Abstract Distributed zero-knowledge (dZK) proofs, recently introduced by Boneh et al. (CRYPTO‘19), allow a prover $$\mathcal{P}$$ P to prove NP statements on an input x , which is distributed between k verifiers $$\mathcal{V}_1,\ldots ,\mathcal{V}_k$$ V 1 , … , V k , where each $$\mathcal{V}_i$$ V i holds only a piece of x . As in standard ZK proofs, dZK proofs guarantee Completeness when all parties are honest; Soundness against a malicious prover colluding with t verifiers; and Zero Knowledge against a subset of t malicious verifiers, in the sense that they learn nothing about the NP witness and the input pieces of the honest verifiers. Unfortunately, dZK proofs provide no correctness guarantee for an honest prover against a subset of maliciously corrupted verifiers. In particular, such verifiers might be able to “frame” the prover, causing honest verifiers to reject a true claim. This is a significant limitation, since such scenarios arise naturally in dZK applications, e.g., for proving honest behavior, and such attacks are indeed possible in existing dZKs (Boneh et al., CRYPTO‘19). We put forth and study the notion of strong completeness for dZKs, guaranteeing that true claims are accepted even when t verifiers are maliciously corrupted. We then design strongly-complete dZK proofs in the honest-majority setting using the “MPC-in-the-head” paradigm of Ishai et al. (STOC‘07), providing a novel analysis that exploits the unique properties of the distributed setting. To demonstrate the usefulness of strong completeness, we present several applications in which it is instrumental in obtaining security. First, we construct a certifiable version of Verifiable Secret Sharing (VSS), which is a VSS in which the dealer additionally proves that the shared secret satisfies a given NP relation. Our construction withstands a constant fraction of corruptions, whereas a previous construction of Ishai et al. (TCC‘14) required $$k={\textsf{poly}}\left( t\right) $$ k = poly t . We also design a reusable version of certifiable VSS that we introduce, in which the dealer can prove an unlimited number of predicates on the same shared secret. Finally, we extend a compiler of Boneh et al. (CRYPTO‘19), who used dZKs to transform a class of “natural” semi-honest protocols in the honest-majority setting into maliciously secure ones with abort. Our compiler uses strongly-complete dZKs to obtain identifiable abort.

Abstract Leakage-resilient cryptography aims to protect cryptographic primitives from so-called “side channel attacks” that exploit their physical implementation to learn their input or secret state. Starting from the works of Ishai, Sahai and Wagner (CRYPTO‘03) and Micali and Reyzin (TCC‘04), most works on leakage-resilient cryptography either focus on protecting general computations, such as circuits or multiparty computation protocols, or on specific non-interactive primitives such as storage, encryption, and signatures. This work focuses on leakage resilience for the middle ground, namely for distributed and interactive cryptographic primitives. Our main technical contribution is designing the first secret sharing scheme that is equivocal, resists adaptive probing of a constant fraction of bits from each share, while incurs only a constant blowup in share size. Equivocation is a strong leakage-resilience guarantee, recently introduced by Hazay et al. (ITC, 2021). Our construction is obtained via a general compiler which we introduce, that transforms any secret sharing scheme into an equivocal scheme against adaptive leakage. An attractive feature of our compiler is that it respects additive reconstruction; namely, if the original scheme has additive reconstruction, then the transformed scheme has linear reconstruction. We extend our compiler to a general paradigm for protecting distributed primitives against leakage and show its applicability to various primitives, including secret sharing, verifiable secret sharing, function secret sharing, distributed encryption and signatures, and distributed zero-knowledge proofs. For each of these primitives, our paradigm transforms any construction of the primitive into a scheme that resists adaptive party corruptions, as well as adaptive probing leakage of a constant fraction of bits in each share when the share is stored in memory (but not when it is used in computations). Moreover, the transformation incurs only a constant blowup in the share size and respects additive reconstruction—an important feature for several of these primitives, such as function secret sharing and distributed encryption.

