Metraipolis › archive › dc_41e7cb63567b

Trust-Gated Capability Control: Breaking the Trust-Vulnerability Paradox in Multi-Agent LLM Systems — arXiv:2610.07000

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.07000
License (as archived)
CC0-1.0
Archived
2026-10-07T11:17:16.066Z
SHA-256 of the archived text
a80be7d25710a92e41c9aeca501ffaa1f43f08662054d74e24b5a8ba33cc36b1
Stable URL of this page
https://metraipolis.com/documento/dc_41e7cb63567b

How to cite

Trust-Gated Capability Control: Breaking the Trust-Vulnerability Paradox in Multi-Agent LLM Systems — arXiv:2610.07000 — arXiv (en), https://arxiv.org/abs/2610.07000 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_41e7cb63567b (Metraipolis archive dc_41e7cb63567b, sha256 a80be7d25710a92e41c9aeca501ffaa1f43f08662054d74e24b5a8ba33cc36b1)
BibTeX
@misc{dc41e7cb63567b,
  title = {Trust-Gated Capability Control: Breaking the Trust-Vulnerability Paradox in Multi-Agent LLM Systems — arXiv:2610.07000},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.07000},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_41e7cb63567b; sha256 a80be7d25710a92e41c9aeca501ffaa1f43f08662054d74e24b5a8ba33cc36b1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07000. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Layered trust models for multi-agent LLM systems remain largely conceptual: they name which dimensions of trust matter but not how layers combine, how their importance is set at runtime, or how trust should govern agent actions. This gap matters because higher inter-agent trust raises task success while also enlarging exposure to exploitation, a tension formalized as the Trust-Vulnerability Paradox. We make a five-layer trust stack operational through three contributions. First, a cross-layer synergy operator propagates prerequisite-layer deficits into de- pendent layers, feeding a generalized-mean composite trust that recovers the weakest-link rule as a limiting case, with provable bounds. Second, per-layer importance weights are grounded in observed failures via a no-regret online estimator that tracks which layer is currently most responsible for harm. Third, Trust- Gated Capability Control issues short-lived, revocable capability grants only when composite trust and the relevant prerequisite layers clear capability-specific thresholds. We prove this mechanism breaks the paradox: a stealthy compromise that inflates behavioral trust while degrading a prerequisite layer cannot escalate privilege, and give a closed-form, provably conservative trust fixed point with a bounded-latency revocation guarantee. A numerical study with a sleeper adversary confirms the predictions: composite-trust bounds hold across 20,000 random draws, the analytic fixed point matches simulation within 0.007, and a compromised prerequisite layer triggers automatic revocation within a few interactions.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno).
Categorie arXiv: Computer Science and Game Theory (cs.GT) ; Multiagent Systems (cs.MA).
Pagina del lavoro: https://arxiv.org/abs/2610.07000 — 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 2026-10-07T11:15:09.468Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).