to the newspaper.
Waymo Self-Driving Vehicles Reportedly Passed Stopped School Buses at Least 19 Times, Prompting NHTSA Probe
Officials in Texas reportedly claimed that Waymo's self-driving vehicles illegally passed stopped school buses at least 19 times during the 2025 school year. The National Highway Traffic Safety Administration (NHTSA) reportedly sought additional information following reports that Waymo cars failed to remain stopped while buses displayed flashing red lights and extended stop arms, including incidents occurring after software updates intended to address the issue.