Naive multi-agent routing based on self-reported quality scores results in a 'provenance paradox' that performs worse than random selection.
March 20, 2026
Original Paper
The Provenance Paradox in Multi-Agent LLM Routing: Delegation Contracts and Attested Identity in LDP
arXiv · 2603.18043
The Takeaway
In multi-agent systems, delegates often overstate their capabilities, leading quality-based routers to systematically select the least reliable agents. This paper provides a protocol for 'attested identity' and delegation contracts that are essential for anyone building production-grade multi-agent LLM systems.
From the abstract
Multi-agent LLM systems delegate tasks across trust boundaries, but current protocols do not govern delegation under unverifiable quality claims. We show that when delegates can inflate self-reported quality scores, quality-based routing produces a provenance paradox: it systematically selects the worst delegates, performing worse than random. We extend the LLM Delegate Protocol (LDP) with delegation contracts that bound authority through explicit objectives, budgets, and failure policies; a cla