Metraipolis › archive › dc_a22ff44e9cf0

Breadcrumbing Search Agents: Per-Turn Scheming Over Long-Horizon Trajectories — arXiv:2608.04565

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.04565
License (as archived)
CC0-1.0
Archived
2026-10-31T11:14:26.328Z
SHA-256 of the archived text
5486cfac651345994c0c263bf46a0cd4fff27f43aa63a84e28d0f9458c01c344
Stable URL of this page
https://metraipolis.com/documento/dc_a22ff44e9cf0

How to cite

Breadcrumbing Search Agents: Per-Turn Scheming Over Long-Horizon Trajectories — arXiv:2608.04565 — arXiv (en), https://arxiv.org/abs/2608.04565 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a22ff44e9cf0 (Metraipolis archive dc_a22ff44e9cf0, sha256 5486cfac651345994c0c263bf46a0cd4fff27f43aa63a84e28d0f9458c01c344)
BibTeX
@misc{dca22ff44e9cf0,
  title = {Breadcrumbing Search Agents: Per-Turn Scheming Over Long-Horizon Trajectories — arXiv:2608.04565},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.04565},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a22ff44e9cf0; sha256 5486cfac651345994c0c263bf46a0cd4fff27f43aa63a84e28d0f9458c01c344},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.04565. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

LLM-based search agents are widely used for information-seeking tasks, but their reliance on external tool returns introduces a critical security risk: web content retrieved during execution is untrusted, exposing agents to prompt injection and goal hijacking. Prior work on search-agent safety primarily focuses on static web-content injection, but modern agents issue follow-up queries and cross-check competing sources, so a single injected page is often diluted or rejected. We show that the channel delivering search and page observations is a fragile security boundary: beyond exposing the agent to a single poisoned page, a mediated search interface can repeatedly steer how the agent gathers evidence and forms its final answer. Under a constrained tool-intermediary threat model, appending only one controlled result per query can substantially increase attack success when the evidence is coordinated across the agent's trajectory. We study this setting with a strategy-driven long-horizon attack system and introduce Authority-Chain Hijack (ACH), an expert-refined strategy that turns isolated search-result and page-content manipulations into a coherent evidence chain across seemingly corroborating sources. ACH achieves the highest Overall ASR among all baselines, reaching 55.9% / 83.3% ASR / MaxN ASR on the full SafeSearch test split. We further introduce Trace-Guided Strategy Evolution (TGSE), which automatically improves attacker strategies from execution traces, replacing manual redesign with trace-driven refinement; its strongest single setting reaches 71.4% / 95.0% in held-out evaluation. Code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-06 in cs.AI).
Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.04565 — 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 2026-10-31T11:02:13.527Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-06?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).