Measuring the Security of Mobile LLM Agents under Adversarial Prompts from Untrusted Third-Party Channels
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917063999619072 |
|---|---|
| author | Du, Chenghao Huang, Quanfeng Tang, Tingxuan Wang, Zihao Nadkarni, Adwait Xiao, Yue |
| author_facet | Du, Chenghao Huang, Quanfeng Tang, Tingxuan Wang, Zihao Nadkarni, Adwait Xiao, Yue |
| contents | Large Language Models (LLMs) have transformed software development, enabling AI-powered applications known as LLM-based agents that promise to automate tasks across diverse apps and workflows. Yet, the security implications of deploying such agents in adversarial mobile environments remain poorly understood. In this paper, we present the first systematic study of security risks in mobile LLM agents. We design and evaluate a suite of adversarial case studies, ranging from opportunistic manipulations such as pop-up advertisements to advanced, end-to-end workflows involving malware installation and cross-app data exfiltration. Our evaluation covers eight state-of-the-art mobile agents across three architectures, with over 2,000 adversarial and paired benign trials. The results reveal systemic vulnerabilities: low-barrier vectors such as fraudulent ads succeed with over 80% reliability, while even workflows requiring the circumvention of operating-system warnings, such as malware installation, are consistently completed by advanced multi-app agents. By mapping these attacks to the MITRE ATT&CK Mobile framework, we uncover novel privilege-escalation and persistence pathways unique to LLM-driven automation. Collectively, our findings provide the first end-to-end evidence that mobile LLM agents are exploitable in realistic adversarial settings, where untrusted third-party channels (e.g., ads, embedded webviews, cross-app notifications) are an inherent part of the mobile ecosystem. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_27140 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Measuring the Security of Mobile LLM Agents under Adversarial Prompts from Untrusted Third-Party Channels Du, Chenghao Huang, Quanfeng Tang, Tingxuan Wang, Zihao Nadkarni, Adwait Xiao, Yue Cryptography and Security Large Language Models (LLMs) have transformed software development, enabling AI-powered applications known as LLM-based agents that promise to automate tasks across diverse apps and workflows. Yet, the security implications of deploying such agents in adversarial mobile environments remain poorly understood. In this paper, we present the first systematic study of security risks in mobile LLM agents. We design and evaluate a suite of adversarial case studies, ranging from opportunistic manipulations such as pop-up advertisements to advanced, end-to-end workflows involving malware installation and cross-app data exfiltration. Our evaluation covers eight state-of-the-art mobile agents across three architectures, with over 2,000 adversarial and paired benign trials. The results reveal systemic vulnerabilities: low-barrier vectors such as fraudulent ads succeed with over 80% reliability, while even workflows requiring the circumvention of operating-system warnings, such as malware installation, are consistently completed by advanced multi-app agents. By mapping these attacks to the MITRE ATT&CK Mobile framework, we uncover novel privilege-escalation and persistence pathways unique to LLM-driven automation. Collectively, our findings provide the first end-to-end evidence that mobile LLM agents are exploitable in realistic adversarial settings, where untrusted third-party channels (e.g., ads, embedded webviews, cross-app notifications) are an inherent part of the mobile ecosystem. |
| title | Measuring the Security of Mobile LLM Agents under Adversarial Prompts from Untrusted Third-Party Channels |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2510.27140 |