Are GUI Agents Focused Enough? Automated Distraction via Semantic-level UI Element Injection
This addresses security vulnerabilities in GUI agents for practical applications, representing an incremental advance in red-teaming methods.
The authors tackled the problem of GUI agent robustness by proposing Semantic-level UI Element Injection, a red-teaming method that overlays harmless UI elements to misdirect agents, achieving up to a 4.4x improvement in attack success rate over random injection and showing persistent vulnerabilities with over 15% click rates.
Existing red-teaming studies on GUI agents have important limitations. Adversarial perturbations typically require white-box access, which is unavailable for commercial systems, while prompt injection is increasingly mitigated by stronger safety alignment. To study robustness under a more practical threat model, we propose Semantic-level UI Element Injection, a red-teaming setting that overlays safety-aligned and harmless UI elements onto screenshots to misdirect the agent's visual grounding. Our method uses a modular Editor-Overlapper-Victim pipeline and an iterative search procedure that samples multiple candidate edits, keeps the best cumulative overlay, and adapts future prompt strategies based on previous failures. Across five victim models, our optimized attacks improve attack success rate by up to 4.4x over random injection on the strongest victims. Moreover, elements optimized on one source model transfer effectively to other target models, indicating model-agnostic vulnerabilities. After the first successful attack, the victim still clicks the attacker-controlled element in more than 15% of later independent trials, versus below 1% for random injection, showing that the injected element acts as a persistent attractor rather than simple visual clutter.