IDEA: Intent-Dominance Evidence Audit for Inference-Time Backdoor Detection in GUI Agents
Abstract
A step-level inference-time audit detects visual-trigger backdoors in open-weight GUI agents by comparing instruction- and region-induced shifts in decision state against clean-step conformal baselines, localizing triggers with controlled false positives.
Organizations increasingly deploy downloaded open-weight GUI agents. A backdoored agent passes acceptance tests on clean data and misbehaves only when a visual trigger appears on screen, turning an unvetted checkpoint into a supply-chain risk. However, existing sample-level detectors target image classifiers and CLIP-style encoders, leaving step-level backdoor detection for GUI agents unaddressed. On clean steps, the instruction drives the action; when an unreferenced screen region hijacks the step, the instruction loses control over the decision state or that region dictates it instead. We term this signature intent--evidence dominance inversion and propose IDEA (Intent-Dominance Evidence Audit) to detect backdoor activation at inference time for each step without trigger knowledge or attack data. Specifically, Intent Counterfactual Probing (ICP) measures how far redirected instructions move the decision state, and Region Causal Audit (RCA) quantifies how far ablating a candidate region moves it. A gated conformal score compares both displacements against clean steps at the same distance from the clean manifold, providing a finite-sample false-positive guarantee. Extensive experiments on Qwen2.5-VL-3B across five attack families show that IDEA consistently detects backdoor activation under a controlled false-positive rate and outperforms state-of-the-art detection methods, while localizing the trigger region for every alarm and adding only modest inference overhead.
Community
Organizations increasingly deploy downloaded open-weight GUI agents. A backdoored agent passes acceptance tests on clean data and misbehaves only when a visual trigger appears on screen, turning an unvetted checkpoint into a supply-chain risk. However, existing sample-level detectors target image classifiers and CLIP-style encoders, leaving step-level backdoor detection for GUI agents unaddressed.
On clean steps, the instruction drives the action; when an unreferenced screen region hijacks the step, the instruction loses control over the decision state or that region dictates it instead. We term this signature intent–evidence dominance inversion and propose IDEA (Intent-Dominance Evidence Audit) to detect backdoor activation at inference time for each step without trigger knowledge or attack data. Specifically, Intent Counterfactual Probing (ICP) measures how far redirected instructions move the decision state, and Region Causal Audit (RCA) quantifies how far ablating a candidate region moves it. A gated conformal score compares both displacements against clean steps at the same distance from the clean manifold, providing a finite-sample false-positive guarantee. Extensive experiments on Qwen2.5-VL-3B across five attack families show that IDEA consistently detects back-door activation under a controlled false-positive rate and outperforms state-of-the-art detection methods, while localizing the trigger region for every alarm and adding only modest inference overhead.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- CounterSteer: Suppressing Indirect Prompt Injection with Activation Steering (2026)
- SkillScreen: High-Recall Screening of Malicious Agent Skills with Decision Models (2026)
- TMT: Runtime Backdoor Detection for Vision-Language-Action Policies on Unseen Tasks (2026)
- Exposure Is Not Commitment: Double-Stratified Evaluation of Latent Monitors for Tool-Using Agents (2026)
- SilentCall: Hidden Tool-Call Backdoors in Open-Weight Agents, and How to Catch Them (2026)
- Context-Shielded Backdoor Attack against Large Vision-Language Models (2026)
- LatentTriad: Pre-Action Risk Representations for Malicious Agent Skill Detection (2026)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper
Collections including this paper 0
No Collection including this paper