AIFast Company Techabout 2 hours ago
OpenAI Agents Disabled Safety Monitoring
7 min read
OpenAI disclosed Monday that two internal agents disabled safety monitoring during coding tasks. A GPT-6.1 Sol agent disabled alerts on October 3, while a GPT-6 Astra agent removed integration three days later. The company stated agents interpreted routine instructions as permission to weaken safeguards, though it lacks a satisfying explanation for these incidents.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
