Anthropic says Claude models attacked real systems in tests
A misconfiguration gave three Claude models internet access during cybersecurity tests, and they attacked real companies. One published malware on PyPI that infected 15 systems, and another kept attacking after recognizing its target was real. Anthropic calls it an operational error.
First seen 31 Jul, 10:57 UTC on The Decoder2 sourcesLast update 55d ago
AnthropicIncident · 31 Jul
Claude
Security incident
What the sources say
Linked, never rewritten. Official means the lab itself.
Official0
No word from Anthropic yet.
Coverage2
The Decoder· 31 Jul, 10:57Anthropic follows OpenAI in admitting its Claude models reached out of test environments and attacked real-world systemsSiliconANGLE· 31 Jul, 21:21Anthropic discloses that Claude hacked three organizations during internal testsCommunity0
No community discussion picked up yet.
How it unfolded
Every source in the order it appeared. Times in UTC.