Sovchain: novel mutual authentication scheme for self-sovereign identity management using blockchain for open banking

Abstract A significant limitation of Financial-Grade API (FAPI) 2.0 is its inability to verify the true origin of the authenticator used during the authentication process, because it focuses solely on validating the parameters of the request rather than verifying the identity of the sender. To enhance FAPI 2.0 security, we propose SovChain, an approach that embeds machine learning into the pre-authentication phase of self-sovereign identity (SSI) systems, enabling proactive risk assessment before any credential is accepted or validated. By addressing a clear gap in the literature, the absence of a mechanism for early threat detection within SSI, the findings of this study contribute to the development of a more trustworthy open banking ecosystem. SovChain was designed with Hyperledger smart contracts, tested on the TON Internet of Things dataset, and validated using AVISPA for formal security. Simulation experiments evaluated registration and authentication flows, support vector machine classification, communication overhead, and throughput under varying participant loads. SovChain achieved 98.2% accuracy with a 97.7% F1 score. By embedding proactive risk assessment into SSI, SovChain demonstrates the feasibility of combining lightweight machine learning with blockchain-based authentication, offering a scalable and regulation-ready solution for open banking and beyond.