{"slug": "encryptedllm-privacy-preserving-large-language-model-inference", "title": "EncryptedLLM: Privacy-Preserving Large Language Model Inference", "summary": "A new paper titled \"EncryptedLLM: Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption\" presents a GPU-accelerated implementation of fully homomorphic encryption (FHE) that lets users send sensitive queries to cloud-hosted large language models and receive outputs without the cloud learning anything about their data. The work also develops methods to evaluate LLMs under FHE while preserving the quality of the model outputs, targeting sensitive domains such as healthcare and finance. The technical contribution centers on speeding up FHE-based private inference, which the authors describe as previously inefficient.", "body_md": "# EncryptedLLM: Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption\n\n### Abstract\n\n### Lay Summary\n\nLarge language models (LLMs) are typically deployed in cloud environments. To use these models, the user's data must be sent to an external cloud machine. For sensitive queries (e.g., topics related to healthcare or finance), this represents a major security concern. This work improves the efficiency of techniques to privately evaluate models over sensitive queries. This allows users to safely send their query to a cloud machine and receive the model output without allowing the cloud to learn anything about their data. The main underlying tool is an advanced cryptography primitive called fully homomorphic encryption (FHE), and a technical contribution of this work is a new GPU-accelerated implementation of FHE. We also develop methods to evaluate LLMs using FHE while preserving the quality of the model outputs.", "url": "https://wpnews.pro/news/encryptedllm-privacy-preserving-large-language-model-inference", "canonical_source": "https://ICML.cc/virtual/2025/poster/45395", "published_at": "2026-09-22 07:53:37+00:00", "updated_at": "2026-09-22 08:24:07.640273+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-safety", "ai-research", "ai-infrastructure"], "entities": ["EncryptedLLM", "fully homomorphic encryption", "large language models"], "alternates": {"html": "https://wpnews.pro/news/encryptedllm-privacy-preserving-large-language-model-inference", "markdown": "https://wpnews.pro/news/encryptedllm-privacy-preserving-large-language-model-inference.md", "text": "https://wpnews.pro/news/encryptedllm-privacy-preserving-large-language-model-inference.txt", "jsonld": "https://wpnews.pro/news/encryptedllm-privacy-preserving-large-language-model-inference.jsonld"}}