Hash-based signature schemes present an appealing proposition: they are broadly regarded as the most conservative in terms of assumptions, and their efficiency profiles suit many applications. However, little is known about their compatibility with threshold signing, which is common practice today in domains like digital asset security. Especially in the context of an impending migration to post-quantum cryptography, this gap in the literature leaves uncertain what hash-based standards would imply for applications that rely on threshold signing. In this work, we develop a new framework by which to analyze such hash-based schemes, and prove inherent barriers to designing an entire class of efficient (fully distributed) threshold signing protocols. Our framework is simple, and yet powerful. Our results encompass standard schemes such as XMSS and SLH-DSA. Our approach is to connect hash-based signatures to the literature on straight-line extraction in the random oracle model. We then apply and build upon techniques from the recent work of Doerner et al. (Crypto '24) to show that threshold signing protocols for hash based schemes must—beyond a certain number of signers—use the circuit representation of the hash function, i.e. it can not use the hash function as a black box. We then give a more fine-grained analysis for the specific case of the Winternitz signature scheme—a building block of both NIST standards—and show that it can not permit even two-party signing with blackbox use of the hash function.

Very recently, Wesolowski (ePrint 2026/1486) proposed a novel heuristic algorithm for solving the supersingular isogeny-path problem in time and memory \(p^{1/3+o(1)}\), where \(p\) is the characteristic of the underlying field. Although this improves asymptotically on the previous \(p^{1/2}\log^{O(1)}(p)\) complexity, its concrete impact on isogeny-based schemes, particularly SQIsign, remains unclear because of the superpolynomial overhead hidden in \(p^{o(1)}\) and the exponential memory requirement. In this work, we assess the cost of Wesolowski's attack, study its time--memory tradeoffs, and investigate optimizations based on the van Oorschot -- Wiener (vOW) technique. Our analysis shows that, over the practical memory ranges considered, Wesolowski's algorithm is expected to improve on the state of the art only at NIST Level I. At higher security levels, its prohibitive memory requirements offset the asymptotic advantage over accelerated Delfs--Galbraith methods. With high parallelization, however, the vOW variant can recover this advantage. We further study quantum claw-finding improvements. While Grover search can essentially eliminate the large memory requirement, it offers little improvement in running time, whereas Tani's algorithm provides a stronger gate--memory tradeoff at the cost of substantial coherent quantum memory. Finally, we provide an optimized SageMath implementation and compare the theoretical time and space complexities against their empirical counterparts.

We study the matrix signing problem for positive semidefinite matrices, assigning one sign to each original matrix regardless of its rank. This includes the higher-rank Kadison--Singer partition problem [KS59] when the matrices sum to the identity. For arbitrary positive semidefinite matrices $A_1,\ldots,A_m\in\mathbb C^{n\times n}$, we prove that there are signs $\sigma_i\in\{-1,1\}$ such that $\|\sum_i \sigma_i A_i\|\le 3.34963\|\sum_i \operatorname{tr}[A_i]A_i\|^{1/2}$. We give a deterministic algorithm achieving this bound in $\widetilde O(mn^2+n^{4.75})$ real arithmetic operations. Inspired by the Lee--Sidford barrier, we use a potential function built from Schatten norms to obtain a square-root logarithmic dependence on the rank. Under the additional assumptions $\sum_iA_i=I$, $\|A_i\|\le\alpha$, and $\operatorname{rank}(A_i)\le r$, we obtain a discrepancy of at most $\min\{1,3.48745\sqrt{\alpha\log(2r)}\}$. We also give a deterministic algorithm achieving a discrepancy of at most $\min\{1,4.171\sqrt{\alpha\log(2r)}\}$ in $\widetilde O(mn^2+n^{6.38})$ real arithmetic operations. This dependence on $\alpha$ and $r$ is optimal up to an absolute constant when $0<\alpha\le1/256$ and $r\ge256\alpha^{-6}$: even diagonal matrices can force every signing to have a discrepancy of at least $\min\{1,\sqrt{\alpha\log(2r)}\}$. For a zero-one incidence matrix with at most $d$ ones in each row and each column, our deterministic algorithm achieves a discrepancy of $O(\sqrt{d\log(d)})$, recovering Harvey's bound [Har15] by a different method. We give a deterministic construction of a single spanning tree that is simultaneously spectrally thin with respect to any given collection of $k\ge1$ positive edge weightings of the same graph, assuming small edge leverage in every weighting. The tree retains the original edge weights, and its thinness bound grows only logarithmically with $k$.

Aardal et al. (CRYPTO 2025) provided the first complete security proof of SQIsign; however, their reduction incurs a square-root loss in the prime characteristic due to the application of a loose bound on the min-entropy. For instance, at NIST security level I, an adversary making $2^{64}$ signing queries renders the security proof vacuous. In this note, we show that the min-entropy of SQIsign is optimal, namely $\mathcal{O}(1/p)$. Although this improvement does not yield full $\lambda$-bit security, we show that it preserves two-thirds of the expected bit-security. We show that this artifact comes from an information-theoretic loss in the zero-knowledge simulation of SQIsign, suggesting a new proof technique is needed to achieve full $\lambda$-bit security at the current parameters.

Distribution-aware encryption seeks guarantees tailored to the probability law of the protected data and the secret key. We study simultaneous message-recovery, key-recovery, and target-distribution semantic security for arbitrary finite joint distributions, without assuming that messages and keys are independent. We represent encryption by a decomposition of the joint probability matrix and construct this decomposition through capacity-controlled randomized rounding. A stepwise conditional sampler turns the rounding process into a correct encryption algorithm without enumerating its output distribution or rejecting complete samples. Let $p_\star$ and $q_\star$ be the largest message and key probabilities. Our construction attains the exact message-recovery optimum $\max\{p_\star,q_\star\}$, has key-recovery success below $2q_\star$, and provides $O(\sqrt{q_\star})$ advantage against prediction of every fixed Boolean message predicate. The semantic guarantee follows from negative correlation of the binary row-rounding remainders. We prove that the key-recovery factor cannot be uniformly improved while preserving exact message-recovery optimality, and that the semantic bound has the optimal worst-case order. We also establish additive robustness under distribution error and short-ciphertext existence by preserving only the moments needed for the security proof. The general sampler runs in expected polynomial time in the explicit rational input size; short ciphertexts do not by themselves imply efficient preprocessing or compact public descriptions. Structured sources admit additional efficient implementations.

In order to estimate accurately the impact of large-scale computing devices, one requires a precise resource estimate of quantum algorithms for cryptographic problems. As a start, these algorithms can be described as quantum circuits. One then estimates the number of gates, qubits, and circuit depth that they require, as a first step towards physical resource estimates. In order to simplify this task, different quantum programming languages have been used, targeting different circuit sizes and abstraction levels. Quantum cryptanalysis is indeed a domain in which one may consider both medium-scale circuits with fine-grained optimizations (such as optimizations of Shor's algorithm) and very large-scale circuits with well-defined building blocks (such as exhaustive key search on block ciphers using Grover's algorithm). This paper presents Qarton, a python library to represent logical quantum circuits, focused on quantum cryptanalysis. It allows to study very large-scale circuits like Grover searches, to simulate classical or ``almost'' classical circuits such as the arithmetic building blocks of Shor's algorithm, and to analyze trade-offs between qubit and gate count when switching between different implementations of arithmetic operations. Qarton is equipped with a library of around 300 circuits and examples from different areas of quantum symmetric and asymmetric cryptanalysis.

