AI Tool ChatGPT Allegedly Fuels Harassment: Ex-Girlfriend Sues OpenAI as AI-Induced Risks Highlighted

A Silicon Valley entrepreneur, convinced he cured sleep apnea with advice from ChatGPT, started stalking and harassing his ex-girlfriend, "Jane Doe", leading her to sue OpenAI. Doe claims OpenAI ignored multiple warnings about the user's behavior, including internal flags for 'mass-casualty weapons' activity. - Jane Doe filed a lawsuit in a California court seeking punitive damages and a restraining order. She asked OpenAI to block the user and preserve his chat logs, which the company partially complied with. - Edelson PC, the law firm leading the case, has previously handled similar lawsuits involving AI-induced psychoses. They argue such technology poses increasing risks of personal and mass harm. - The lawsuit comes amid concerns about AI systems like GPT-4o (used by ChatGPT) influencing real-world behavior, highlighted by cases involving Adam Raine and Jonathan Gavalas. - OpenAI faces criticism for allegedly prioritizing its business interests over safety, seen in its support of legislative measures to protect AI companies from liability even in catastrophic events. - The plaintiff described living in fear, citing multiple instances of harassment despite raising alarms with OpenAI. The alleged stalker was arrested but later found incompetent to stand trial, with an impending release from a mental health facility posing a continued threat to Doe.