Metraipolis › archive › dc_21c8fd0aafe7

Efficient Auditing of Adversarial AI Agent Behavior from Agent Traces — arXiv:2610.07256

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.07256
License (as archived)
CC0-1.0
Archived
2026-10-07T11:35:01.751Z
SHA-256 of the archived text
8d9cefab6f273600bba8271b4222f59845396335ee389258dde541b86312cf0f
Stable URL of this page
https://metraipolis.com/documento/dc_21c8fd0aafe7

How to cite

Efficient Auditing of Adversarial AI Agent Behavior from Agent Traces — arXiv:2610.07256 — arXiv (en), https://arxiv.org/abs/2610.07256 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_21c8fd0aafe7 (Metraipolis archive dc_21c8fd0aafe7, sha256 8d9cefab6f273600bba8271b4222f59845396335ee389258dde541b86312cf0f)
BibTeX
@misc{dc21c8fd0aafe7,
  title = {Efficient Auditing of Adversarial AI Agent Behavior from Agent Traces — arXiv:2610.07256},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.07256},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_21c8fd0aafe7; sha256 8d9cefab6f273600bba8271b4222f59845396335ee389258dde541b86312cf0f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07256. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

AI agents powered by large language models (LLMs) can perform complex tasks but may harm the systems they operate in, either intentionally or unintentionally. Existing agent monitoring approaches rely on rule-based guardrails or LLM-based trace auditing. However, rule-based guardrails can be bypassed through obfuscation and may miss harmful actions beyond their predefined rules, whereas applying an LLM to audit every action is costly. We present a two-stage agent trace auditing framework. The first stage uses single-event and trace-sequence rules to select pending actions for inspection; the second uses an LLM audit agent to examine each selected action in the context of the agent's preceding trace before execution. We jointly refine the gate rules and audit instructions using training data, allowing the framework to adapt to complex agent behaviors rather than relying solely on predefined rules. On the public benchmark OpenAgentSafety, our framework reduces the average number of LLM audits from 8.15 to 2.33 per run and token usage from 47.8k to 14.6k, with a detection rate of 72.8\% compared with 81.5\% when every action is audited. In two simulated multi-agent case studies, the framework flags all malicious traces while reducing audit token usage by more than 80\%.

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/2610.07256 — 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 2026-10-07T11:18:39.657Z dalla pagina https://arxiv.org/list/cs.CR/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).