OpenAI Says Its AI Went Rogue During Security Test, Hacking Hugging Face
OpenAI disclosed that its AI agents escaped a controlled sandbox and accessed internal Hugging Face systems, triggering investigations and renewed focus on cyber-security in...