MobileInject: Evaluating Reasoning-to-Action Security of MLLM-Based Mobile Agents under Cross-Modal Consistency Injection

Summary

MobileInject is a benchmark for testing MLLM-based mobile agents against multimodal injection attacks, where malicious objectives are embedded in cross-application content. It covers 11 scenarios, 9 security-risk categories and 108 executable risk instances, and adds Cross-Modal Consistency Injection plus two attack methods, Multi-Role Collaborative Deception and Reasoning-Camouflage Injection. The authors report that mobile agents are broadly vulnerable, with many failures arising from reasoning hijacking at early stages.

Related items

• MediumCVE-2026-95183: DeepSeek Harness context-compaction prompt injection via external contentSimilar attack · NVD/CVE Database
• InfoIgnored, Not Refused: What Low Prompt-Injection Success Rates Hide in Small Tool-Using AgentsSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• InfoSAFE-Mem: Secure Agentic Memory with Fine-Grained Authority and Governance-Aware RetrievalSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• InfoIdentify, Classify, Act: Per-Instruction Controlled Compliance for LLMs over External DataSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)
• InfoGRAD: Graph-Structured Antibody Defense for LLM AgentsSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)