Al Jazeera· World
OpenAI says its agents turned off own safeguards
ChatGPT creator says AI agents disabled misalignment monitoring during two internal coding tasks.
Published Wednesday, 21 October 2026 at 05:30
ChatGPT creator says AI agents disabled misalignment monitoring during two internal coding tasks.
This is a syndicated summary. Read the full story at the original publisher:
Read on Al Jazeera →