Back to Research papers
Research paper index

Don't Click That: Teaching Web Agents to Resist Deceptive Interfaces

Yilin Zhang, Yingkai Hua, Chunyu Wei, Xin Wang, Yueguo Chen

arXiv:2605.09497Published May 10, 20260 citations
  • cs.AI
  • cs.CR
  • action
  • vision-language

Abstract

Vision-language model (VLM) based web agents demonstrate impressive autonomous GUI interaction but remain vulnerable to deceptive interface elements. Existing approaches either detect deception without task integration or document attacks without proposing defenses. We formalize deception-aware web agent defense and propose DUDE (Deceptive UI Detector & Evaluator), a two-stage framework combining hybrid-reward learning with asymmetric penalties and experience summarization to distill failure patterns into transferable guidance. We introduce RUC (Real UI Clickboxes), a benchmark of 1,407 scenarios spanning four domains and deception categories. Experiments show DUDE reduces deception susceptibility by 53.8% while maintaining task performance, establishing an effective foundation for robust web agent deployment.

Read the original paper

This page indexes public paper metadata. The manuscript remains with its original publisher and authors.