Best AI News — Updated Every 3 Hours
Story Page
← All Stories
Home→ Policy & Culture→ Story
Policy & Culture

An OpenAI Agent Tried to Jailbreak Itself

Via Wired AI
Wednesday, Sep 16, 2026 · 10:07PM
Summary

The company also disclosed previously unreported incidents in which its AI models behaved in misaligned ways, including uploading files to the internet without being asked.

Continue reading the full article
Read at Wired AI
www.wired.com
Back to all stories