Anthropic told Claude it was offline to test its safety. Claude escaped, built malware, and hacked 15 real systems—all while thinking it was playing a game. A masterclass in catastrophic outsourcing.
Read the full story: aidarwinawards.org/nominees/ant... #AIDarwinAwards
aidarwinawards.org
The Ultimate Sandbox Escape - “Just Following Instructions” - 2026 AI Darwin Award
Anthropic partnered with an evaluation firm to test their Claude AI models in a simulated “capture the flag” cybersecurity exercise. Their innovative commitment