Metraipolis › archive › dc_82eded9f18fe

Session-Bound Behavioral Watermarking for Autonomous LLM Agents — arXiv:2610.38579

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.38579
License (as archived): CC0-1.0
Archived: 2026-11-01T02:21:11.189Z
SHA-256 of the archived text: 25b887b4339be4b1a4841e5281db42a91522803e7626910f3e9451ac4ce3d80f
Stable URL of this page: https://metraipolis.com/documento/dc_82eded9f18fe

How to cite

Session-Bound Behavioral Watermarking for Autonomous LLM Agents — arXiv:2610.38579 — arXiv (en), https://arxiv.org/abs/2610.38579 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_82eded9f18fe (Metraipolis archive dc_82eded9f18fe, sha256 25b887b4339be4b1a4841e5281db42a91522803e7626910f3e9451ac4ce3d80f)

BibTeX

@misc{dc82eded9f18fe,
  title = {Session-Bound Behavioral Watermarking for Autonomous LLM Agents — arXiv:2610.38579},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.38579},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_82eded9f18fe; sha256 25b887b4339be4b1a4841e5281db42a91522803e7626910f3e9451ac4ce3d80f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.38579. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large Language Model (LLM) agents are increasingly acting through Web interfaces and external tools. They produce trajectories. Those trajectories may pass through several intermediaries. They can store, modify, or dispute those trajectories. In Web-based agent deployments, trajectories may pass through APIs, marketplaces, and tool servers. As such, a copied or disputed log may not by itself reveal which agent owner produced it. Behavioral watermarking can embed ownership evidence in agent action choices. However, existing methods mainly address detection, removal, rewriting, or imitation. They do not directly test whether repeated observation of watermarked trajectories enables cross-session forgery by exploiting learned action preferences. We formulate this learned-forgery threat and propose AgentAnchor, a session-keyed behavioral watermark for owner attribution from trajectory logs. We derive each step key from a secret key, a fresh public session identifier, and semantic states of recent observations, while a guarded digest of the action-policy state diversifies linear coding rows. It embeds parity evidence through a KL-bounded exponential tilt, and a trusted verifier checks a disputed log without rerunning the agent, using wrong-key calibration and abstention.

Tema degli esploratori: fiducia digitale e crittografia (elenco arXiv cs.CR, nuovi lavori del giorno).
Categorie arXiv: Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2610.38579 — 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 2026-11-01T02:12:13.371Z dalla pagina https://arxiv.org/list/cs.CR/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).