The Utilization of Homomorphic Encryption Technology Grounded on Artificial Intelligence for Privacy Preservation
Guangze Su, Jiufan Wang, Xiaonan Xu, Yufu Wang, Cankun Wang
Abstract
The advent of Artificial Intelligence (AI) and Machine Learning (ML), particularly deep learning, has escalated the demand for computing resources. However, the high hardware requirements pose challenges for companies, compelling them to outsource ML tasks to the cloud. Nevertheless, concerns about cloud trustworthiness limit such applications. Encrypting data before uploading it to the cloud is a straightforward solution to ensure data security. However, traditional encryption schemes render ciphertext data unable to participate in operations within the ciphertext domain, posing challenges for data analysis. This paper delves into the pivotal role of homomorphic encryption in addressing the critical issue of privacy protection in machine learning.