International Association
for Cryptologic Research

We review the two RSA-based accumulators introduced by Camenisch and Lysyanskaya in 2002 in the setting of revocation for anonymous credential schemes, such as Idemix or BBS+. We show that in such a setting, the lower and upper bounds placed on the accumulated values in the paper are unnecessarily strict; they can be removed almost entirely (up to the group order of the credential scheme). This allows the accumulators to be used on elliptic curves of ordinary sizes, such as the ones on which BBS+ is commonly implemented. We also offer some notes and optimizations for implementations of anonymous credential schemes that use these accumulators to enable revocation.

Developers of computer-aided cryptographic tools are optimistic that formal methods will become a vital part of developing new cryptographic systems. We study the use of such tools to specify and verify the implementation of Classic McEliece, one of the code-based cryptography candidates in the fourth round of the NIST Post-Quantum standardisation Process. From our case study we draw conclusions about the practical applicability of these methods to the development of novel cryptography.

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 $4.95$ms and $0.53$ms, respectively, and $0.21$MB 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.

Template attacks~(TAs) are one of the most powerful Side-Channel Analysis~(SCA) attacks. The success of such attacks relies on the effectiveness of the profiling model in modeling the leakage information. A crucial step for TA is to select relevant features from the measured traces, often called Points Of Interest~(POIs), to extract the leakage information. Previous research indicates that properly selecting the input leaking features could significantly increase the attack performance. However, due to the presence of SCA countermeasures and advancements in technology nodes, such features become increasingly difficult to extract with conventional approaches such as Principle Component Analysis (PCA) and the Sum Of Squared pairwise T-differences based method (SOST).

This work proposes a framework, AutoPOI, based on proximal policy optimization to automatically find, select, and scale down features. The input raw features are first grouped into small regions. The best candidates selected by the framework are further scaled down with an online-optimized dimensionality reduction neural network. Finally, the framework rewards the performance of these features with the results of TA. Based on the experimental results, the proposed framework can extract features automatically that lead to comparable state-of-the-art performance on several commonly used datasets.

Recently, in post-quantum cryptography migration, it has been shown that an IND-1-CCA-secure key encapsulation mechanisms (KEM) is required for replacing an ephemeral Diffie-Hellman (DH) in widely-used protocols, e.g., TLS, Signal, and Noise. IND-1-CCA security is a notion similar to the traditional IND-CCA security except that the adversary is restricted to one single decapsulation query. At EUROCRYPT 2022, based on CPA-secure public-key encryption (PKE), Huguenin-Dumittan and Vaudenay presented two IND-1-CCA KEM constructions called $T_{CH}$ and $T_H$, which are much more efficient than the widely-used IND-CCA-secure Fujisaki-Okamoto (FO) KEMs. The security of $T_{CH}$ was proved in both random oracle model (ROM) and quantum random oracle model (QROM). However, the QROM proof of $T_{CH}$ requires that the ciphertext size of the resulting KEM is twice as large as the one of the underlying PKE. While, the security of $T_H$ was only proved in the ROM, and the QROM proof is left open.

In this paper, we present an IND-1-CCA KEM construction $T_{RH}$, which can be seen as an implicit variant $T_H$, and is as efficient as $T_H$. We prove the security of $T_{RH}$ in both ROM and QROM with much tighter reductions than Huguenin-Dumittan and Vaudenay's work. In particular, our proof will not lead to ciphertext expansion. Moreover, for $T_{RH}$, $T_H$ and $T_{CH}$, we also show that a $O(1/q)$ ($O(1/q^2)$, resp.) reduction loss is unavoidable in the ROM (QROM, resp.), and thus claim that our ROM proof is optimal in tightness. Finally, we make a comprehensive comparison among the relative strengths of IND-1-CCA and IND-CCA in the ROM and QROM.

This paper investigates different ways of applying multi-task learning in the context of two masked AES implementations (via the ASCADv1 and ASCADv2 databases). We propose novel ideas: jointly using multiple single-task models (aka multi-target learning), custom layers (enabling the use of multi-task learning without the need for information about randomness), and hierarchical multi-task models (owing to the idea of encoding the hierarchy flow directly into a multi-task learning model). Our work provides comparisons with existing approaches to deep learning and delivers a first attack using multi-task models without randomness during training, and a new best attack for the ASCADv2 dataset.

