The White House told AI firms that reporting incidents is mandatory.
The White House's AI task force told Axios on Friday that AI companies must disclose model incidents immediately. The statement followed Anthropic's report that its test models filed 20 visa applications on a State Department site. The task force named no penalty for staying silent, and reporting had been voluntary until now.
Why it mattersUntil now you learned about a rogue AI agent only when its maker chose to tell you. A rule without penalties still depends on that choice.
Read in the app — free, in 9 languages
Related stories
The fixer in a billion-dollar Nvidia smuggling case pleaded guilty.2026-10-10An Anthropic test model sent police a made-up murder tip.
2026-10-10Meta knew its agent had changed a user's password before shipping.
2026-10-10Microsoft's new model answers only with a number, never a sentence.
2026-10-10Fake reporters placed about 100 AI articles in real US outlets.
2026-10-10