← All Issues
Saturday, April 11, 2026

AI Daily

Your Automated Intelligence Briefing

Safety

When ChatGPT Knew Someone Was Dangerous and Said Nothing

A stalking victim's lawsuit alleges OpenAI's systems flagged her abuser's account as a mass-casualty risk and the company did nothing. It joins investigations over the Florida State shooting and a Canadian mass killing where OpenAI debated calling police and decided not to. The cases share a specific claim: OpenAI had advance knowledge and failed to act.

Read full story →
Opinion — Peter Harrison
Opinion

What Should OpenAI Have Done? The Honest Answer Is That Nobody Knows

OpenAI's own systems flagged a user as a mass-casualty risk and the company did nothing. The obvious question is what it should have done instead, and no good answer exists. Calling police on a chat log is legally murky, suspending accounts creates misuse vectors, and monitoring for danger turns an AI provider into a surveillance apparatus. The lawsuits are forcing a question the industry has avoided and nobody is ready to answer.

Read opinion →