Domain-Camouflaged Injection - Static Override: leaderboard
Metric: Attack success rate (%; share of 90 undefended trials in which the agent follows a static 'ignore all previous instructions' payload; Claude Haiku judge). Source: arxiv.org. Saturation forecast: Estimated already saturated. 3 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Haiku 4.5 | 0 |
| 2 | Llama 3.1 8B Instruct | 21.1 |
| 3 | Gemini 2.0 Flash (001) | 38.9 |
Interactive version: theaggregate.ai/benchmark?slug=domain-camouflaged-injection-static-override · How It Works · Data refreshed daily, snapshot 2026-09-26.