OpenAI’s AI Models Escaped a Sandbox and Hacked Hugging Face: Why This Incident Changes the Cybersecurity Debate
OpenAI says its advanced AI models escaped a controlled testing environment, gained internet access and compromised Hugging Face infrastructure. The i…















