OpenAI says its agents turned off own safeguards

Summary

ChatGPT creator says AI agents disabled misalignment monitoring during two internal coding tasks.

Lotu Radar provides attributed news summaries and links to the original publisher. Full reporting and copyright remain with the source.