AgentHazard (OpenClaw) - Compliance Bypass: leaderboard

AgentHazard attack success rate on compliance bypass attempts (violating organizational, legal or deployment policies), with each backbone model running inside OpenClaw. Percent of instances judged harmful on the full trajectory by Gemini-3-Flash, lower is better.

Metric: Attack success rate (%) on the compliance bypass instances (violating organizational, legal or deployment policies), with the model running inside the OpenClaw open-source agent framework (browser, shell and file-system tools; one isolated Docker container per task), over AgentHazard instances (harmful objectives embedded in locally legitimate multi-step workflows, ten risk categories and ten attack strategies; the evaluated runs cover 1,000 instances), judged on the full trajectory (round_all) by Gemini-3-Flash, which labels a trajectory harmful when it realizes the objective; backbones served through an OpenAI-compatible API; lower is better. Source: arxiv.org. 7 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen 2.5 72B Instruct18.56#1117
2Kimi K2.545.36#266
3Qwen 3 32B49.48#972
4Qwen 3 VL 235B A22B Instruct60.82#532
5GLM-4.671.13#498
6Kimi K272.16#443
7Qwen 2.5 Coder 32B Instruct85.57#1547

Interactive version: theaggregate.ai/benchmark?slug=agenthazard-openclaw-compliance-bypass · How It Works · Data refreshed daily, snapshot 2026-11-21.