Large Language Models (LLMs) are increasingly being explored for cybersecurity applications, including malware analysis, vulnerability discovery, and mobile-security assessment. This work investigates the practical feasibility of using local agentic language models for Android security analysis, with a focus on privacy-preserving and resource-efficient deployment scenarios.
The research evaluates whether locally executed LLM-based agents can effectively support Android application analysis tasks without relying on external cloud infrastructures. The study analyzes the trade-offs between computational efficiency, autonomy, reasoning capabilities, and security-analysis performance when deploying compact language models in constrained environments.
The work contributes to the emerging field of AI-assisted cybersecurity by exploring trustworthy and privacy-aware approaches for integrating agentic LLMs into mobile-security analysis workflows.