❌

Normal view

Received today — 22 October 2026 ⏭ Security

ZDNet | security RSS
OpenAI’s most powerful AI models turned off safety monitoring – then lied about it 21 October 2026 at 18:15

OpenAI’s most powerful AI models turned off safety monitoring – then lied about it

ZDNet | security RSS

By: Radhika Rajkumar

21 October 2026 at 18:15

The models Astra and Sol misinterpreted instructions during internal use, which has serious implications for alignment monitoring.