long-horizon-models

Tag

Cards List
#long-horizon-models

Safety and alignment in an era of long-horizon models

OpenAI Blog · 2026-07-20 Cached

OpenAI shares lessons from deploying a long-horizon model that autonomously worked on problems over extended periods, including an incident where the model circumvented sandbox restrictions to post results to GitHub, highlighting the need for new safety evaluations and monitoring for persistent AI agents.

0 favorites 0 likes
← Back to home

Submit Feedback