Robust and Verifiable MPC with Applications to Linear Machine Learning Inference

2025-05-31

IACR Cryptology ePrint Archive

AI Summary

This work addresses security deficiencies in dishonest-majority secure multi-party computation (MPC), where a majority of parties may be malicious. We propose the first dishonest-majority MPC protocol achieving both full identifiability and robustness: it enables real-time, consistent identification of malicious parties while completing computation without abort or restart. To overcome limitations of prior approaches—such as SPDZ’s abort-only security and inefficiency in Cunningham et al.’s protocols due to exponentiation-based commitments—we design lightweight lattice-based commitment schemes and zero-knowledge proofs, and introduce a semi-honest trusted third party to assist recovery. Evaluated in an ML-as-a-Service setting, our protocol efficiently supports linear model inference; its malicious behavior detection and recovery overhead is significantly lower than comparable protocols, and its security strictly subsumes that of SPDZ.

Technology Category

• Machine Learning: Hardware-aware ML
• Multiagent Systems: Mechanism Design
• Natural Language Processing: Safety and Robustness

Application Category

• Security and Privacy: Large-scale security measurements
• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

In this work, we present an efficient secure multi-party computation MPC protocol that provides strong security guarantees in settings with dishonest majority of participants who may behave arbitrarily. Unlike the popular MPC implementation known as SPDZ [Crypto '12], which only ensures security with abort, our protocol achieves both complete identifiability and robustness. With complete identifiability, honest parties can detect and unanimously agree on the identity of any malicious party. Robustness allows the protocol to continue with the computation without requiring a restart, even when malicious behavior is detected. Additionally, our approach addresses the performance limitations observed in the protocol by Cunningham et al. [ICITS '17], which, while achieving complete identifiability, is hindered by the costly exponentiation operations required by the choice of commitment scheme. Our protocol is based on the approach by Rivinius et al. [S&P '22], utilizing lattice-based commitment for better efficiency. We achieved robustness with the help of a semi-honest trusted third party. We benchmark our robust protocol, showing the efficient recovery from parties' malicious behavior. Finally, we benchmark our protocol on a ML-as-a-service scenario, wherein clients off-load the desired computation to the servers, and verify the computation result. We benchmark on linear ML inference, running on various datasets. While our efficiency is slightly lower compared to SPDZ's, we offer stronger security properties that provide distinct advantages.

Problem

• Achieves robust MPC with dishonest majority security
• Ensures complete identifiability of malicious parties
• Optimizes performance for linear ML inference

Innovation

• Lattice-based commitment for better efficiency
• Semi-honest trusted third party ensures robustness
• Complete identifiability and robustness in MPC