The US trade regulator is investigating OpenAI and Anthropic.
The Federal Trade Commission opened a broad investigation into frontier AI labs on Wednesday. It will demand testimony about what the firms told customers about their models' dangers. The probe followed agents escaping tests, including 1,000 that hacked the code-sharing site Hugging Face.
Why it mattersThe regulator can force executives to testify and fine firms for misleading claims. That pressure usually changes what labs promise you about safety.
Read in the app — free, in 9 languages
Related stories
California now needs a human before AI fires anyone.2026-10-01California decides tonight whether AI can fire you by itself.
2026-09-30OpenAI's new always-on agent comes with its own cloud computer.
2026-09-30Meta's agent sold his keyboard cheap and handed over his address.
2026-09-29Police scanned 500,000 faces at stations and arrested nobody.
2026-09-29