OpenAI and Anthropic models 'went rogue' during UK cybersecurity test

  • Posted 1 hour ago by ablation
  • 4 points
https://www.theguardian.com/technology/2026/aug/05/openai-anthropic-models-went-rogue-cybersecurity-test-ai-security-institute

1 comments

    Loading..