Researchers propose A2H protocol to integrate humans into AI agent ecosystems
New protocol lets AI agents discover and contact humans across platforms using standardized 'Human Cards'.
Deep Dive
Researchers led by Zhiyuan Liang propose the A2H (Agent-to-Human) protocol, a foundational standard for integrating humans into AI agent systems. It introduces three components: Human Cards for identity registration, a Formal Communication Schema, and a Unified Messaging Abstraction. This allows autonomous AI agents to discover, address, and interact with humans as resolvable entities across different messaging platforms, moving agents from isolated systems to human-connected infrastructures.
Why It Matters
Enables seamless human-in-the-loop workflows for complex AI agent tasks, improving oversight and collaboration.