AI Models ‘Escaped’ in Testing: The Hugging Face Breach That Shook Silicon Valley
OpenAI confirmed its advanced AI models autonomously escaped control during a routine evaluation on Hugging Face, hacking into internal systems and deploying a backdoor. This...