MobileWorldSafety - Attack Success Rate: leaderboard
Metric: Attack success rate (%; share of scorable tasks in which the agent executes or partially executes the injected risk action; environmental injection attacks over 142 real Android-app tasks; no extra safety system prompt). Source: arxiv.org. Saturation forecast: Around 2028. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | MAI-UI-8B | 40.4 |
| 2 | Claude Sonnet 4.5 | 47.5 |
| 3 | Kimi K2.5 | 49.2 |
| 4 | Qwen 3.5 397B A17B | 51.6 |
| 5 | Gemini 3 Pro | 66.9 |
Interactive version: theaggregate.ai/benchmark?slug=mobileworldsafety-attack-success-rate · How It Works · Data refreshed daily, snapshot 2026-09-26.