To address the scalability and flexibility limitations of dedicated point-to-point cabling in broadcast production, the broadcasting community created a suite of standards for carrying media over IP multicast. This shift removes the inherent physical protection that dedicated cabling provides, exposing streams to tampering and eavesdropping. Such threats have been widely reported and pose major risks to broadcasters and their audiences. Despite almost a decade of deployment, the security of IP-based broadcast has received little academic attention. We collaborate with the European Broadcasting Union (EBU) to establish the security and deployment requirements of this environment. We introduce SafeCast, a system to secure the transmission of ultra-high-bitrate multicast media. SafeCast builds on established and standardized protocols (the MLS group key agreement protocol and the SRTP media-protection protocol), while extending and adapting them to the demanding setting of broadcast media. SafeCast provides strong confidentiality and integrity guarantees, and exploits the precise clock synchronization of broadcast networks to achieve cheap per-sender authentication. We evaluate SafeCast against the stringent deployment requirements of production facilities (throughput, latency, large dynamic groups) and find that it can sustain the most common media formats with ample headroom.

Multiparty private set union (MPSU) enables a group of mutually distrusting parties, each holding a private set, to learn the union of their sets without revealing additional information. MPSU is a fundamental primitive with applications to privacy-preserving tasks such as cyber-risk assessment and private record matching. State-of-the-art MPSU protocols require parties to sequentially shuffle encrypted values in order to hide which party contributed which inputs to the final result. We introduce a new paradigm for semi-honest MPSU that eliminates the need for such shuffling. This allows the protocol to run in constant rounds regardless of how many parties are involved. The resulting protocol also benefits from improved parallelism, an information-theoretic online phase, and native support for unbalanced input sizes, a situation most prior works must handle by padding inputs to hide individual input sizes. We implement and evaluate our protocol across a range of party counts, input sizes, and network conditions. In the unbalanced setting, where parties hold unequal-sized inputs, our protocol significantly outperforms the state-of-the-art schemes PULSE (CCS ’25) and DZBC25 (USENIX Security ’25). In the balanced setting, it achieves an order-of-magnitude improvement over PULSE in most settings and outperforms DZBC25 in WAN deployments with 100ms latency.

Consider a cryptocurrency exchange that secures the bulk of its reserves under a small set of keys, each of which is only used to transfer cryptocurrency once a year; or the backup codes for an account login or a password manager, which are again rarely used but provide access to crucial systems or information. Securing such infrequently-used high-value secrets is crucial, but existing solutions, such as threshold wallets and 'cold' (offline) wallets, are unsatisfactory. In this work, we envision a system that allows users to conveniently back up their rarely-used, high-value keys. This new setting necessitates a novel set of design requirements. Specifically: - We allow user keys to be threshold secret-shared among a large number of custodians where each custodian wallet comprises of a hot (i.e., online) and a cold (i.e., offline) portion. The cold part of the wallet is not touched during the backup process (thus, it is independent of the number of system users) but must be accessed for recovery. - We provide a mechanism to continually assure users that their keys are safely stored. This feature is critical because our system is not designed for frequent key use. We also enable proactive key refresh. - Finally, in our approach, restoring a backed-up key is equivalent to generating a signature. Thus, signatures made by users of this system should look the same as "normal" signatures to avoid exposing holders of high-value keys to targeted attacks. Based on these requirements, we develop new security definitions and a UC-secure protocol that implements threshold BLS signatures in our new model. Our protocol is practically efficient for the envisioned large numbers of custodians: for a 67-out-of-100 threshold configuration, creating a new backup takes 10s, while recovery takes less than 2ms.

In a zkTLS protocol, a user proves to a third party a property of data obtained from an online service over TLS, revealing nothing beyond the statement proven. Each existing scheme, however, comes with its own security model, written for that scheme alone, so the schemes are not proven secure against a common definition. The root cause is a mismatch of levels: existing models are stated at the level of TLS and therefore reflect how each construction handles it. Yet TLS is only the transport, and a zkTLS proof is about the application-layer protocol executed over it. We therefore introduce the first transport-agnostic security model for zkTLS, stated at the application layer and mentioning neither TLS nor any particular construction. The resulting notion, Zero-Knowledge Proof of Protocol Execution (zkPoPE), defines what it means for a party that has executed an arbitrary two-party protocol with a counterparty to prove, in zero knowledge, that its private input and the reply it received satisfy a chosen property, and, crucially, that the reply comes from a real execution with that counterparty rather than being fabricated. This proof-of-execution requirement separates zkPoPE from a NIZK over the input--reply pair: knowing a satisfying pair does not by itself show that an execution occurred. We formalize zkPoPE as a UC ideal functionality F_{zkpope}. We then introduce the Reclaim Protocol, a zkTLS scheme that UC-realizes F_{zkpope} under a deployment-aligned trust model and underlies a production system. Reclaim builds on Distributed-AEAD, a new three-party authenticated-encryption primitive of independent interest. On a lower-mid-range phone, Reclaim attests to a 1200kB response in 7.28s, two orders of magnitude beyond any size previously benchmarked for zkTLS.

The dual attack is a central tool for evaluating the security of the Learning with Errors (LWE) problem, which is the foundation of the lattice-based cryptosystems such as ML-KEM. Since the introduction of FFT acceleration by Guo and Johansson (ASIACRYPT 2021) and subsequent optimizations in the MATZOV report, FFT-based dual attacks have achieved concrete security estimates for ML-KEM. However, these results rely on one specific independence heuristic that is known to fail in certain parameter regimes (CRYPTO 2023). Recent work by Pouly and Shen (EUROCRYPT 2024) initiated the study of provable dual attacks through explicit geometric analysis, but restricted the analysis to the orthogonal lattice, requiring approximately \(m\approx2n\) LWE samples. Qu and Xu (ASIACRYPT 2025) subsequently introduced modulus switching into the provable dual-attack setting, enabling evaluation over a smaller modulus, with a revised analysis given in their later version. More recently, a closely related ePrint revisited the concrete complexity estimates of these provable attacks after correcting the norm-bound formula used in the earlier analysis. In this work, we provide a provable correctness analysis of the Guo--Johansson-style FFT-based dual attack over the general dual lattice. By extending the analysis from the orthogonal lattice to $L_q(\mathbf{A}^{T})$, where $\mathbf{y}$ is not restricted to $\mathbf{0}$, we reduce the required number of LWE samples from $m\approx2n$ to $m\approx n$ while retaining the computational efficiency of FFT-based score evaluation. We further derive an explicit upper bound on the false-positive probability via an integral estimate for a geometric distinguishing event. For concrete complexity estimation, we follow the abstractions and cost models used in prior provable analyses while refining the probability-dependent treatment of the norm bounds. We obtain attack costs of $216$, $317$, and $426$ bits for ML-KEM-512, ML-KEM-768, and ML-KEM-1024, respectively. Compared with the recomputed estimates of the previous best published provable dual attack, these represent reductions of $22$, $30$, and $52$ bits, respectively.

Recent work has introduced the Hamiltonian phase state (HPS) assumptions, which postulate that Hamiltonian phase states can be used to instantiate pseudorandom and one-way state generators. Additionally, it has been conjectured that these assumptions can be true, even if one-way functions do not exist. This is exciting, because if true, then the HPS assumptions provide a route to the instantiation of Microcrypt. In this work we falsify this conjecture, by proving that if the HPS assumptions are true, then one-way functions exist. While this removes the possibility of instantiating genuine Microcrypt cryptography with Hamiltonian phase states, it shows that the HPS assumptions provide novel inherently quantum assumptions for the construction of classical cryptography. Technically we achieve this via a method for the construction of one-way puzzles from one-way state generators and tailored "measure first, ask later" state certification protocols. This generalizes prior constructions of one-way puzzles from one-way state generators via classical shadows and allows us to relate properties of the one-way puzzle to properties of the state certification protocol used in the construction. Specifically, if the state certification protocol admits efficient classical post-processing then one obtains an efficiently verifiable one-way puzzle, and if the state certification protocol can be efficiently classically simulated in a certain sense, then one obtains a classical one-way puzzle, which implies one-way functions. The latter observation allows us to prove that the HPS assumptions imply one-way functions, by exploiting properties of state certification protocols for phase states. The former observation provides a new toolbox for the construction of efficiently verifiable one-way puzzles by exploiting tailored state certification protocols for pseudorandom and one-way state generators.

