Metraipolis › archive › dc_770b0b0fcd25
HE-Guardrail: A Homomorphic Guardrail Against Jailbreak Attacks for Encrypted Large Language Model Inference — arXiv:2609.21484
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.21484
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T01:08:31.692Z
- SHA-256 of the archived text
0dd4cb9812aad84b4b7bb100bba9776629f474bff1c31aa4a213024842295389- Stable URL of this page
https://metraipolis.com/documento/dc_770b0b0fcd25
How to cite
HE-Guardrail: A Homomorphic Guardrail Against Jailbreak Attacks for Encrypted Large Language Model Inference — arXiv:2609.21484 — arXiv (en), https://arxiv.org/abs/2609.21484 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_770b0b0fcd25 (Metraipolis archive dc_770b0b0fcd25, sha256 0dd4cb9812aad84b4b7bb100bba9776629f474bff1c31aa4a213024842295389)
BibTeX
@misc{dc770b0b0fcd25,
title = {HE-Guardrail: A Homomorphic Guardrail Against Jailbreak Attacks for Encrypted Large Language Model Inference — arXiv:2609.21484},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.21484},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_770b0b0fcd25; sha256 0dd4cb9812aad84b4b7bb100bba9776629f474bff1c31aa4a213024842295389},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21484. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Homomorphic encryption (HE) has emerged as a promising approach to privacy-preserving machine learning (PPML), enabling computation directly over encrypted data. In HE-based PPML, a client submits an encrypted input to the server, which evaluates models such as large language models (LLMs) without access to the underlying plaintext. However, we identify a critical security vulnerability in this setting: HE-LLM inference is vulnerable to malicious clients that submit adversarial prompts, such as jailbreak attacks. The same confidentiality that protects benign clients also prevents the server from inspecting incoming prompts or generated responses, making adversarial attempts difficult to detect or block and potentially allowing successful attacks to remain entirely invisible to the server. To address this vulnerability, we propose HE-Guardrail, a framework that evaluates guardrail mechanisms entirely over encrypted data and homomorphically controls whether the target-model response is returned to the client. We instantiate HE-Guardrail with three representative guardrails - Llama Guard, JBShield, and GradSafe. Our results show that HE-Guardrail closely reproduces the decisions of the corresponding plaintext guardrails in the encrypted domain, with distinct security-efficiency-utility trade-offs. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.AI). Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.21484 — 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 2026-10-10T00:23:42.541Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).