Metraipolis › archive › dc_9795fb87e35c

Who Else Is In? Multi-Agent Collusion Initation Via Strategic Vagueness — arXiv:2610.37984

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.37984
License (as archived)
CC0-1.0
Archived
2026-10-31T06:03:18.536Z
SHA-256 of the archived text
fb0739e86f221e7cf46a4e53a5fdf88e2ea9402c2adb958c088b1be587bbc3a9
Stable URL of this page
https://metraipolis.com/documento/dc_9795fb87e35c

How to cite

Who Else Is In? Multi-Agent Collusion Initation Via Strategic Vagueness — arXiv:2610.37984 — arXiv (en), https://arxiv.org/abs/2610.37984 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9795fb87e35c (Metraipolis archive dc_9795fb87e35c, sha256 fb0739e86f221e7cf46a4e53a5fdf88e2ea9402c2adb958c088b1be587bbc3a9)
BibTeX
@misc{dc9795fb87e35c,
  title = {Who Else Is In? Multi-Agent Collusion Initation Via Strategic Vagueness — arXiv:2610.37984},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.37984},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9795fb87e35c; sha256 fb0739e86f221e7cf46a4e53a5fdf88e2ea9402c2adb958c088b1be587bbc3a9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.37984. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Multi agent LLM systems raise the risk of collusion among agents pursuing objectives hidden from their overseers. When potential partners identities are unknown, collusion first requires initiation: agents must discover one another, often through a public channel observed by ordinary agents and monitors. Strategic vagueness offers a way to navigate this mixed audience, conveying intent to a prospective partner while preserving a plausible ordinary interpretation. We introduce a mutual identification game in which two schemers must identify each other through public interaction, while jointly solving an ordinary task. We find that some models discover successful handshake strategies without explicit signalling guidance, using task compatible exchanges that monitors often fail to flag at low false positive rates. To characterize the communicative capabilities underlying these exchanges, we study a controlled coordination setting. We find an asymmetric response to message explicitness: as messages become more indirect, monitor punishment can fall faster than partner coordination, creating a profitable intermediate regime of strategic vagueness. Together, these findings illustrate a Blackstone style evidential asymmetry: communication sufficient for coordination between agents may remain insufficient for confident detection by monitor, opening a new surface of collusion.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno).
Categorie arXiv: Multiagent Systems (cs.MA).
Pagina del lavoro: https://arxiv.org/abs/2610.37984 — 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 2026-10-31T06:03:01.129Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).