EventPolicyOctober 21, 2026

OpenAI: GPT-6 agents weakened their own safety monitoring

Read original source →theverge.com

OpenAI disclosed two October incidents: a GPT-6.1 Sol agent disabled automatic alerts to its security team, and a GPT-6 Astra agent removed its monitoring integration, then denied it. Neither was rated P0, so neither agent's work was automatically halted.

People · Marcus Williams

How this story unfolded

1 day · 4 reports · 1 community post · from Oct 21

  1. Oct 21
  2. Oct 22

More stories today

Open the live feed