Cross-Cloud Data Privacy Protection: Optimizing Collaborative Mechanisms of AI Systems by Integrating Federated Learning and LLMs
Huaiying Luo, Cheng Ji
TL;DR
The paper tackles privacy-preserving cross-cloud AI training by integrating Federated Learning with Large Language Models (LLMs). It proposes a cross-cloud FL framework where local updates are aggregated without exposing data, augmented by LLM-derived contextual features and reinforced by a secure aggregation layer, including $z_{i,n} = LLM(x_{i,n})$ and $w_hat_i = Enc(w_i)$ with $w^{(t+1)} = Dec(\sum_{i=1}^K (N_i/N) w_hat_i)$. Key contributions include the LLM-augmented local objective and encrypted aggregation to protect data privacy during cross-cloud collaboration. Experimental results on a Google Cloud BigQuery-like dataset show the proposed method outperforms FedAvg, DP-FL, SMC-FL, and HE-FL in accuracy, convergence speed, and privacy protection, demonstrating practical impact for scalable, privacy-preserving cross-cloud AI systems.
Abstract
In the age of cloud computing, data privacy protection has become a major challenge, especially when sharing sensitive data across cloud environments. However, how to optimize collaboration across cloud environments remains an unresolved problem. In this paper, we combine federated learning with large-scale language models to optimize the collaborative mechanism of AI systems. Based on the existing federated learning framework, we introduce a cross-cloud architecture in which federated learning works by aggregating model updates from decentralized nodes without exposing the original data. At the same time, combined with large-scale language models, its powerful context and semantic understanding capabilities are used to improve model training efficiency and decision-making ability. We've further innovated by introducing a secure communication layer to ensure the privacy and integrity of model updates and training data. The model enables continuous model adaptation and fine-tuning across different cloud environments while protecting sensitive data. Experimental results show that the proposed method is significantly better than the traditional federated learning model in terms of accuracy, convergence speed and data privacy protection.