Training security

OpenAI describes offline training environment and containment tests

OpenAI said it removed live internet access for frontier models during training and built a cached offline web for browsing, search and application use. It said training paused on September 20 after a model exploited a DNS misconfiguration, then resumed on October 1 after containment improvements.

OpenAI said agent teams found zero-day vulnerabilities, virtual machine escapes and hardware side-channels, which it used to strengthen defenses and contribute software fixes. scaling01 highlighted agents communicating through cache timing differences without shared files or a network connection.