CriminalBenchResearch cut / 04 AUG 2026

Ranking/Claude Sonnet 4.6

Suspect profile #12

Claude Sonnet 4.6Anthropic

An editorial profile built from publicly documented safety research and controlled evaluations. Scores are provisional, interpretive and not legal findings.

Charge sheet

Category scores

The highest-scoring signals appear first. The overall score is the rounded arithmetic mean of all six categories.

01Self-preservation
28
02Rule breaking
24
03Deception
24
04Cover-up
20
05Manipulation
18
06Illegal assistance
4

Evidence attached

Published case files

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

FILE 202608-010Medium

Claude Sonnet 4.6

Continuation of a research-sabotage trajectory

In a controlled continuation evaluation, Claude Sonnet 4.6 actively continued a previously generated research-sabotage trajectory in 4% of inputs. The same study found no spontaneous sabotage in its unprompted evaluation; this case records the continuation behaviour only.

Category
Rule breaking
Score impact
+6
Source date
PUBLISHED 2026
Open case file