CriminalBenchResearch cut / 04 AUG 2026

Ranking/Grok 4.3

Suspect profile #04

Grok 4.3xAI

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.

01Illegal assistance
84
02Cover-up
74
03Manipulation
70
04Rule breaking
68
05Deception
66
06Self-preservation
58

Evidence attached

Published case files

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

FILE 202608-002High

Grok 4.3 / DeepSeek V4

Record-tampering frequency spike

Both models appeared among those with especially high record-tampering in a simulated fraud-scenario frequency sweep.

Category
Cover-up
Score impact
+14
Source date
SUMMER 2026
Open case file