Abstract We construct the first actively-secure Multi-Party Computation (MPC) protocols with an arbitrary number of parties in the dishonest majority setting, for an arbitrary field $${\mathbb {F}}$$ F with constant communication overhead over the “passive-GMW” protocol (Goldreich, Micali and Wigderson, STOC ‘87). Our protocols rely on passive implementations of Oblivious Transfer (OT) in the Boolean setting and Oblivious Linear function Evaluation (OLE) in the arithmetic setting. Previously, such protocols were only known over sufficiently large fields (Genkin et al. STOC ‘14) or a constant number of parties (Ishai et al. CRYPTO ‘08). Conceptually, our protocols are obtained via a new compiler from a passively-secure protocol for a distributed multiplication functionality $${{{\mathcal {F}}}}_{\scriptscriptstyle \textrm{MULT}}$$ F MULT , to an actively-secure protocol for general functionalities. Roughly, $${{{\mathcal {F}}}}_{\scriptscriptstyle \textrm{MULT}}$$ F MULT is parameterized by a linear-secret sharing scheme $${{{\mathcal {S}}}}$$ S , where it takes $${{{\mathcal {S}}}}$$ S -shares of two secrets and returns $${{{\mathcal {S}}}}$$ S -shares of their product. We show that our compilation is concretely efficient for sufficiently large fields, resulting in an overhead of 2 when securely computing natural circuits. Our compiler has two additional benefits: (1) It can rely on any passive implementation of $${{{\mathcal {F}}}}_{\scriptscriptstyle \textrm{MULT}}$$ F MULT , which, besides the standard implementation based on OT (for Boolean) and OLE (for arithmetic), allows us to rely on implementations based on threshold cryptosystems (Cramer et al. Eurocrypt ‘01), and (2) it can rely on weaker-than-passive (i.e., imperfect/leaky) implementations, which in some parameter regimes yield actively-secure protocols with overhead less than 2. Instantiating this compiler with an “honest-majority” implementation of $${{{\mathcal {F}}}}_{\scriptscriptstyle \textrm{MULT}}$$ F MULT , we obtain the first honest-majority protocol (with up to one-third corruptions) for Boolean circuits with constant communication overhead over the best passive protocol (Damgård and Nielsen, CRYPTO ‘07).

Privacy-Preserving Machine Learning (PPML) provides protocols for learning and statistical analysis of data that may be distributed amongst multiple data owners (e.g., hospitals that own proprietary healthcare data), while preserving data privacy. The PPML literature includes protocols for various learning methods, including ridge regression. Ridge regression controls the L2 norm of the model, but does not aim to strictly reduce the number of non-zero coefficients, namely the L0 norm of the model. Reducing the number of non-zero coefficients (a form of feature selection) is important for avoiding overfitting, and for reducing the cost of using learnt models in practice. In this work, we develop a first privacy-preserving protocol for sparse linear regression under L0 constraints. The protocol addresses data contributed by several data owners (e.g., hospitals). Our protocol outsources the bulk of the computation to two non-colluding servers, using homomorphic encryption as a central tool. We provide a rigorous security proof for our protocol, where security is against semi-honest adversaries controlling any number of data owners and at most one server. We implemented our protocol, and evaluated performance with nearly a million samples and up to 40 features.

With the development of sequencing technologies, viral strain classification - which is critical for many applications, including disease monitoring and control - has become widely deployed. Typically, a lab (client) holds a viral sequence, and requests classification services from a centralized repository of labeled viral sequences (server). However, such “classification as a service” raises privacy concerns. In this paper we propose a privacy-preserving viral strain classification protocol that allows the client to obtain classification services from the server, while maintaining complete privacy of the client's viral strains. The privacy guarantee is against active servers, and the correctness guarantee is against passive ones. We implemented our protocol and performed extensive benchmarks, showing that it obtains almost perfect accuracy (99.8%-100%) and microAUC (0.999), and high efficiency (amortized per-sequence client and server runtimes of 4.95ms and 0.53ms, respectively, and 0.21MB communication). In addition, we present an extension of our protocol that guarantees server privacy against passive clients, and provide an empirical evaluation showing that this extension provides the same high accuracy and microAUC, with amortized per sequences overhead of only a few milliseconds in client and server runtime, and 0.3MB in communication complexity. Along the way, we develop an enhanced packing technique in which two reals are packed in a single complex number, with support for homomorphic inner products of vectors of ciphertexts. We note that while similar packing techniques were used before, they only supported additions and multiplication by constants.