Zero-Knowledge Proofs in Anti-Money Laundering Multiparty Computation

Money laundering presents substantial challenges to financial institutions, necessitating effective risk assessment mechanisms while maintaining client confidentiality. This study introduces a blockchain-integrated framework that utilizes Secure Multiparty Computation (MPC) and Zero-Knowledge Proofs to facilitate privacy-preserving Anti-Money Laundering (AML) score computation among banks. The pro-posed system adopts a hybrid architecture, wherein banks conduct off-chain computations on encrypted client data. The innovation of this paper lies in its formulation of an AML score, underpinned by Bayesian analysis, developed in a trustless and private manner on a blockchain ledger. It exploits a variety of client parameters that represent financial activities and risk factors. The evaluation of results using simulated data indicates the framework’s ability to accurately assess the likelihood of money laundering activities. To ensure secure and transparent computation, we propose integrating the SPDZ protocol for MPC with Circom to generate proofs of correct and unbiased computation results. Ultimately, the solution’s workflow and architecture ensure that computation results are publicly verifiable on-chain, allowing for additional regulatory investigation when high-risk clients are identified.