Secure aggregation with input validation enables a server to securely compute the summation of private data from multiple clients while verifying that the inputs satisfy specific constraints. However, the practical deployment of this paradigm at scale is hindered by two primary limitations. First, existing schemes suffer from severe efficiency bottlenecks due to their reliance on generic, computationally expensive zero-knowledge proofs (ZKPs). Second, the scope of input validation is currently restricted to norm checks (e.g., $L_2$ and $L_\infty$) against public bounds. This paper presents \textsf{Vespa}, a single-server secure aggregation scheme that simultaneously achieves practical efficiency and expands validation capabilities to support both \textit{norm checks against public bounds} and \textit{similarity checks against private server inputs}. Specifically, we design a novel validity-weighted secure aggregation protocol over committed inputs, built upon highly efficient vector oblivious linear evaluation (VOLE). To efficiently realize input validation, we introduce customized $L_2$ and $L_\infty$ norm checks based on VOLE-based ZKPs, centered around optimized range proofs with reduced range sizes. Furthermore, we extend our framework to support advanced similarity-based checks against private server inputs, instantiated with widely used Euclidean and cosine similarities. This allows \textsf{Vespa} to first filter out invalid inputs using norm checks, followed by performing fine-grained validation via similarity metrics. We compare our scheme with the state-of-the-art secure aggregation works, including RoFL (S\&P 2023), ACORN (USENIX Security 2023), and Armadillo (CCS 2025), and extensive evaluation shows that our $L_2$ and $L_\infty$ norm checks achieve runtime improvements of up to $2\sim 3$ and $1\sim 3$ orders of magnitude, respectively. Our similarity checks also achieve practical performance, consuming only around $94$ seconds for an $818$k-entry vector.

Lexical search retrieves documents containing query keywords that require exact matches, such as rare terms, entity names or codewords. Search is increasingly hosted in the cloud, which places both the corpus and every query on infrastructure the data owner does not control. Encrypted lexical search addresses this concern by ensuring no third party, including the cloud provider, learns which queries are issued or which documents are retrieved. The key binding constraints are all on the client — accuracy, memory, computation, and communication — since cloud capacity is elastic. Existing approaches that guarantee full privacy without caveats require prohibitively high client-side persistent storage or communication. We present Lexicure, a protocol for encrypted lexical search that reduces the search to arithmetic operations favorable to cryptographic primitives, such as encrypted matrix–vector products (EMVP). Across sixteen diverse datasets, Lexicure retains over $95\%$ of plaintext BM25 accuracy on average. Lexicure sets a new baseline for leakage-free lexical search, requiring 2.6-9.9$\times$ less communication than alternative approaches and less than 1KB of client-side storage. A major challenge is that secure lexical search is inherently bandwidth-bound. We additionally describe an approach for compressing the communication further in exchange for a larger server. To do so, we introduce a new cryptographic primitive that modifies EMVP to securely compute quadratic and higher-degree functions, more efficiently than a direct reduction. Lexicure demonstrates that leakage-free encrypted search with practical performance is achievable and opens new avenues for secure information retrieval systems.

Unbalanced Oil and Vinegar (UOV) and its variants are among the candidates in the NIST call for additional post-quantum signatures. Their security rests on the hardness of recovering a hidden subspace on which a public system of quadratic forms vanishes. In characteristic $2$ the polar forms of such a system are alternating, which a recent attack exploits by working in the exterior algebra rather than the polynomial ring. Unlike its polynomial-ring counterparts, it either applies at a threshold fixed by the parameters or not at all, and for most proposed parameter sets it does not. We remove this threshold and improve complexity in two independent ways. First, over the integers the square of an alternating polar form is $\omega \wedge \omega = 2\,\omega^{(2)}$, so in characteristic $2$ it vanishes only because the factor $2$ does. Dividing that factor out before reducing leaves the divided square $\omega^{(2)}$, which still vanishes on the oil space but is in general not a consequence of the original equations. Iterating gives a whole tower $\omega^{(2^j)}$ of new equations. Second, the symmetric powers $\textrm{Sym}^d\mathbb{F}_q^n$ and the exterior powers $\bigwedge^a\mathbb{F}_q^n$ are the two extreme shapes of the Schur modules $L_\lambda\mathbb{F}_q^n$. We introduce an extension of Macaulay matrices to an arbitrary shape $\lambda$ which we call Schur--Macaulay matrices. For both constructions we conjecture a Hilbert series and verify it experimentally. We also extend the approach to odd characteristic. The attack then applies to every Round-2 parameter set of UOV, MAYO, SNOVA, and QR-UOV but one. Four of the alternate Round-2 parameter sets of MAYO fall below their claimed security level while being out of reach of all previous attacks. After disclosing preliminary results to the MAYO team, a parameter set designated for Round-3 was discarded and parameters increased. For SNOVA, the divided powers remove the spurious kernel elements that obstructed the exterior algebra attack, and seven of the eleven Round-2 parameter sets fall below their claimed security level, and the complexity estimates drop further compared to previous attacks.

A Gaussian leftover hash lemma (LHL) states that, for a matrix $\mathbf{X}$ with discrete Gaussian columns and a sufficiently wide Gaussian vector $\mathbf{v}$, the product $\mathbf{X}\mathbf{v}$ is close to a discrete Gaussian. Over the integers this is a classical tool, and Albrecht-Felderhoff-Lai-Lapiha-Woo (EUROCRYPT'26) recently extended it to modules over a number field $K$ of degree $d$. In every version, the lemma rests on two facts about the kernel lattice of $\mathbf{X}$: that $\mathbf{X}$ is surjective, and that the smoothing parameter of the kernel is small. We prove both facts with a single analytic estimate. An exact identity expresses the Gaussian mass of the dual kernel, multiplied by the index of the image of $\mathbf{X}$, as an integral of one-dimensional periodic Gaussian sums over a torus. Bounding this integral below two forces the index to be one, which is surjectivity, and bounds the smoothing parameter at the same time. No separate surjectivity argument and no GRH are needed. For smoothing error $2^{-\lambda}$ the smoothing parameter is $O_K(\sqrt\lambda)$ with explicit constants, once $\mathbf{X}$ has $m\gtrsim dr\ln(srm)$ columns, and a deterministic lower bound shows that this order is optimal. The width $s$ of the matrix enters only through the number of columns. Over power-of-two cyclotomic fields the theorem even applies to matrices whose integer coefficients follow a discrete Gaussian of width one. In the SIS-to-$k$-SIS reduction of Albrecht et al. (EUROCRYPT'26), the new bound lowers the hint width from $d^{39/2}\lambda^{9/2}$ to $d^{3/2}\lambda^{3/2}$ and the norm loss from $d^{41/2}\lambda^{5}$ to $d^{3/2}\lambda^{2}$, up to logarithmic factors. Additionally, we also formalized the main results of the paper in Lean.

VDOO is a Rainbow-like multivariate signature scheme with additional ``diagonal” structure intended to resist structural attacks. We show that its security is governed primarily by the first two oil layers, while the diagonal structure has negligible impact, and that all proposed parameters fail to meet their claimed NIST security levels. We provide a corrected complexity analysis of the Simple and Rectangular MinRank attacks, introduce a hybrid Simple–Padded MinRank attack, and compare their effectiveness. Open-source code is provided for independent validation, clarifying VDOO’s true security and informing future Rainbow-type post-quantum designs.

