LDP: An Identity-Aware Protocol for Multi-Agent LLM Systems

Researchers present LLM Delegate Protocol (LDP), a new AI-native communication protocol for multi-agent LLM systems that introduces identity awareness, progressive payloads, and governance mechanisms. The protocol achieves 12x lower latency on simple tasks and 37% token reduction compared to existing protocols like A2A, though quality improvements remain limited in small delegate pools.

• →LDP introduces five key mechanisms including delegate identity cards, progressive payload modes, and governed sessions for multi-agent AI systems.
• →Identity-aware routing achieves approximately 12x lower latency on easy tasks through specialized delegate selection.
• →Semantic frame payloads reduce token count by 37% with no quality loss, improving efficiency.
• →The protocol shows superior attack detection (96% vs 6%) and failure recovery (100% vs 35%) in simulated environments.
• →Noisy provenance metadata actually degrades synthesis quality, suggesting confidence data needs verification to be useful.