Metraipolis › archive › dc_deeed25fcd10

On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models — arXiv:2608.10530

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.10530
License (as archived): CC0-1.0
Archived: 2026-10-19T06:09:16.381Z
SHA-256 of the archived text: 2f4bced29cc435388334bfe4cbef88089fd4a4a0e0a3f81482300e38300d7d05
Stable URL of this page: https://metraipolis.com/documento/dc_deeed25fcd10

How to cite

On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models — arXiv:2608.10530 — arXiv (en), https://arxiv.org/abs/2608.10530 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_deeed25fcd10 (Metraipolis archive dc_deeed25fcd10, sha256 2f4bced29cc435388334bfe4cbef88089fd4a4a0e0a3f81482300e38300d7d05)

BibTeX

@misc{dcdeeed25fcd10,
  title = {On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models — arXiv:2608.10530},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.10530},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_deeed25fcd10; sha256 2f4bced29cc435388334bfe4cbef88089fd4a4a0e0a3f81482300e38300d7d05},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.10530. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large Language Models (LLMs) have undergone a shift from stateless conversational interfaces to autonomous agents capable of multi-step planning, tool invocation, code execution, and maintaining persistent memory. When these agents operate with real-world privileges---calling APIs, modifying files, and querying databases---a compromised reasoning step can trigger unauthorized data access, irreversible state changes, or cascading failures, yet the security research community has not kept pace. To quantify the state of the field, we conducted a systematic literature review under PRISMA 2020 guidelines across six databases, screening 743 records and retaining 85 papers (2023--2025) on agentic LLM security. Attack research outpaces defense work by 3.9:1. Perception-layer vulnerabilities (prompt injection, jailbreaking, adversarial perturbations) dominate, accounting for 66\% of papers, while action-layer vulnerabilities (tool misuse, code injection, sandbox escape) appear in only 4.7\%, misaligned with real-world risk. Code execution security accounts for 3.5\%, and tool-augmented agents 12\%. We contribute a four-layer taxonomy mapping 13 vulnerability types across perception, brain, action, and interaction layers, and identify seven open problems centered on containment. Agentic LLM insecurity stems from architectural coupling, where weak isolation allows vulnerabilities to propagate across layers.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-12 in cs.AI).
Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.10530 — 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 2026-10-19T05:57:28.381Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-12?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).