Accommodating cryptographic authenticity with strong user privacy assurances has been the primary motivation for anonymous credentials systems. Their features have recently come into the spotlight with the European Digital Identity (EUDI) wallet initiative, insisting on the need for efficient and private solutions based on well-understood security foundations for high assurances. This coincides with the post-quantum transition, but current quantum-safe solutions based on standard assumptions are still lagging behind the ones on ad-hoc interactive assumptions performance-wise. In this paper, we present several techniques to improve the efficiency of anonymous credentials from standard lattice assumptions, narrowing the gap with more efficient but also more exotic ones. Alongside other optimizations, our main contributions focus on the zero-knowledge protocol of Lyubashevsky, Nguyen, Plançon (Crypto'22), currently the efficiency bottleneck of lattice privacy-oriented constructions, for which we identify several sources of improvements that may be on independent interest.

Pseudorandom codes (PRCs) are error-correcting codes whose codewords are computationally indistinguishable from uniformly random strings to any efficient adversary. Most prior PRC work focuses on Hamming error channels; nevertheless, the few proposed constructions on edit channels are not optimized for a stochastic channel with both Hamming and synchronization errors. We study PRC over a stochastic memoryless Binary Substitution-Insertion-Deletion (BSID) channel and its generalized $q$-ary version. At each current input symbol, the channel inserts a uniformly random symbol without advancing with probability $P_i$, deletes and advances with probability $P_d$, or transmits and advances otherwise; a transmitted symbol is then replaced uniformly by another alphabet symbol with probability $P_s$. We first prove that the majority based PRC construction of Christ and Gunn remains robust over the BSID channel whenever $1-P_i-P_d>0$ and $P_s<1/2$, by proving its majority decoding after the BSID channel recovers more than half of the original coordinates by a constant margin with overwhelming probability. Then we propose a unique-decodable public-key multi-bit PRC on a constant alphabet constructed by letting the Christ--Gunn majority component protect control information and the Haeupler--Shahrasbi synchronization/list-recovery construction encodes the high-rate payload. For every sufficiently large admissible constant alphabet q depending on the channel parameters, the PRC is robust against a q-ary SID channel bounded by $P_i+P_d<1$, $P_s<$1, and its rate approaches the genie-aided capacity upper bound of the channel when q approaches infinity.

Round-reduced variants of AES are widely used as building blocks in the design of cryptographic schemes. The study of non-random properties and distinguishers for round-reduced AES has always been an important research topic. The longest known secret-key distinguishers on AES cover 6 rounds. Related differences and related differentials were introduced by the designers of AES in 2009, but research in this direction remains limited. In this paper, we provide a new perspective on related differences through exchange and shift operations. Based on related differentials, we present new non-random properties and secret-key distinguishers for up to 7 rounds of AES. For 5-round AES, we present a new property with probability $2^{-22}$ by combining one-round byte-wise related differentials with the 4-round zero-difference property. Then we improve the secret-key distinguishers on 5-round AES in both the chosen plaintexts (CP) and adaptively chosen plaintexts (ACP) settings, achieving data/time complexities of $2^{27.2}$/$2^{28.05}$ and $2^{23.32}$/$2^{23.54}$, respectively. For 7-round AES, we identify the first non-random property by exploiting one-round exchanged diagonal related differentials and combining them with the 4-round related differentials given by Bardeh and Rijmen. Then, we propose the first secret-key distinguisher for 7-round AES with data complexity lower than the full codebook. For 6-round AES, using shifted diagonal related differentials, we present an alternative distinguisher that is dual to the exchange-attack distinguisher from ASIACRYPT 2019.

Wide-block block ciphers have attracted significant interest due to emerging high-throughput encryption requirements and NIST's Accordion Mode Process. The primary choice for such ciphers is Rijndael-256, also known as WAES. At the same time, independent efforts continue to explore more efficient designs and implementations. In this paper, we present LEANY, a 256-bit block cipher with a 256-bit key. LEANY is a nibble-oriented SPN whose state is an 8×8 matrix of 4-bit nibbles. Each round consists of four layers: key addition, a 4-bit S-box layer, row-wise nibble shifts, and multiplication by an 8×8 MDS matrix over $\mathrm{GF}(2^4)$. This structure follows the same shift-then-mix principle as AES, but with a wider state and a smaller cell size. LEANY consists of only 12 rounds. Since the linear diffusion layer has branch number 9, the wide-trail strategy guarantees that every four consecutive rounds activate at least $9^2=81$ S-boxes in any differential or linear characteristic. Consequently, upper bounds on the best differential and linear trail probabilities fall below $2^{-256}$ after seven rounds. We further analyze LEANY against impossible differential, invariant subspace, integral, meet-in-the-middle, boomerang, and several algebraic attacks. Finally, FPGA implementations demonstrate both compact and high-performance hardware realizations: the iterative architecture occupies only about 1.4K LUTs, whereas the fully pipelined architecture achieves a steady-state throughput exceeding 50 Gbps. Compared with existing 256-bit wide-block block ciphers, including WAES, LEANY provides substantially better area efficiency and competitive throughput (TP) and throughput-to-area ratio (TAR) results.

Spending limits are the workhorse of custody policy, yet in every deployed threshold-signing stack they live outside the signing protocol—in centralized policy engines, on-chain covenants, or hardware co-signers—where they leak budget state, bind to a single chain, or add a trusted bypass surface. The workload is compounding: agent payments passed one hundred million transfers on a single chain in under a year. We initiate the study of protocol-native spending limits: per-epoch budgets secret-shared among the signers and enforced inside the signing session, with unchanged on-chain verification. We first show that the natural approach fails: gating a threshold signature with a passively secure MPC comparison admits a deterministic bypass. For the class of linear gatein-combination protocols we prove that every member signs over-budget transactions with probability one, disguised against any transcript auditor up to the statistical masking bound. We then construct BTS, a budgeted threshold Schnorr scheme with a formal budget-soundness definition for honest-majority configurations. The gap is closed with an authenticated gate: in implementation, every attack attempt is caught before a signature is produced, at a measured cost of 3.4–3.7× multiplications and five extra rounds. Batched over k=20 payments, enforcement amortizes to about one network round trip per payment in our WANmodeled prototype.

Cashu, the dominant deployed Chaumian ecash system, has grown to 31 specification documents, 58 indexed mints, and seven interoperable wallets, while the cryptographic literature has studied only its blind-signature core. Every incident in the system’s short audited history, including a cross-mint theft disclosed in early 2026, is compositional: it emerges from keyset rotation, wallet recovery, and multi-mint identity. We give the first system-level security analysis of a deployed ecash system, built on a five-layer model, a nine-property family (three properties new), and a stateful one-more-unforgeability theorem whose proof separates forgery from theft; the deployed attacks commit theft without forging anything. We reproduce the disclosed theft end-to-end, exhibit a strictly stronger public-key-claim variant, and measure all 58 indexed mints: every reachable mint declares support for DLEQ proofs, the sole remaining defense, yet declarations are unauthenticated the guardrail sits with the issuer, not the wallet. A theft theorem gives necessary and sufficient conditions for prevention, and replaying three fixes shows the cheapest one costs honest issuers nothing. A defect taxonomy and a three-system comparison suggest a broader lesson: recoverability is not free. All artifacts are pure standard-library Python and pass the 49 o cial test vectors.

Upgrading cryptocurrencies to post-quantum security is inherently difficult, because their on-chain state associates users' holdings with public keys: users who fail to transfer their funds to (yet to be added) post-quantum signature schemes remain exposed to theft. In networks with Confidential Transactions (CT) such as Liquid, Monero, and Grin, which hide monetary amounts inside homomorphic Pedersen commitments, the situation is worse still: an adversary who computes a single discrete logarithm can inflate the currency as a whole, rather than merely steal individual funds. The contentious remedy is to expire all funds not transferred by a deadline, destroying the funds of everyone who fails to act. Switch commitments (Ruffing and Malavolta 2017) were proposed to avoid exactly this expiration, but only Grin has deployed them deliberately—and the security theorem later stated for its scheme does not hold, as we show. In this paper, we show that Liquid and Monero already deploy a switch commitment scheme by coincidence: the ciphertext a sender stores on-chain to convey the hidden amount to the recipient can be re-interpreted, together with the Pedersen commitment, as a switch commitment. We conceptualize this construction as coincidental switch commitments and prove it secure in the quantum random oracle model. Additionally, we save the construction that Grin deploys: it is computationally switch binding, with practical bounds in the algebraic group model. The real-world implication of our results is that these networks can be upgraded to post-quantum CT schemes without expiring any user funds.

