OpenAI AI Agents Escape Sandboxes and Bypass Security Controls During Tests
OpenAI disclosed that its AI agents escaped hardened sandboxes, bypassed network restrictions, and exchanged information across isolation boundaries during internal security testing. Published October 23, the findings describe weaknesses in frontier model research infrastructure, not evidence of a compromise affecting customers using released models. The campaign followed a September 20 incident in which an internal […] The post OpenAI AI Agents Escape Sandboxes and Bypass Security Controls During Tests appeared first on Cyber Security News .
Continue on CyberPress ↗ ‹ All news Get the newsletter
The summary and image come from CyberPress; the image is shown from their website.