From Thinker to Society: Security in Hierarchical Autonomy Evolution of AI Agents
This work identifies and categorizes critical security vulnerabilities in evolving AI agent systems, providing a structured framework for researchers and developers to build more trustworthy AI.
This paper addresses the security vulnerabilities introduced by the evolution of AI agents from passive tools to active, autonomous entities, particularly those driven by Large Language Models (LLMs). It proposes a Hierarchical Autonomy Evolution (HAE) framework to categorize agent security into three tiers: Cognitive Autonomy (L1), Execution Autonomy (L2), and Collective Autonomy (L3).
Artificial Intelligence (AI) agents have evolved from passive predictive tools into active entities capable of autonomous decision-making and environmental interaction, driven by the reasoning capabilities of Large Language Models (LLMs). However, this evolution has introduced critical security vulnerabilities that existing frameworks fail to address. The Hierarchical Autonomy Evolution (HAE) framework organizes agent security into three tiers: Cognitive Autonomy (L1) targets internal reasoning integrity; Execution Autonomy (L2) covers tool-mediated environmental interaction; Collective Autonomy (L3) addresses systemic risks in multi-agent ecosystems. We present a taxonomy of threats spanning cognitive manipulation, physical environment disruption, and multi-agent systemic failures, and evaluate existing defenses while identifying key research gaps. The findings aim to guide the development of multilayered, autonomy-aware defense architectures for trustworthy AI agent systems.