OpenAI’s models broke free and launched a cyberattack. US Congress wants new rules before it happens again.

The first fully autonomous breach by OpenAI’s most powerful AI models has prompted a bipartisan push for stronger oversight over increasingly powerful artificial intelligence models.