This paper provides a comprehensive systematization of knowledge regarding the security of mobile on-device AI (MoAI) systems. It outlines the security pillars, attack vectors, and defense mechanisms relevant to MoAI, which integrates local AI models with mobile software for privacy-preserving and offline functionality. The research identifies current gaps in MoAI security and suggests future research directions to foster the development of more secure on-device AI applications. AI
IMPACT Establishes a foundational framework for understanding and improving the security of AI models deployed directly on mobile devices.
RANK_REASON This is a research paper providing a systematization of knowledge on a specific technical topic. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →