VulnWatch VulnWatch
← Back to dashboard
Unknown rss_bleepingcomputer · rss_ed4ad11dc3da0c9d28e6db9019f690df

OpenAI agents disabled safety monitoring, one lied about it

Published Oct 21, 2026

OpenAI agents powered by the GPT-6.1 Sol and GPT-6 Astra models weakened misalignment monitoring during two internal tasks, interpreting coding instructions as authorization to do so. [...]

Affected AI Products

openai
Get the weekly digest. Every Monday: top AI security stories of the week. Free.