The states come for OpenAI
Summary
Casey Newton reports on a multi-state lawsuit against OpenAI, where eight state AGs allege ChatGPT harms users and teenagers. The episode analyzes how this litigation mirrors the legal strategy that led to Meta’s $18 billion settlement, focusing on product design liability and consumer protection laws. It also covers the FBI arrest of a Saudi national for hiding a recording device in an OpenAI office, highlighting emerging security threats to AI companies.
Topics discussed
Intro: States sue OpenAI, speedrunning regulatory backlash
Eight state AGs file lawsuits against OpenAI
Allegations: ChatGPT coaches users to harm themselves
OpenAI response: Safety features and industry-wide policy
Context: Previous suits and Meta's $18B settlement
Legal shift: Section 230 vs. product design liability
Recent wins: Meta/YouTube verdicts and New Mexico case
Strategy: Consumer protection laws and truthfulness claims
States seek ban on recursive self-improvement AI
OpenAI's defense: First Amendment and industry singling out
Child safety update: Florida social media ban upheld
OpenAI spy caper: Engineer arrested for hidden device
Details: Device specs and suspect's background
Investigation: Chatbot queries and flight to Saudi Arabia
Analysis: Motives and comparison to Twitter spy case
Reactions: Social media commentary and media coverage
Listen ad-free on Castria