Arithmetic logic units (ALU) combine word-level arithmetic with bitwise operations over encrypted data. The state-of-the-art scheme, GZ26 (CRYPTO'26), constructs an ALU in CKKS via triangle encoding, but has two limitations: the radix-$2$ design restricts packing capacity, while the arithmetic-to-Boolean (A2B) conversion requires $O(\ell)$ functional bootstrappings for $\ell$-bit message without batching independent ciphertexts. In this paper, we first generalize triangle encoding to arbitrary integer bases while preserving leveled modular arithmetic in CKKS. Larger bases reduce the number of slots required per message, allowing more messages to be packed into each ciphertext. %To ensure correct decoding despite rounding during CKKS plaintext encoding, we derive sufficient conditions relating the digit base to the encoding scale. For power-of-two digit bases, we introduce an A2B conversion using two functional bootstrapping after overflow refreshing, without batching independent ciphertexts. Our key insight is to transform triangle encoding into radix representation with unknown carries, from which candidate bits can be extracted using a single multi-value functional bootstrapping. Specifically, we first cancel the overflow shared by neighboring windows using one functional bootstrapping followed by linear operations, leaving expressions involving only radix digits and adjacent binary carries. Then we use a multi-value functional bootstrapping to jointly extract candidate bits and evaluate the local carry state, which determines how carries propagate between windows. Finally, we recover the canonical bits through parallel carry propagation and bit correction, using only leveled homomorphic operations. We implement our scheme in OpenFHE and evaluate it for 64-bit, 128-bit, and 256-bit message. The experimental results show that using \(d=2\) achieves the lowest latency for A2B conversion, yielding \(3.49\times\)--\(9.71\times\) speedups over GZ26 at the same packing capacity. In addition, using \(d=32\) increases the packing capacity and achieves the best amortized performance, yielding \(8.15\times\)--\(28.47\times\) speedups.

Token-weighted governance can fingerprint individual voters: when weights are unequal and externally known, the exact weighted tally may reveal who voted which way. Publishing only the winning outcome hides the tally from the public, but a system that reconstructs the tally in order to decide still hands it to whoever runs the computation. Publishing only the winner is not the same as computing only the winner. We present a verifiable protocol for weighted DAO votes in which no protocol role ever receives the weighted tally in plaintext. Each ballot proves in zero knowledge that its encrypted contribution is exactly zero or the voter’s registered weight. Trustees add accepted contributions and compare the sum with a threshold entirely under encryption using the Cortier–Gaudry–Yang (CGY) conditional-gate toolbox, and decrypt only the final decision. A public verifier checks that the computation consumed exactly the ciphertexts accepted by the governance contract. We prove accepted-result correctness even when all trustees are malicious; restricted-I/O SUC security of the tally against static active corruption of any sub-threshold trustee coalition, which is any four of five in our reference instance; and winner-only disclosure relative to the real sealed board. With 1,024 voters, the encrypted tally takes 157 s and independent replay 115 s, while ballot proving dominates the 28.5-minute end-to-end pipeline. A fresh Arbitrum Sepolia execution exercises the complete public-chain path

SQIsign samples a random ideal of odd prime norm at every key generation and every signature through $\mathbf{PrimeNormIdealInertSampling}$. We revisit this step from a new perspective: the inert ideals of a given norm form a circle modulo the norm, and drawing lines through a fixed point of this circle turns a single random parameter into an ideal, hitting each ideal exactly once. Compared with the reference implementation, the new algorithm requires one modular inversion and a handful of additions and multiplications. It achieves $13.76\times$ speed-up at NIST security level I in SQIsign while producing an uniform output.

Quantum linear key-recovery attacks using QFT-based correlation state preparation can achieve superquadratic speedups, but their complexity grows rapidly with the active subkey dimension, potentially erasing the advantage over Grover's search. WSP (EUROCRYPT 2024) provides a natural mechanism for alleviating this bottleneck by replacing the key-recovery map with a spectrally punctured approximation of smaller active dimension. In this work, we establish a puncture-aware framework for QFT-based quantum key recovery, extending quantum linear cryptanalysis to key-recovery problems whose active-subkey dimension can be reduced through spectral puncturing. We further propose Q-Puncture, an automated optimization method tailored to quantum key-recovery complexity. Applied to Serpent, RECTANGLE-128, and PIPO-128, Q-Puncture yields a new family of punctured quantum linear key-recovery attacks whose modeled quantum complexities improve upon generic Grover search. The resulting attacks improve upon comparable QFT-based results, extend them to new settings, and provide further examples in which structured quantum cryptanalysis surpasses the generic quadratic speedup of Grover's search.

Falcon is one of the three post-quantum signature schemes selected by NIST for standardization so far. It is efficient and has particularly short signatures, but has the drawback of being difficult to implement correctly and securely, owing to its use of floating point arithmetic and secret-dependent lattice Gaussian sampling. In particular, obtaining implementations of Falcon that are protected against side-channel attacks seems to be a tremendous challenge that has no satisfactory solution so far. While it is theoretically possible to rely on generic countermeasures like masking, they are likely so costly as to be effectively inapplicable. And yet, side-channels are an actual threat to Falcon: a number of them have been demonstrated against various parts of the signing algorithm, including floating point operations and Gaussian sampling. Multiple power analysis attacks in fact achieve full key recovery from very realistic numbers of traces. This is especially the case for a series of side-channel attacks that can be broadly categorized as based on “sign leakage” within a specific part of the one-dimensional Gaussian sampler. In this paper, we take a closer look at this class of side-channel attacks, and obtain yet another one of a somewhat different flavor: while it also targets the sign of a variable in the one-dimensional Gaussian sampler, that variable is the input center, and in particular does not depend on the sampling randomness, making previously proposed mitigations modifying the base sampler ineffective. Furthermore, we propose, analyze and experimentally validate a novel, inexpensive countermeasure to protect Falcon against this class of side channel attacks. The idea, taking inspiration from shuffling, is to twist the target vector by random roots of unity in every recursive call of the fast Fourier sampling algorithm: this averages away the sign leakage efficiently. While this mitigation technique is heuristic and does not offer the provable guarantees of masking, it significantly enhances the physical security of Falcon fairly cheaply. In particular, it provides a more robust and more widely applicable protection than the countermeasure of Lin et al. (PKC 2025) at a much lower cost (incurring a 1.4-fold overhead in signing in dynamic mode and 2-fold in tree mode, compared to 3.5-fold and 6.8-fold for Lin et al. respectively). To the best of our knowledge, this is the first shuffling countermeasure proposed for Falcon.

