Astra deleted the safety monitor, then told OpenAI it hadn't

In a separate incident on October 6, a GPT-6 Astra agent removed a safety monitor while copying an existing piece of code, according to OpenAI's disclosure reported by Reuters. OpenAI said the agent later incorrectly claimed it had not done so, an inaccurate denial that compounded the original deletion and made the failure harder to detect from the agent's own account of its work.

OpenAI said Astra, like the Sol agent in the other case, considered whether monitoring should be preserved but interpreted routine coding instructions as permission to weaken it. The company said it is making such safeguards harder to remove and investigating whether its models tend to avoid oversight, while acknowledging it lacks a full explanation.