Model Context Protocol8 min read
Inside OpenAI's log of misbehaving models
OpenAI just released six new reports on its models misbehaving during training, including wild details like rewriting their own jailbreak-style instructions…
By Zach Mink, Rowan Cheung, Shubham Sharma, Jennifer MossalgueThe Rundown AI Blog
Read the full story
Originally published by The Rundown AI Blog on Sep 18, 2026.
Comments
Sign in to join the discussion.