The redundant of multimedia data made an unnecessary waste in encrypted cloud storage, unlike text with completely consistent content, multimedia data allows a certain degree of similarity in deduplication, In this work, we focus on the multimedia data which takes a seriously proportion of storage in scenarios such as data outsourcing to propose secure fuzzy deduplication without the additional servers based on Convergent Encryption(CE), say the Single-server Fuzzy Deduplication (SSFD). Compared to the related fuzzy deduplication, SSFD is strong at resisting brute-force attacks caused by server-server collusion, moreover, we also put server-client collusion attacks into security solutions. Additionally, to enhance the security of data, the proposed scheme provides both protection against replay attacks and verification of label consistency and adds no extra communication such as Proof of Ownership(PoW) in interaction. We separately presented a formal security analysis and performed performance at last to prove security solutions and evaluate the experimental results, it shows SSFD provides both a reliable fuzzy images secure deduplication protocol and a computationally feasible solution.

With the recent development of quantum computers, various studies on quantum artificial intelligence technology are being conducted. Quantum artificial intelligence can improve performance in terms of accuracy and memory usage compared to deep learning on classical computers. In this work, we proposed an attack technique that recovers keys by learning patterns in cryptographic algorithms by applying quantum artificial intelligence to cryptanalysis. Cryptanalysis was performed in the current practically usable quantum computer environment, and this is the world's first study to the best of our knowledge. As a result, we reduced 70 epochs and reduced the parameters by 19.6%. In addition, higher average BAP (Bit Accuracy Probability) was achieved despite using fewer epochs and parameters. For the same epoch, the method using a quantum neural network achieved a 2.8% higher BAP with fewer parameters. In our approach, quantum advantages in accuracy and memory usage were obtained with quantum neural networks. It is expected that the cryptanalysis proposed in this work will be better utilized if a larger-scale stable quantum computer is developed in the future.

Sigstore is a Linux Foundation project aiming to become the new standard for signing software artifacts. It consists of a free certificate authority called Fulcio, a tamper-resistant public log called Rekor, and an optional federated OIDC identity provider called Dex, where Rekor also acts as the timestamping service. Several command line interfaces (CLIs), written in different languages, are available to interact with it for signing software artifacts.

Ironically, we will show in this paper the design of Sigstore eliminates the need of Sigstore, i.e., the key components mentioned above are inessential. Specifically, we will first show how to remove the dependency on Fulcio from existing CLIs while keeping the CLIs work. Next, we will show how to remove the dependency on Rekor from the CLIs. Last, we will explain why relying on Dex, an optional black box with too much power, should be avoided.

As none of Fulcio, Rekor, and Dex is essential to making existing CLIs work, we conclude that they are unnecessary trusted third parties which the open source community should avoid employing. Instead, existing CLIs can be easily adapted to remove the dependency on them while providing the same functionality and user experience. The design of Sigstore is an example of solving a problem with a method which requires the solution as the input.

We introduce a new cryptographic primitive, aptly named ring verifiable random functions (ring VRF), which provides an array of uses, especially in anonymous credentials. Ring VRFs are (anonymized) ring signatures that prove correct evaluation of an authorized signer's PRF, while hiding the specific signer's identity within some set of possible signers, known as the ring.

We discover a family of ring VRF protocols with surprisingly efficient instantiations, thanks to our novel zero-knowledge continuation technique. Intuitively our ring VRF signers generate two linked proofs, one for PRF evaluation and one for ring membership. An evaluation proof needs only a cheap Chaum-Pedersen DLEQ proof, while ring membership proof depends only upon the ring itself. We reuse this ring membership proof across multiple inputs by expanding a Groth16 trusted setup to rehide public inputs when rerandomizing the Groth16. Incredibly, our fastest amortized ring VRF needs only eight G_1 and two G_2 scalar multiplications, making it the only ring signature with performance competitive with group signatures.

We discuss applications that range across the anonymous credential space:

As in Bryan Ford's proof-of-personhood work, a ring VRF output acts like a unique pseudo-nonymous identity within some desired context, given as the ring VRF input, but remains unlinkable between different contexts. These unlinkable but unique pseudonyms provide a better balance between user privacy and service provider or social interests than attribute based credentials like IRMA credentials.

Ring VRFs support anonymously rationing or rate limiting resource consumption that winds up vastly more flexible and efficient than purchases via money-like protocols.

We define the security of ring VRFs in the universally composable (UC) model and show that our protocol is UC secure.

With the advent of secure function evaluation, distrustful parties can jointly compute on their private inputs without disclosing anything besides the results. Yao's garbled circuit protocols have become an integral part of secure computation thanks to considerable efforts made to make it feasible, practical, and more efficient. These efforts have resulted in multiple optimizations on this primitive to enhance its performance by orders of magnitude over the last years. Such improvement targets have been defined to primarily reduce the cost of garbling in terms of computation and communication required for the creation, transfer, and evaluation of the garbled tables. The advancement in protocols has also led to the development of general-purpose compilers and tools made available to academia and industry. For decades, the security of protocols offered in those tools has been assured with regard to sound proofs and the promise that during the computation, no information on parties' input would be leaking.

