Metraipolis › archive › dc_20cd24bb575d

Connecting the Dots in Agentic AI Security: A Cross-Dimensional Threat Taxonomy, Evaluation Maturity, and Open Challenges — arXiv:2609.23894

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.23894
License (as archived): CC0-1.0
Archived: 2026-10-09T07:01:05.351Z
SHA-256 of the archived text: 2f4cd61362c6c0d79406a9b7055e5a26c507c1ae40cff4ef83cc71e6bc04ef9f
Stable URL of this page: https://metraipolis.com/documento/dc_20cd24bb575d

How to cite

Connecting the Dots in Agentic AI Security: A Cross-Dimensional Threat Taxonomy, Evaluation Maturity, and Open Challenges — arXiv:2609.23894 — arXiv (en), https://arxiv.org/abs/2609.23894 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_20cd24bb575d (Metraipolis archive dc_20cd24bb575d, sha256 2f4cd61362c6c0d79406a9b7055e5a26c507c1ae40cff4ef83cc71e6bc04ef9f)

BibTeX

@misc{dc20cd24bb575d,
  title = {Connecting the Dots in Agentic AI Security: A Cross-Dimensional Threat Taxonomy, Evaluation Maturity, and Open Challenges — arXiv:2609.23894},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.23894},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_20cd24bb575d; sha256 2f4cd61362c6c0d79406a9b7055e5a26c507c1ae40cff4ef83cc71e6bc04ef9f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.23894. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agentic AI extends LLM security beyond generated content to persistent state, autonomous actions, tool use, and interactions with humans and other agents. Existing threat classifications often emphasize individual dimensions, obscuring connections among entry points, affected components, and security consequences. The known threat landscape also differs from the coverage demonstrated by empirical research. Through a structured review of 66 studies published from 2022 to 2026, we introduce T={S, B, P, A}, a cross-dimensional representation linking affected functional or system surfaces {S}, interaction or trust boundaries {B}, violated security properties {P}, and empirically examined architectures {A}. We analyze 22 artifact-backed red-teaming studies and 11 representative security benchmarks to characterize empirical coverage and evaluation maturity. Within the selected studies, evidence concentrates on prompt/reasoning, memory, and tool-mediated attacks, predominantly in single-agent settings. Persistent, Human--Agent, complex multi-agent, systemic, and long-horizon threats receive less coverage. These findings describe the selected corpus rather than establish gaps across all empirical research. Heterogeneous metrics, limited adaptive defense evaluation, architectural imbalance, and incomplete execution-state capture further constrain comparison and reproducibility. We derive 13 open research questions to guide more systematic, architecture-aware, and reproducible security evaluation of agentic AI.

Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CR).
Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.23894 — 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 2026-10-09T06:56:00.986Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).