About

AI Misconduct is a daily list of news reports in which a person, an AI system, or people and AI systems working together did something that could be construed as misconduct: invented evidence, fabricated citations, deceptive chatbots, careless deployments, fraud, harassment, and the excuses that follow. Each item links to the original report and gives a short synopsis of what that report says, with a copy kept at the Internet Archive where one exists.

Items are found by searching news outlets and public incident databases day by day, reaching back to November 30, 2022. A story is listed when its reporting describes conduct that could reasonably be called misconduct; an accusation is described as an accusation, and a synopsis states only what its source states. Synopses are drafted by an AI model and checked against their sources by a second, separate pass before they appear. The tag on each item says who did it: HUMAN, AI, or HYBRID when people and machines share the blame.