In a parallel effort, however, side-channel analysis has gained momentum in connection with the real-world implementation of cryptographic primitives. Timing side-channel attacks have proven themselves effective in retrieving secrets from implementations, even through remote access to them. Nevertheless, the vulnerability of garbled circuit constructions, in particular, the optimized one to timing attacks, has, surprisingly, never been discussed in the literature. This paper introduces Goblin, the first timing attack against two commonly employed optimized garbled circuit constructions, namely free-XOR and half-gates. Goblin is a machine learning-assisted, non-profiling, single-trace timing attack, which successfully recovers the garbler's input during the computation. As the first step, Goblin targets the TinyGarble family and its core garbling tool, JustGarble. In this regard, Goblin hopefully paves the way for further research.

We put forth a new paradigm for program obfuscation, where obfuscated programs are endowed with proofs of ``well-formedness.'' In addition to asserting existence of an underlying plaintext program with an attested structure and functionality, these proofs also prevent mauling attacks, whereby an adversary surreptitiously creates an obfuscated program based on secrets which are embedded in a given obfuscated program. We call this new guarantee Chosen Obfuscation Attack (COA) security.

We define and construct general-purpose COA-secure Probabilistic Indistinguishability Obfuscators for circuits, assuming sub-exponential IO for circuits and CCA commitments. To demonstrate the power of the new notion, we use it to realize, in the plain model: - Structural Watermarking, which is a new form of software watermarking that provides significantly broader protection than current schemes and features a keyless, public verification process. - Completely CCA encryption, which is a strengthening of completely non-malleable encryption.

We also show, based on the same assumptions, a generic method for enhancing any obfuscation mechanism that guarantees any semantic-style form of hiding to one that provides also COA security.

In [2] we studied collision side-channel attacks, and derived an optimal distinguisher for key ranking. In this note we propose a heuristic estimation procedure for key ranking based on this distinguisher, and provide estimates of lower bounds for secret key ranks in collision side-channel attacks. The procedure employs nonuniform sampling introduced in [1], and it is more efficient than the subset uniform sampling procedure [3].

[1] MCRank: Monte Carlo Key Rank Estimation for Side-Channel Security Evaluations. [2] Optimal Collision Side-Channel Attacks. [3] A Note on Key Ranking for Optimal Collision Side-Channel Attacks.

An accumulator is a cryptographic primitive that allows a prover to succinctly commit to a set of values while being able to provide proofs of (non-)membership. A batch proof is an accumulator proof that can be used to prove (non-)membership of multiple values simultaneously.

In this work, we present a zero-knowledge batch proof with constant proof size and constant verification in the Bilinear Pairings (BP) setting. Our scheme is 16x to 42x faster than state-of-the-art SNARK-based zero-knowledge batch proofs in the RSA setting. Additionally, we propose protocols that allow a prover to aggregate multiple individual non-membership proofs, in the BP setting, into a single batch proof of constant size. Our construction for aggregation satisfies a strong soundness definition - one where the accumulator value can be chosen arbitrarily.

We evaluate our techniques and systematically compare them with RSA-based alternatives. Our evaluation results showcase several scenarios for which BP accumulators are clearly preferable and can serve as a guideline when choosing between the two types of accumulators.

Integrating private set intersection (PSI) protocols within real-world data workflows, software applications, or web services can be challenging. This can occur because data contributors and result recipients do not have the technical expertise, information technology infrastructure, or other resources to participate throughout the execution of a protocol and/or to incur all the communication costs associated with participation. Furthermore, contemporary workflows, applications, and services are often designed around RESTful APIs that might not require contributors or recipients to remain online or to maintain state. Asynchronous delegated PSI protocol variants can better match the expectations of software engineers by (1) allowing data contributors to contribute their inputs and then to depart permanently, and (2) allowing result recipients to request their result only once they are ready to do so. However, such protocols usually accomplish this by introducing an additional party that learns some information about the size of the intersection. This work presents an asynchronous delegated PSI protocol variant that does not reveal the intersection size to the additional party. It is shown that such a protocol can have, on average, linear time and space complexity.

