Paper 2023/1136

Secure Multiparty Computation with Identifiable Abort from Vindicating Release

Abstract

In the dishonest-majority setting, generic secure multiparty computation (MPC) protocols are fundamentally vulnerable to attacks in which malicious participants learn their outputs and then force the protocol to abort before outputs are delivered to the honest participants. In other words, generic MPC protocols typically guarantee security with abort. This flavor of security permits denial-of-service attacks in many applications, unless the cheating participants who cause aborts are identified. At present, there is a substantial performance gap between the best known protocols that are secure with non-identifiable abort, and the best known protocols that achieve security with identifiable abort (IA). Known constructions with IA rely on generic zero-knowledge proofs, adaptively secure oblivious transfer (OT) protocols, or homomorphic primitives. We present a novel approach for realizing functionalities with a weak form of input-revealing IA, which is based on delicate and selective revealing of committed input values. We refer to this new approach as vindicating release. When our approach is applied to several well-known protocols---including a variant of PVW OT, Softspoken OT extension, DKLs multiplication, and MASCOT generic MPC---the resulting protocols can be combined to realize any sampling functionality with (standard) IA. Such a realization is statistically secure given a variant of statically-corruptable ideal OT, and it differs minimally in terms of cost, techniques, and analysis from the equivalent realization (using the same well-known protocols, unmodified) that lacks identifiability. Using our protocol to sample the correlated randomness of the IOZ compiler reduces the compiler's requirements from an adaptively secure OT protocol to a variant of statically-corruptable ideal OT.