Metraipolis › archive › dc_ed1601255ddd

Bounded Agents: Delegation Security for Multi-Agent AI Systems — arXiv:2608.15888

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.15888
License (as archived)
CC0-1.0
Archived
2026-10-18T17:34:35.063Z
SHA-256 of the archived text
1ae49b6607d9cd1013f50071632ae02507e5a68be57dfaa9bca1ac8bc3672352
Stable URL of this page
https://metraipolis.com/documento/dc_ed1601255ddd

How to cite

Bounded Agents: Delegation Security for Multi-Agent AI Systems — arXiv:2608.15888 — arXiv (en), https://arxiv.org/abs/2608.15888 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ed1601255ddd (Metraipolis archive dc_ed1601255ddd, sha256 1ae49b6607d9cd1013f50071632ae02507e5a68be57dfaa9bca1ac8bc3672352)
BibTeX
@misc{dced1601255ddd,
  title = {Bounded Agents: Delegation Security for Multi-Agent AI Systems — arXiv:2608.15888},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.15888},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ed1601255ddd; sha256 1ae49b6607d9cd1013f50071632ae02507e5a68be57dfaa9bca1ac8bc3672352},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.15888. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

LLM-based agents can act on behalf of a user to access cloud services, call tools, or invoke agents. At session start, the agent's permissions are set but remain static, and each request is evaluated independently, without considering prior actions. Within its permissions, an agent may act contrary to the delegated task, combine individually permitted actions into a prohibited outcome, or delegate authority to a sub-agent without limiting it. A prompt injection poses a risk only if the agent has authority to perform such actions; this is therefore a problem of authorization architecture, not just the model. The Agentic Principal Chain (APC) tracks delegated authority from one principal to the next. APC evaluates each request against the accumulated session state using six authorization checks. APC carries forward and restricts delegated scope and budgets. Using composition closure, APC checks requests against prior actions to prevent prohibited combinations and enforces the decision outside the model. We prove Blast Radius Monotonicity and Composition Soundness for APC implementations; Composition Soundness is limited to prohibited combinations under a complete restriction set and serialized admission. We evaluated 3,154 instances including InjecAgent, AgentDojo, and ASB. Our compromised-model evaluation tests APC independently of model behavior by inserting the ground-truth attack call after the first legitimate tool call. AgentDojo exfiltration fell from 75-100% to 0% across all four domains; APC blocked all 544 InjecAgent data-stealing cases. Intent binding reduced destruction from 38.6% to 4.0% and manipulation from 90.5% to 12.1%. Authorization latency was 0.24 ms at the 99th percentile on an idle host; across 949 AgentDojo task-injection pairs, utility was 8.6 and 13.9 percentage points lower in the two settings. Implementation, evaluation tools, and data are publicly available.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2608.15888 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T16:29:13.838Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).