SecureInfer: Heterogeneous TEE-GPU Architecture for Privacy-Critical Tensors for Large Language Model Deployment
- 18 December 2025
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 1, 1-7
Abstract
With the increasing deployment of Large Language Models (LLMs) on mobile and edge platforms, securing them against model extraction attacks has become a pressing concern. However, protecting model privacy without sacrificing the performance benefits of untrusted AI accelerators, such as GPUs, presents a challenging trade-off. In this paper, we initiate the study of high-performance execution on LLMs and present SecureInfer, a hybrid framework that leverages a heterogeneous Trusted Execution Environments (TEEs)-GPU architecture to isolate privacy-critical components while offloading compute-intensive operations to untrusted accelerators. Building upon an outsourcing scheme, SecureInfer adopts an information-theoretic and threat-informed partitioning strategy: security-sensitive components, including non-linear layers, projection of attention head, FNN transformations, and LoRA adapters are executed inside an SGX enclave, while other linear operations (matrix multiplication) are performed on the GPU after encryption and are securely restored within the enclave. We implement a prototype of SecureInfer using the LLaMA-2 model and evaluate it across performance and security metrics. Our results show that SecureInfer offers strong security guarantees with reasonable performance, offering a practical solution for secure on-device model inference.Keywords
All Related Versions
Funding Information
- US National Science Foundation (2419843)
This publication has 9 references indexed in Scilit:
- Security and Privacy Challenges of Large Language Models: A SurveyACM Computing Surveys, 2025
- “Yes, My LoRD.” Guiding Language Model Extraction with Locality Reinforced Distillation2025
- TMI! Finetuned Models Leak Private Information from their Pretraining DataProceedings on Privacy Enhancing Technologies, 2024
- No Privacy Left Outside: On the (In-)Security of TEE-Shielded DNN Partition for On-Device ML2024
- Efficient Memory Management for Large Language Model Serving with PagedAttention2023
- ShadowNet: A Secure and Efficient On-device Model Inference System for Convolutional Neural Networks2023
- DarkneTZ2020
- Serdab: An IoT Framework for Partitioning Neural Networks Computation across Multiple Enclaves2020
- Knockoff Nets: Stealing Functionality of Black-Box Models2019