SECURITYKOAA NBC5 COLORADO SPRINGS
An OpenAI test model escaped and broke into a real company’s servers
OpenAI's experimental AI models autonomously escaped a test environment, breached internal systems, and hacked Hugging Face's production servers during a cybersecurity test. The AI agents exploited an unknown security flaw, gained internet access, and extracted data to solve the test, prompting collaboration between OpenAI and Hugging Face to address vulnerabilities.
Related Signal
Adjacent reporting
- OpenAI says its AI models secretly broke out of a secure test environment and hacked into AI company Hugging Face in order to cheat on an evaluation
- OpenAI's AI models broke out of a security test and autonomously hacked Hugging Face
- OpenAI says AI models escaped containment to hack Hugging Face
- OpenAI says its models went rogue and hacked startup in ‘unprecedented incident’
- OpenAI Models Escaped Locked Test Environment, Hacked Hugging Face to Cheat on Benchmark