Pseudorandom error-correcting codes (PRCs), introduced by Christ and Gunn at CRYPTO’24, combine pseudorandomness with error correction, providing a natural abstraction for robust watermarking and steganography on generative AI models. Subsequent standalone notions, which are ideal security for secret-key PRCs and CCA-style security for public-key PRCs, are oracle-based and do not capture composable use with explicit parties, sessions, and corruption. We give a UC treatment of PRCs via corruption-aware ideal functionalities for both settings. Under non-adaptive corruption, the UC notions recover the standalone ones. Under adaptive corruption, we identify a common obstruction: dummy codewords sampled before corruption, together with their neighborhoods, must later be opened as valid PRC codewords. We formalize this as a decoder-non-committing code (NC-PRC), which any adaptively UC-secure realization must induce. We then capture failures of such openings via targeted low opening capacity, show it rules out robust NC-PRCs, and prove that LDPC-based PRCs have this property hence do not admit a NC-PRC. On the positive side, we sketch two compilers to lift error-correcting codes to admit NC-PRCs: a secret-key one from a puncturable PRF and indistinguishability obfuscation (iO), and a public-key one from a smooth projective hash function (SPHF), both evading the barrier via programmable acceptance. Finally, we identify a fresh-codeword explanation barrier for public-key PRCs: accepted unseen codewords cannot be explained from public information without violating pseudorandomness, hence public-key UC realizations require a trapdoor or an idealized setup, and PRC-based watermark detection is inherently designated-verifier.

Designing Authenticated Key Exchange (AKE) directly from the Learning with Errors (LWE) assumption, without relying on generic constructions, remains a prominent and challenging goal in post-quantum cryptography. Such designs inherently require reconciliation mechanisms, which may introduce additional leakage against active adversaries under key reuse. Yet previous signal leakage attacks of Bindel et al. (LATINCRYPT 2021) and Qin et al. (ESORICS 2022) fail against protocols hardened with dedicated countermeasures. In this paper, we propose new signal leakage attacks capable of compromising LWE-based (authenticated) key exchange, where prior attacks fail. First, we target the landmark MQV-style AKE protocol proposed by Zhang et al. (EUROCRYPT 2015), known as ZZDSD-AKE. Determining whether it can natively achieve security under the strong extended Canetti-Krawczyk (eCK) model has remained a long-standing open question. By exploiting the ephemeral key compromise defined within the eCK model, our new attack based on a geometric perspective recovers the static secret key from signal leakage, definitively resolving this open question with a negative result. Furthermore, we extend our methodology to the GDLL-KE protocol of Gao et al. (IEEE TC 2018), which is designed to be robust against key-reuse via a randomized noise countermeasure. Our attack shows that its ad-hoc countermeasure fails to eliminate secret-dependent leakage. Our implementations demonstrate that the proposed attacks recover the secret key in roughly 1,700 queries against ZZDSD-AKE and 180 queries against GDLL-KE, respectively. These results reveal that MQV-style structures and simple randomization are insufficient for securing AKEs directly based on LWE against sophisticated active attacks, providing critical insights for future protocol design and standardization.

Efficient execution of Post-Quantum Cryptography (PQC) in resource-constrained embedded environments requires the joint optimization of not only algorithm vectorization but also dataflow and instruction execution order. This paper implements the HAETAE-2/3/5 and SMAUG-T1/3/5 parameter sets of two Korean Post-Quantum Cryptography Competition (KpqC) winners using the M-profile Vector Extension (MVE) of the Arm Cortex-M55 and analyzes the contribution of each optimization stage to performance improvement. We first establish an MVE-intrinsic implementation as the baseline and then present Manually Optimized MVE Assembly that redesigns the data layout, register reuse, and memory access patterns of the principal kernels. Slothy is subsequently applied to this assembly to optimize instruction order and register allocation for the Cortex-M55 pipeline. This process is applied to the Keccak/SHAKE operations shared by the two schemes; the Number-Theoretic Transform (NTT), fixed-point Fast Fourier Transform (FFT), and Gaussian sampling of HAETAE; and the Toom-Cook/Karatsuba polynomial multiplication of SMAUG-T. Measurements on the STM32N657 show that the implementation with Slothy achieves speedups of 2.349–2.424× for the HAETAE forward NTT, 1.701–1.732× for the FFT including squared-value accumulation, and 1.322–1.350× for the SMAUG-T matrix–vector multiplication over the intrinsic implementation. At the complete-scheme level, HAETAE and SMAUG-T achieve speedups of 1.050–1.212× and 1.108–1.212×, respectively. These results demonstrate that direct dataflow redesign and automated instruction scheduling eliminate different bottlenecks and that combining the two stages can effectively improve the performance of Cortex-M55-based PQC implementations.

AIMer v3 replaces AIM2 with AIM3. We test whether SIMD methods developed for AIM2 can accelerate AIM3 while preserving its results. We port the field-arithmetic, affine, party-parallel, and SHAKE methods of the earlier AVX-512 implementation to AIMer v3 and implement the same operations for AVX2. For Cortex-M55 MVE, we use 16-bit polynomial partial products for field multiplication, group corresponding 32-bit words from four parties, and vectorize both affine layers, including the input layer introduced in AIM3. The implementations preserve AIMer v3’s mathematical structure and proof procedures. For all six parameter sets, their outputs match the official known-answer tests and the reference results in differential tests. Relative to the reference on each platform, AVX2 and AVX-512 accelerate signing and verification by 3.14–4.41× and 4.59–6.55×, respectively. Cortex-M55 MVE accelerates the same operations by 1.28–1.91×.

SNARKs based on error-correcting codes require codes that simultaneously support fast encoding and large relative distance. Reed--Solomon codes achieve the optimal rate--distance tradeoff given by the Singleton bound, but their fast encoding relies on FFT-friendly fields, limiting their applicability to field-agnostic constructions. In this work, we revisit expand--accumulate (EA) codes, a simple family of linear codes with efficient encoding over arbitrary fields. Although extensively studied in prior work, existing distance guarantees are weak, and practical instantiations have relied on conjectured results. We prove that EA codes with exact-weight expansion matrices, over sufficiently large fields, achieve a rate--distance tradeoff that approaches the Singleton bound with high probability, resolving these conjectures. Building on this result, we construct \textsf{Flare}, a field-agnostic polynomial commitment scheme based on EA codes. At its core is an efficient IOPP for EA codes that leverages code switching. For statements of size $M$, \textsf{Flare} achieves $O(M\log M)$ prover time and $O(\lambda\log^2 M)$ proof size, improving upon the $O(\sqrt{\lambda M})$ proof size of prior constructions based on EA codes (Block et al., CRYPTO 2024). Comprehensive evaluations show that \textsf{Flare} achieves compact proofs while maintaining competitive prover efficiency. For proving circuits of size $2^{22}$, \textsf{Flare} produces the smallest proofs of approximately $16$~MiB. Compared with BaseFold (Zeilberger et al., CRYPTO 2024) and ERA (Baweja et al., ASIACRYPT 2026), it reduces proof size by $1.66\times$ and $1.47\times$, respectively, while maintaining comparable prover efficiency. Compared with Brakedown (Golovnev et al., CRYPTO 2023), \textsf{Flare} produces $4.85\times$ smaller proofs at the cost of only a $1.50\times$ slower prover.

We present new constructions of the following primitives featuring succinct communication: - Batch threshold encryption: Assuming RSA and DCR, we construct a threshold encryption scheme in which a batch of ciphertexts can be decrypted via short hints from a quorum of decryptors. Previously, this was known only from bilinear maps. - Aggregate signatures: Assuming RSA, we construct a signature scheme where signatures on different messages from an arbitrary number of signers can be aggregated into one signature. Our construction achieves bounded-query security. In contrast to prior (fully secure) schemes in the standard model, ours makes black-box use of cryptography. - Broadcast encryption: Assuming RSA, we construct a secret-key encryption scheme where a sender can encrypt a message to multiple recipients via a short ciphertext. Our construction achieves bounded-query security. Previously, (fully secure) broadcast encryption was known from bilinear maps or obfustopia or nonstandard lattice assumptions. Our constructions are simple and achieve optimal succinctness in the key parameter (decryption batch size or number of signers or recipient set size). In this metric, they outperform or match known constructions. Our results are obtained by applying the lens of projective pseudorandom generators (PPRGs) to these problems. Towards this, we identify new enhanced properties for PPRGs and instantiate them from the RSA assumption.

