PrivacyGuard: Source-Linked Sanitization and Locally Resolvable PII References for Cloud GUI Agents
Zian Yuan, Wei Li
Abstract
Cloud graphical user interface (GUI) agents send instructions and screenshots to remote planners, exposing personally identifiable information (PII). Existing placeholder defenses mask detected spans and reuse exact strings. Fragmented or changed forms can expose PII or break entity continuity across turns. PrivacyGuard addresses this private reference problem with source-linked sanitization (SAN) and a structured reference lifecycle (REF). SAN reconstructs fragmented or overlapping evidence and rewrites the complete source extent. REF links changed appearances to session entities and supports locally resolvable references without returning raw values to the upload path. In five-turn sessions scored against common gold, PrivacyGuard reduces audited source-value leakage by 33.55\%--82.60\% relative to the closest prior full-pipeline system in all four reader--language combinations. REF raises strict cross-form continuity from 0.04 under exact-value reuse to 0.81. In reference-required Android tasks, verified completion rises from 8/30 and 7/30 to 17/30 and 20/30. On AndroidLab, PrivacyGuard achieves the highest or tied-highest task-success point estimate among protection pipelines for all four planners. On PrivScreen, it reduces fixed-reader field exposure by 62.72\%--62.84\% relative to the same baseline. PrivacyGuard lowers upload exposure while preserving a locally resolvable task path.
Create a lesson
Related papers
CWE-bench: Refocusing Agentic Security from Exploitation to Defense
Gonzalo Gonzalez-Pumariega, Ajay Jagannath, Adit Jain et al.
Security and robustness analysis of encrypted communications for local home automation: ESPHome Native API and MQTT over TLS 1.3
Carlos Roberto Gonzalez-Escarpeta, Juan Eduardo Garcia Sanchez, José Antonio Hernandez Reyes et al.
Toward Realistic Privacy Risk in FL-Enabled OT Intrusion Detection Systems
Assunta Imperatrice, Luigi ROmano
Beyond User-Side Defenses: Image Protection in Open-Weight Diffusion Models
Jihye Kang, Jiwon Park, Namhyuk Ahn
StableJudge: A Lightweight Adversarial Defense Framework for LLM-as-a-Judge
Anning Hu, Yangyu Li, Yuqi Huang et al.
Logit Leakage through Floating-Point Bias Addition
Shaomin Chen, Xiaorui Gong