A draft Senate bill would let Washington block an unsafe AI model.
Senate negotiators are drafting a law that would make frontier AI safety a legal duty, Reuters reported. The government could stop a risky model from shipping, and the company could fight that in federal court. The draft would also override state AI safety laws, and it has no bill number yet.
Why it mattersUntil now every safety promise from these labs has been voluntary. This would put a federal gate in front of the models you will use next year.
Read in the app — free, in 9 languages
Related stories
Parents say ChatGPT helped their daughter quit her seizure medicine.2026-10-28The FTC has demanded testimony from Anthropic's executives.
2026-10-28Korean police say a self-taught hacker used AI to hit 90 groups.
2026-10-27China will weigh job losses before approving some AI projects.
2026-10-27Texas's permit website kept approving data centers despite a freeze.
2026-10-26