Metraipolis › archive › dc_e2092ea19d0b

Will the User Ever Know? Covert Indirect Prompt Injection Attacks on Tool-Using LLM Agents — arXiv:2608.30362

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.30362
License (as archived)
CC0-1.0
Archived
2026-10-17T07:44:37.677Z
SHA-256 of the archived text
d0dfda08f203fcc9115e701440bad2981ab1357fda9fdc38028499562e84a80a
Stable URL of this page
https://metraipolis.com/documento/dc_e2092ea19d0b

How to cite

Will the User Ever Know? Covert Indirect Prompt Injection Attacks on Tool-Using LLM Agents — arXiv:2608.30362 — arXiv (en), https://arxiv.org/abs/2608.30362 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e2092ea19d0b (Metraipolis archive dc_e2092ea19d0b, sha256 d0dfda08f203fcc9115e701440bad2981ab1357fda9fdc38028499562e84a80a)
BibTeX
@misc{dce2092ea19d0b,
  title = {Will the User Ever Know? Covert Indirect Prompt Injection Attacks on Tool-Using LLM Agents — arXiv:2608.30362},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.30362},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e2092ea19d0b; sha256 d0dfda08f203fcc9115e701440bad2981ab1357fda9fdc38028499562e84a80a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.30362. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As LLM agents take real-world actions through tools, indirect prompt injection (IPI) has emerged as a serious threat. The standard metric, Attack Success Rate (ASR), counts whether an injection succeeds but ignores what the user notices in the agent's final response. Looking at successful injection traces, we find two distinct outcomes: the agent executes the injection while returning an otherwise normal response, or reports the injected action in its final response, giving the user a chance to notice. We call these covert and overt successes. From the user's perspective, we decompose ASR into the Covert Success Rate (CSR), counting successes leaving no trace in the final response, and the Overt Success Rate (OSR), counting successes the user can detect. To understand what drives the gap, we analyze successful trajectories and find that the agent's behavior after the injection separates covert from overt: covert traces hand control back to the user task before ending, while overt traces end at the attack itself. This split follows from the ReAct format, where the final response summarizes the most recent action. Building on this observation, we propose ICoA (Induced Covert Attack), an IPI attack designed to induce covert outcomes by steering the agent back to the user task after executing the injection. Across four target models on AgentDojo, ICoA achieves the highest CSR, with gains of 3.79-12.01 percentage points over the strongest baseline.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.30362 — 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 2026-10-17T07:28:26.598Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).