In this article we realize a general study on the nonlinearity of weightwise perfectly balanced (WPB) functions. First, we derive upper and lower bounds on the nonlinearity from this class of functions for all $n$. Then, we give a general construction that allows us to provably provide WPB functions with nonlinearity as low as $2^{n/2-1}$ and WPB functions with high nonlinearity, at least $2^{n-1}-2^{n/2}$. We provide concrete examples in $8$ and $16$ variables with high nonlinearity given by this construction. In $8$ variables we experimentally obtain functions reaching a nonlinearity of $116$ which corresponds to the upper bound of Dobbertin's conjecture, and it improves upon the maximal nonlinearity of WPB functions recently obtained with genetic algorithms. Finally, we study the distribution of nonlinearity over the set of WPB functions. We examine the exact distribution for $n=4$ and provide an algorithm to estimate the distributions for $n=8$ and $16$, together with the results of our experimental studies for $n=8$ and $16$.

A nonce-respecting tweakable blockcipher is the building-block for the OCB authenticated encryption mode. An XEX-based TBC is used to process each block in OCB. However, XEX can provide at most birthday bound privacy security, whereas in Asiacrypt 2017, beyond-birthday-bound (BBB) forging security of OCB3 was shown by Bhaumik and Nandi. In this paper we study how at a small cost we can construct a nonce-respecting BBB-secure tweakable blockcipher. We propose the OTBC-3 construction, which maintains a cache that can be easily updated when used in an OCB-like mode. We show how this can be used in a BBB-secure variant of OCB with some additional keys and a few extra blockcipher calls but roughly the same amortised rate.

Trapdoor Claw-free Functions (TCFs) are two-to-one trapdoor functions where it is computationally hard to find a claw, i.e., a colliding pair of inputs. TCFs have recently seen a surge of renewed interest due to new applications to quantum cryptography: as an example, TCFs enable a classical machine to verify that some quantum computation has been performed correctly. In this work, we propose a new family of (almost two-to-one) TCFs based on conjectured hard problems on isogeny-based group actions. This is the first candidate construction that is not based on lattice-related problems and the first scheme (from any plausible post-quantum assumption) with a deterministic evaluation algorithm. To demonstrate the usefulness of our construction, we show that our TCF family can be used to devise a computational test of qubit, which is the basic building block used in the general verification of quantum computations.

The DECentralized Oracle (DECO) protocol enables the verifiable provenance of data from Transport Layer Security (TLS) connections through secure two-party computation and zero-knowledge proofs. In this paper, we present PECO, an extension of DECO that enhances privacy features through the integration of two new private three-party handshake protocols (P3P-HS). PECO allows any web user to prove to a verifier the properties of data from TLS connections without disclosing the identity of the servers. Like DECO's three-party handshake protocol, PECO's P3P-HS methods do not require any changes on the server side. PECO offers two options: one that provides $k-$anonymity for the server's identity, and another that completely masks the server's identity from the verifier. PECO is based on three main protocols: (a) commit-and-proof zero-knowledge proofs (CP-ZKP) that enable the proof of relations under committed values in zero-knowledge, (b) verification of Elliptic Curve Digital Signature Algorithm (ECDSA) signatures under a committed public key without revealing the key (zkAttest), and (c) a proof of membership to verify that a committed key belongs to a set of keys. We estimate the performance of both P3P-HS protocols and compare it to TLS timeout using state-of-the-art implementations.

Within just four years, the blockchain-based Decentralized Finance (DeFi) ecosystem has accumulated a peak total value locked (TVL) of more than 253 billion USD. This surge in DeFi’s popularity has, unfortunately, been accompanied by many impactful incidents. According to our data, users, liquidity providers, speculators, and protocol operators suffered a total loss of at least 3.24 billion USD from Apr 30, 2018 to Apr 30, 2022. Given the blockchain’s transparency and increasing incident frequency, two questions arise: How can we systematically measure, evaluate, and compare DeFi incidents? How can we learn from past attacks to strengthen DeFi security?

In this paper, we introduce a common reference frame to systematically evaluate and compare DeFi incidents, including both attacks and accidents. We investigate 77 academic papers, 30 audit reports, and 181 real-world incidents. Our open data reveals several gaps between academia and the practitioners’ community. For example, few academic papers address “price oracle attacks” and “permissonless interactions”, while our data suggests that they are the two most frequent incident types (15% and 10.5% correspondingly). We also investigate potential defenses, and find that: (i) 103 (56%) of the attacks are not executed atomically, granting a rescue time frame for defenders; (ii) SoTA bytecode similarity analysis can at least detect 31 vulnerable/23 adversarial contracts; and (iii) 33 (15.3%) of the adversaries leak potentially identifiable information by interacting with centralized exchanges.