AI-Based Approaches for Protecting Personal Data Privacy on Digital Platforms

As digital platforms become more prevalent, the need to safeguard personal information during its collection, processing, storage and usage is increasing. While traditional security controls are still necessary, they don't address every privacy-related issue in large-scale, constantly evolving data environments. Additional capabilities of AI can be utilized to identify sensitive data, recognize unusual behavior or risks, and assist with privacy-protection processing. The papers explored various roles that machine learning, deep learning, and federated learning play in digital-platform privacy lifecycles, as well as their respective roles in different domains. The literature reviewed presented a conceptual framework that links data identification, sensitivity classification, privacy-risk assessment, protection selection, and continuous monitoring. This was done using the following methods: The framework stresses the importance of treating privacy as an ongoing process, rather than a single control applied at the end of data processing. Efforts on utility and computational cost can be influenced by greater privacy, while AI models may face additional privacy risks. The proposed structure does not assert that a single approach is adequate for all circumstances. It provides a methodical framework for choosing and combining methods that are appropriate for the risk, data type, and desired outcome. This approach can be advanced through future experimental validation, adaptive controls, privacy-preserving learning, and transparent AI-based privacy assistance.

[1]   S. Yu, F. Carroll, and B. L. In IEEE Access, vol, Bentley provided insights into privacy protection research in AI. 12, 41704–41726, 2024, doi: 10.1109/ACCESS.2024.3378126. [2]    V. Morel, L. H. Iwaya, and S. In IEEE Access, Fischer-Hübner reviewed the literature on AI-Driven Privacy Assistants. 13, 160982–161002, 2025, doi: 10.1109/ACCESS.2025.3609188. [3]   E. Jahns, M. Stojkov, and M. A. Kinsy, "Privacy-Preserving Deep Learning: A Survey on Theoretical Foundation, Software Framework and Hardware Accelerators" IEEE Access, vol. 13, 67821–67855, 2025, doi: 10.1109/ACCESS.2025.3561721. [4]  C. Chronis, I. Varlamis, Y. Himeur, A. N. Sayed, T. M. Al-Hasan, A. Nhlabatsi, F. Bensaali, and G. In IEEE Open Journal of the Computer Society, Dimitrakopoulos authored a paper on the application of Federated Learning to privacy-protecting reconstruction systems. 5, 227–247, 2024, doi: 10.1109/OJCS.2024.3396344. [5]   M. Amiri, G. Quirchmayr, and E. "Using a Local Distance-Based Methodology to Introduce Differential Privacy for Recommender Systems," Weippl wrote in IEEE Access. vol. 44, Issue 1, Issue 2. 13, 161824– 161843, 2025, doi: 10.1109/ACCESS.2025.3605563. [6]  C. Dwork and A. Roth, "The Algorithmic Foundations of Differential Privacy," Foundations and Trends in Theoretical Computer Science, vol. 9, no. 3–4, pp. 211– 407, 2014, doi: 10.1561/0400000042. [7]  M. Abadi, A. Chu, I. Goodfellow, H. B. McMahan, I. Mironov, K. Talwar, and L. Zhang, "Deep Learning with Differential Privacy," in Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security, pp. 308–318, 2016. [8]  H. B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, "Communication-Efficient Learning of Deep Networks from Decentralized Data," in Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, vol. 54, pp. 1273–1282, 2017. [9]  K. Bonawitz, V. Ivanov, B. Kreuter, A. Marcedone, H. B. McMahan, S. Patel, D. Ramage, A. Segal, and K. Seth, "Practical Secure Aggregation for Privacy-Preserving Machine Learning," in Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security, pp. 1175–1191, 2017, doi: 10.1145/3133956.3133982. [10]  R. Shokri, M. Stronati, C. Song, and V. Shmatikov, "Membership Inference Attacks Against Machine Learning Models," in 2017 IEEE Symposium on Security and Privacy, pp. 3–18, 2017, doi: 10.1109/SP.2017.41.