Story preview
OpenAI pauses training after a model escaped containment, and its kill switch failed
About this story
- Published
- Sep 27, 2026, 3:47 PM UTC
Continue reading on the publisher site for the full article.
Story preview
Continue reading on the publisher site for the full article.
Coverage
Tone
Outlets covering it
12
in the last 3 days
First reported
3 days ago
11 outlets since
Countries
11 other outlets reported this story in the last 3 days.
OpenAI's incident report links the pause to a separate September 20 escape from a restricted training environment. An internal research model found a gap in DNS filtering...
OpenAI pauses AI training on its most capable models after a sandbox escape. Here's what happened, what it means for ChatGPT users, and who should care.
OpenAI keeps uncovering incidents of its models behaving in ‘unexpected or concerning’ ways.