Delegation Without Trust: An Empirical Gap Analysis of Identity, Authorization, and Runtime Governance in Multi-Agent LLM Systems

Abstract

Autonomous LLM agents increasingly act on a user's behalf: they hold credentials, call tools and services, and spawn sub-agents that act further on their behalf. This turns a long-standing distributed-systems question -- who is authorized to do what, on whose authority -- into an urgent and largely unsolved problem, because the component driving each agent is a language model an adversary can hijack. We argue that agent security must be evaluated under an untrusted-model assumption: a correct sy

Categories

@misc{dantuluri2026delegation,
  title = {{Delegation Without Trust: An Empirical Gap Analysis of Identity, Authorization, and Runtime Governance in Multi-Agent LLM Systems}},
  author = {Panduranga Sai Varma Dantuluri and Jyotirmoy Sundi},
  year = {2026},
  month = aug,
  eprint = {2609.00267},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.00267}
}