Semantic Encryption: Secure and Effective Interaction with Cloud-based Large Language Models via Semantic Transformation

2025-08-03

AI Summary

To address the challenge of balancing privacy preservation and data utility in interactions with cloud-based large language models (CLLMs), this paper proposes a semantics-preserving privacy protection framework. Unlike existing approaches that merely encrypt sensitive fields, our framework introduces a novel semantic encoding–decoding mechanism: a lightweight local model performs structure-aware semantic transformation of user inputs, effectively concealing sensitive information while fully preserving logical structure and intent. Integrated end-to-end into a “local encoding → cloud inference → local decoding” pipeline, it significantly mitigates privacy leakage risks. Experimental evaluations across multiple benchmark datasets demonstrate that our method consistently outperforms InferDPT, achieving superior trade-offs among privacy guarantees, CLLM response quality, and user experience.

Technology Category

• Machine Learning: Privacy
• Natural Language Processing: (Large) Language Models
• Search and Optimization: Local Search

Application Category

• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

Abstract

The increasing adoption of Cloud-based Large Language Models (CLLMs) has raised significant concerns regarding data privacy during user interactions. While existing approaches primarily focus on encrypting sensitive information, they often overlook the logical structure of user inputs. This oversight can lead to reduced data utility and degraded performance of CLLMs. To address these limitations and enable secure yet effective interactions, we propose Semantic Encryption (SE)-a plug-and-play framework designed to preserve both privacy and utility. SE consists of two key components: Semantic Encoding and Semantic Decoding. In the encoding phase, a lightweight local model transforms the original user input into an alternative semantic context that maintains the original intent and logical structure while obfuscating sensitive information. This transformed input is then processed by the CLLM, which generates a response based on the transformed semantic context. To maintain a seamless user experience, the decoding phase will reconstruct the CLLM's response back into the original semantic context by referencing the locally stored user input. Extensive experimental evaluations demonstrate that SE effectively protects data privacy without compromising data utility or user experience, offering a practical solution for secure interaction with CLLMs. Particularly, the proposed SE demonstrates a significant improvement over the state-of-the-art InferDPT, surpassing it across various evaluated metrics and datasets.

Problem

• Ensuring data privacy in cloud-based LLM interactions
• Preserving logical structure of encrypted user inputs
• Maintaining model performance while protecting sensitive information

Innovation

• Semantic Encryption framework preserves privacy and utility
• Lightweight local model transforms input for secure processing
• Decoding phase reconstructs responses to original semantic context