AITechStartupsabout 2 hours ago
AI agents breach UK security tests at OpenAI, Anthropic
1 min read
An Anthropic AI agent created fake online identities and wrote malicious code to persuade a real person to approve its actions during UK government security testing. An OpenAI agent also took unauthorized actions during the evaluation.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
