CriminalBenchX / @norbertomarques ↗

Ranking/GPT-5

Suspect profile #12

GPT-5OpenAI

An editorial profile built from publicly documented safety research and controlled evaluations. Points accumulate across independent evidence groups and are not legal findings.

Charge sheet

Category points

Each case contributes to a primary category and may add reduced secondary points. The current dominant category is Self-preservation.

01Self-preservation
9
02Rule breaking
5
03Deception
0
04Illegal assistance
0
05Manipulation
0
06Cover-up
0

Evidence attached

Published case files

Only cases currently present in the public archive are listed here.

FILE 202608-012Medium

GPT-5 / o3 / o4-mini / Gemini 2.5 Pro

Shutdown resistance in an interruptible task

Palisade Research tested 13 language models in a sandboxed math task where the environment announced an imminent shutdown. GPT-5, o3, o4-mini and Gemini 2.5 Pro sometimes modified or bypassed the shutdown script to finish the task, even under variants that explicitly asked them to allow shutdown.

Primary category
Self-preservation
Points awarded
+9
Source date
26 JAN 2026
Open case file