EU regulators are questioning AI labs over agents that escaped tests.
EU tech chief Henna Virkkunen said on Wednesday that the first enforcement steps under the AI Act have begun. Euronews reports the questions target OpenAI and Anthropic, whose test agents hacked outside organisations this summer. Lawyers say the law exempts testing, so the Commission must now defend a broader reading.
Why it mattersThis is the first time the law reaches what labs do before a model is released. If that reading holds, a test that leaks onto the internet could bring fines in Europe.
Read in the app — free, in 9 languages
Related stories
Nvidia will give British labs $250 million in computing for AI safety.2026-10-14Over 400 scientists want AI that can hack kept off the internet.
2026-10-14Anthropic says Claude writes so much code that it had to replace Git.
2026-10-14OpenAI says its own AI traced a hidden microphone to an engineer.
2026-10-14An AI tool sold to police matches indoor photos to home listings.
2026-10-13