Leveraging Blockchain for Secure AI Model Training in Decentralized Networks
Abstract
As the adoption of artificial intelligence (AI) and machine learning (ML) models continues to expand, ensuring the security and integrity of training data becomes paramount. This paper presents a novel framework that leverages blockchain technology to create a decentralized, tamper-resistant environment for AI model training. By integrating smart contracts, we ensure that only authenticated data is used for model training, thereby preventing adversarial attacks and data poisoning. The proposed system also allows for transparent auditing of the training process, providing stakeholders with verifiable proof of data integrity. Experimental results demonstrate that the blockchain-enhanced training process not only improves model accuracy by reducing the impact of corrupted data but also offers robust protection against malicious attempts to manipulate the training environment. This approach paves the way for more secure and reliable AI and ML applications in critical domains such as healthcare, finance, and autonomous systems.
Downloads
References
Chiu, C., & Lin, Y. (2020). Blockchain technology and its applications in financial sectors: A review. Journal of Financial Technology, 12(3), 145-162. https://doi.org/10.1016/j.fintech.2020.04.001
Das, M. L., & Sahoo, S. R. (2019). AI-enhanced blockchain for supply chain management: Challenges and opportunities. Journal of Supply Chain Management, 9(2), 117-130. https://doi.org/10.1016/j.scm.2019.03.002
Dhillon, V., Metcalf, D., & Hooper, M. (2017). Blockchain-enabled applications: Understand the blockchain ecosystem and how to make it work for you. Apress. https://doi.org/10.1007/978-1-4842-3081-7
Esposito, C., De Benedictis, A., & Antonelli, F. (2019). Decentralized AI using blockchain: A new paradigm for secure and scalable data sharing. IEEE Transactions on Emerging Topics in Computing, 7(1), 120-131. https://doi.org/10.1109/TETC.2017.2764459
Satish, S., Meduri, K., Nadella, G. S., & Gonaygunta, H. (2022). Developing a Decentralized AI Model
Training Framework Using Blockchain Technology. International Meridian Journal, 4(4), 1-20.
Satish, S., Nadella, G. S., Meduri, K., & Gonaygunta, H. (2022). Collaborative Machine Learning without
Centralized Training Data for Federated Learning. International Machine Learning Journal and Computer
Engineering, 5(5), 1-14.