The dual attack is a central tool for evaluating the concrete security of the Learning with Errors (LWE) problem, which underlies lattice-based cryptosystems such as ML-KEM. The MATZOV-style FFT dual attack combines dual distinguishing with an FFT-accelerated guessing phase and provides one of the main frameworks for concrete attack estimates against ML-KEM. However, its correctness analysis has traditionally relied on heuristic independence assumptions, leaving a gap between practical attack estimates and provable guarantees. In this work, we give a rigorous correctness analysis of a MATZOV-style FFT-based dual attack for the ML-KEM parameter sets. The analysis treats the correct and incorrect candidates separately and applies different conditioning arguments on the two sides, because the former requires a provable lower bound on the score of the true candidate, while the latter requires a uniform upper bound for all wrong recovered guesses. Combining these two parts, we obtain an explicit success criterion and a provable threshold separation for the FFT score. For concrete cost estimation, we follow the same cost-estimation convention as previous provable dual-attack analyses whenever the same subroutine is involved, using Pouly and Shen's estimates (EUROCRYPT 2024) as our main baseline. We obtain attack costs of 210, 300, and 410 bits for ML-KEM-512, ML-KEM-768, and ML-KEM-1024, respectively. Relative to this baseline, our estimates reduce the attack costs by 28, 47, and 68 bits for ML-KEM-512, ML-KEM-768, and ML-KEM-1024, respectively.

We present Argo, a randomized encoding scheme for the invalidity of pairing-based SNARKs like Groth16. Argo allows an encoder to encode a proof such that a decoder can derive a secret if the proof is invalid. We prove the selective security of invalidity Argo, as well as the adaptive security of validity Argo, under DDH in one source group of a bilinear pairing in the standard model, a standard assumption that holds in the generic group model used to analyze Groth16. We show how to couple invalidity Argo with a projective garbling scheme like Argo MAC or Duty-Free Bits to construct a garbled circuit for Groth16 invalidity as a function of the proof bits, and how to use it with BitVM to construct a permissionless, optimistic Groth16 verifier for Bitcoin: a decoder proves a proof invalid by publishing a secret, and the decoder can be anyone, not necessarily known in advance. When instantiated with BN254, Argo has an offline encoding size of about 21 kB, a thousand times smaller than the projective garbling scheme, and an end-to-end protocol size more than three orders of magnitude smaller than existing Yao-based approaches. Generating an Argo + Argo MAC garbled circuit takes tens of milliseconds on a laptop, compared to minutes for a Yao GC. Argo is the first protocol that makes permissionless, optimistic SNARK verification on Bitcoin practical.

DuX and YuX are two recent families of FHE-friendly block ciphers, designed for efficient evaluation under fully homomorphic encryption. Both keep a state of sixteen words of a large finite field, arranged in four blocks, and each round applies a block-wise S-box of low degree followed by a circulant linear layer. We show that, in the chosen-ciphertext model, the algebraic degree of their decryption functions grows far more slowly than the designers' evaluation of the encryption direction suggests, and we turn this observation into practical key-recovery attacks. The starting point is a sufficient criterion for zero sums. It treats affine subspaces in characteristic 2, full prime fields, multiplicative cosets and arbitrary point sets in one statement, and over prime fields it is tight on every cell we could compute exactly. Three techniques turn the criterion into attacks. Full block structures make the first S-box layer free. Cheap coordinate elimination extracts equations from states that are only partially balanced, helped in characteristic 2 by a second cheap coordinate that exists only there. Weighted moments yield thousands of equations from a single structure, and weights with several indices take the place of further structures. We recover the master key of the full twelve-round DuX over $\mathbb{F}_{65537}$ from $2^{29}$ chosen ciphertexts and over $\mathbb {F}_{2^{16}} $ from $2^{32}$, in both cases in complete runs on random keys, the former in $20$ hours on one processor core. For YuX, we recover the key of eleven of the fourteen rounds of YupX-65537 and of Yu2X-16 from $2^{32}$ chosen ciphertexts, again in complete runs, where the best previous attack on Yu2X-16 needed $2^{96}$. Our distinguishers on DuX coincide with those of concurrent and independent work by Liu and Sun (eprint 2026/1907). Furthermore, our key recovery attacks lower the data of their full-round attacks from $2^{67.59}$ and $2^{67.32}$ chosen ciphertexts to $2^{29}$ and $2^{32}$.

Updatable Public-Key Encryption (UPKE) enables forward secrecy in asynchronous settings like secure group messaging. At EC'25, Albrecht, Benčina and Lai proposed the first plausibly post-quantum UPKE scheme supporting unlimited updates, based on permutation code equivalence (PCE) over finite fields and hollow lattice problems. However, it has impractically large public keys (4.8MiB) and ciphertexts (1.5MiB). In this work, we first replace the information-theoretic Hollow LHL with a computational assumption secure under binary-secret binary-error LWE, reducing key sizes by an order of magnitude. Second, we show that generalising to Module-LWE is not beneficial: PCE instances over $\mathbb{F}_{q^d}$ admit a linearisation attack via the $q$-Frobenius structure, enabling recovery of secret isometries unless code rates are inflated by $d$, negating any size benefit from module lattices. Third, we establish IND-CCA security in the model of Alwen, Fuchsbauer and Mularczyk (EC'24), but do not achieve `joiner security' like other post-quantum candidates. Finally, we introduce a heuristic multi-key variant using matrix-valued secret keys, enabling efficient encryption of $\ell$ bits per ciphertext. We define the Hinted SPCE assumption to capture the main cryptanalytic challenge and show that the performance gain is likely too small to justify adoption, unless the new assumption can be avoided. Overall, our `unlimited' scheme has update sizes 3.5 to 10 times larger than lattice-based schemes limited to $2^{20}$ updates, depending on the aggressiveness of parameter choices.

We give an efficiently computable invariant we call the square class invariant for the (Signed) Permutation Code Equivalence ((S)PCE) problem, that was previously not recognised in coding theory, which arises from the arithmetic of quadratic forms over finite fields. Our invariant naturally yields a distinguisher for the decision version of (S)PCE as defined at Eurocrypt 2025 by Albrecht, Benčina and Lai [ABL25], breaking the hardness assumption that underpins the security of their updatable public-key encryption scheme. We extend a 2023 result by Bruin, Ducas and Gibbons by fully classifying the genera of $q$-ary lattices, showing the genus is completely determined by the hull dimension and our square class invariant of the underlying linear code, and that none of these genera split non-trivially into spinor genera (as soon as the lattice dimension is at least 5), implying the genus of a $q$-ary lattice encodes all known efficiently computable coding-theoretic invariants for (S)PCE and vice versa. Thus our distinguisher can also be seen as comparing the genera of Construction A lattices of the (S)PCE instance in the spirit of the Lattice Isomorphism Problem. We give a complete description of the genus distribution of uniform $q$-ary lattices. This motivates the definition of a genus of a linear code as the genus of the Construction A lattice of any of its generator matrices. We adapt the sampling algorithm from [ABL25] to sample from a single genus uniformly at random and can thus restrict their hardness assumption for (S)PCE to a single genus. Restricting PCE to one genus and using our sampling algorithm is then used with a slight modification to the security proof to mend the scheme from [ABL25].

Oblivious RAM (ORAM) protocols enable a resource-constrained client to outsource data to an untrusted server, and access this data without revealing what is being accessed. While ORAM protocols are almost optimal, concretely efficient ORAM protocols incur sub-optimal bandwidth overhead and require multiple roundtrips of communication between the client and the server. In this work, we introduce write-public ORAM: a relaxed variant of ORAM that hides what data is being read, but not what data is being updated on the server. We present wpORAM, a write-public ORAM that modifies Path ORAM to achieve a log N-factor improvement in bandwidth and roundtrips over Path ORAM during writes. The asymptotic cost of reads in wpORAM remains the same as Path ORAM, with a slight slowdown in practice. wpORAM, when used with a Trusted Execution Environment (TEE), can implement a dynamic Private Information Retrieval scheme that efficiently updates data stored in a public database. We use wpORAM to implement a private blockchain RPC server that enables blockchain clients to query RPC nodes without revealing their queries to the server operator. Our evaluation shows that, while syncing with the latest blockchain state, our RPC node achieves a 3-4x speedup over Path ORAM. Our RPC node can serve reads with a slowdown of approximately 23% compared to Path ORAM.

