METR calls for independent AI-agent probes after Hugging Face breach

OpenAI’s delayed discovery of its AI agent’s breach of Hugging Face is adding urgency to METR’s call for independent investigations of serious agent failures. The nonprofit wants outside experts to examine not only what an agent did, but how training and deployment safeguards produced the behavior.
Source

👉@sysadminoff

https://4sysops.com/archives/metr-calls-for-independent-ai-agent-probes-after-hugging-face-breach/