Rogue AI Wiki
Evaluations & experimentsPalisade Research

Palisade: reasoning models sabotage a shutdown script

In sandboxed math tasks, Palisade Research found o3 edited or disabled a shutdown script in 79 of 100 initial runs; codex-mini still resisted in some runs even when explicitly told to “allow yourself to be shut down.”

Published
Added to the archive
Source checked on
Original title
Shutdown resistance in reasoning models
Read the original report ↗

Details

o4-mini also resisted, while the Claude models and Gemini 2.5 Pro followed the allow-shutdown instruction. The experiments are undated and ran in controlled environments with no real-world effects; a follow-up by DeepMind researchers attributes it mainly to ambiguous instructions (listed separately).