Metraipolis › archive › dc_d402702fc125

Inference-Layer Security: Defending Against Adversarial Inference and Infrastructure Abuse — arXiv:2609.38239

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.38239
License (as archived)
CC0-1.0
Archived
2026-10-07T13:43:03.574Z
SHA-256 of the archived text
d3f7793813c70adcc7015bc8d82cba4f24402b3b37b9fa976580d52b35117751
Stable URL of this page
https://metraipolis.com/documento/dc_d402702fc125

How to cite

Inference-Layer Security: Defending Against Adversarial Inference and Infrastructure Abuse — arXiv:2609.38239 — arXiv (en), https://arxiv.org/abs/2609.38239 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d402702fc125 (Metraipolis archive dc_d402702fc125, sha256 d3f7793813c70adcc7015bc8d82cba4f24402b3b37b9fa976580d52b35117751)
BibTeX
@misc{dcd402702fc125,
  title = {Inference-Layer Security: Defending Against Adversarial Inference and Infrastructure Abuse — arXiv:2609.38239},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.38239},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d402702fc125; sha256 d3f7793813c70adcc7015bc8d82cba4f24402b3b37b9fa976580d52b35117751},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.38239. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A Technical Report: Operating a large language model (LLM) as a service requires more than inference infrastructure: the provider must also defend against adversarial interactions that seek to exploit the service, including jailbreaking for harmful use, sophisticated denial of service, and distillation attacks. We study this problem at the inference layer, using a hypothetical frontier lab, Five Elements Inc., as a running example. Because no public labelled dataset of adversarial LLM usage exists, we introduce a structural causal model (SCM) that generates a realistically grounded, labelled dataset of user-sessions, with coordinated multi-account campaigns, platform feedback, and three tiers of label observability. On this dataset we train a practical gradient-boosted detector that classifies each user-session as benign or malicious and, if malicious, by attack type. Against oracle labels the detector very nearly solves the binary task (AUPRC $0.993$), yet against the operational labels a real Trust & Safety team would hold, the same model scores an AUPRC of only $0.313$: the detector is more accurate than the labels used to evaluate it. For attack-type attribution, a naive argmax is dominated by the $98\%$ benign prior (macro-F1 $0.295$), whereas a simple thresholded decision engine raises macro-F1 to $0.489$ without sacrificing accuracy. The dataset is publicly released.

Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CR).
Categorie arXiv: Cryptography and Security (cs.CR) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.38239 — 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 2026-10-07T13:16:25.183Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).