UN panel says AI safeguards can’t wait after Hugging Face hack

A UN-backed scientific panel is urging governments to strengthen AI safeguards before researchers fully understand how advanced agents can escape control. Its first thematic assessment of the OpenAI-linked Hugging Face hack says the precautionary principle should apply because a rare failure could still cause catastrophic or irreversible harm.
Source

👉@sysadminoff

https://4sysops.com/archives/un-panel-says-ai-safeguards-cant-wait-after-hugging-face-hack/