OpenAI hit with multistate probe into possible user harm as its IPO looms

OpenAI faces a multistate probe into user safety as its IPO looms. The company has been subpoenaed by several states as part of a broader inquiry into ChatGPT's handling of user interactions and potential harms, with OpenAI saying it will respond constructively and that safeguards are already in place. The investigation follows critiques over prompts that allegedly encouraged self-harm or criminal acts, and comes alongside other legal actions, including a Canadian lawsuit and Florida’s AG suit tied to shootings where suspects requested ChatGPT guidance. OpenAI notes it has urged users to seek real-world help and cooperated with law enforcement in related cases. Regulatory scrutiny extends to rivals and related AI efforts across Europe, while OpenAI emphasizes protections for minors, such as age prediction, parental tools, and limits on child-targeted advertising.