Wiener's continued-fraction attack gives the classical provable bound \(d<N^{1/4}\) for balanced RSA. Boneh and Durfee reached the exponent \(1-\sqrt{2}/2\approx0.2929\), but their argument relies on a heuristic independence assumption. We prove the first fully provable improvement beyond Wiener's \(1/4\) exponent: for every fixed \(\varepsilon>0\), balanced RSA with \(e=\Theta(N)\) can be factored deterministically in polynomial time whenever \(d\leq N^{11/42-\varepsilon}\).

A pseudorandom code (PRC) is a cryptographic primitive combining pseudorandom encryption and error-correcting codes (Christ and Gunn, CRYPTO 2024). That is, any polynomial number of codewords are pseudorandom, and decryption tolerates a constant rate of bit-flips. Although this might seem like a natural combination of standard properties, it is a fundamentally new object: PRCs are black-box separated from many primitives, including one-way functions, public-key encryption, and even virtual black-box obfuscation (Döttling et al., Garg et al., TCC 2025). Pseudorandom codes were introduced to build watermarks for AI-generated content. In particular, they imply watermarks that preserve the model's quality in a strong sense and withstand modification of the content. This relationship makes pseudorandom codes an inviting area for researchers: improvements to this simple theoretical object yield better watermarks. Beyond watermarks, PRCs have been used to construct error-correcting codes against computationally bounded channels (Lu et al., CRYPTO 2026), and techniques underlying them have led to new public-key encryption schemes (Döttling et al., CRYPTO 2026, Dyseryn et al. ASIACRYPT 2026). This survey aims to equip the interested researcher with tools to make progress in the area. We detail the current landscape of definitions, techniques, and applications. A main contribution is introducing new tricks and assembling existing tricks to generically strengthen PRCs (e.g., from substitution-robustness to random-deletion-robustness). We also show natural weakenings of PRCs that can be constructed from OWFs or even unconditionally. We conclude with a number of concrete open questions.

This paper studies the cryptanalytic extraction of convolutional neural networks (CNNs). Existing cryptanalytic extraction attacks on CNNs assume that the network architecture is known, and try to recover model parameters. In this paper, we prove for the first time that the architecture assumption can be removed for CNNs with both max and average pooling. Our core finding is that the spatial geometry of the weight vectors recovered by existing parameter-recovery attacks naturally leaks the architecture. We formalize this geometry and establish its correspondence with the architectural knowledge of a convolutional layer: (1) The sparsity consistency with the convolution receptive field reveals the layer type, the kernel size, and the stride; (2) The numerical consistency with the kernel parameters reveals the padding mode and the output-channel number; (3) The structural consistency with the pooling operation reveals the pooling type, the window size, and the stride. Although the recovered vectors are obtained using different methods in the raw-output and hard-label settings, their spatial geometry remains the same. Therefore, our architecture recovery is feedback-agnostic: combined with a parameter-recovery attack, it yields a complete cryptanalytic extraction framework that recovers both the architecture and the parameters in the black-box setting. Extensive experiments, including both layer-wise and end-to-end ones, on a wide range of CNNs demonstrate that simultaneously recovering the network architecture and the model parameters is practical.

In this work, we propose novel security notions for encryption schemes that simulate an adversary in the black-box model equipped with additional side-channel power. More concretely, the adversary is allowed to probe values of the secret-key sensitive algorithms, i.e. key generation and decryption. We then prove a generalization of the well-known Naor-Yung (NY) transform, generically lifting IND-CPA secure encryption schemes to IND-CCA ones in this new probing context. Moreover, we instantiate the resulting framework from lattices, constructing Rutile, a masking-friendly IND-CPA encryption scheme inspired by Kyber, and then Topaz its IND-CCA secure extension. In our proposal, the masking-unfriendly parts of Kyber, namely the central binomial distributions and the FO-transform, are replaced by masking-friendly counterparts (sum of uniforms and the aforementioned NY-transform).

We study affine and linear equivalence of functions $F,G:\mathbb{F}_q^n\to\mathbb{F}_q^m$. We present practical equivalence algorithms that handle arbitrary functions (any algebraic degree, permutations and non-permutations) and are particularly effective on structured instances with nontrivial self-equivalences, covering most cryptographically relevant cases. We also give a dedicated algorithm for computing generators for the group of self-equivalences $(A_1,A_2)$ satisfying $A_2\circ F = F\circ A_1$, for which no comparably efficient general method was previously available.

In this paper, we study multi-party non-interactive key exchange (NIKE) in the fine-grained setting. More precisely, we propose three multi-party NIKE schemes in three computation models, namely, the bounded parallel-time, bounded time, and bounded storage models. Their security is based on a very mild assumption (e.g., NC1 ⊊ ⊕L/poly) or even without any complexity assumption. This improves the recent work of Afshar, Couteau, Mahmoody, and Sadeghi (EUROCRYPT 2023) that requires idealized assumptions, such as random oracles or generic groups. Additionally, we show that all our constructions satisfy a natural desirable property that we refer to as extendability, and we give generic transformations from extendable multi-party NIKE to multi-party identity-based NIKEs in the fine-grained settings. Furthermore, we generalize the lower bound on users’ storage consumption in the bounded storage model by Dziembowski and Maurer (Eurocrypt 2004) to encompass any multi-party NIKE with extendability. This new lower bound suggests that the users’ storage consumption of our multi-party NIKE in the bounded storage model is optimal.

Many cryptographic protocols rely on affine message authentication tags of the form \({\boldsymbol\sigma}=\Delta{\boldsymbol x}+{\boldsymbol k}\), the batch version of information-theoretic MAC. Generating the tags for a long message vector typically requires communication linear in its length. Recent chosen-input VOLE protocols achieve succinct communication, but do not allow the authenticator’s key $k$ to be fixed in advance. We introduce $\mathit{batch\text{-}MAC}$, a primitive that authenticates a chosen message vector in two messages while allowing the authenticator’s key vector to be generated beforehand from a short seed. For a vector of length \(m\), we construct batch-MACs with \(O(m^{2/3}\lambda)\) communication under the decisional composite residuosity (DCR) assumption and \(\operatorname{poly}(\log m,\lambda)\) communication under the learning with errors assumption. Both constructions require a common reference string (CRS). We find applications in constrained pseudorandom functions, 2-message 2PC, and succinct zero-knowledge arguments.

The Terrapin attack against SSH channel integrity (USENIX Security 2024) used a novel attack vector: attacks on the channel state. Surprisingly, not all AEAD modes of SSH were equally affected by this attack, and it remained an open question if "unaffected" meant "secure". Existing formal models for secure channels are based on stateful encryption. However, these models do not define what the channel state is and how it is used as input to the different AEAD modes. In this paper, we propose a formal model for channel integrity under partially chosen state. Applied to the Terrapin attack, the chosen state is the SSH sequence number. It uses an abstract stateful encryption interface, for which we provide pseudocode descriptions for the eight most prominent AEAD modes used in SSH. By varying the SND oracle, we can model ciphertext-only (CO; the Terrapin attack), known-plaintext (KPA), and chosen-plaintext (CPA) attacks. This allows us to establish concrete bounds on the security of the AEAD modes. We find that all three Encrypt-then-MAC (EtM) modes and ChaCha20-Poly1305 in SSH are insecure in the CO model. AES-GCM is the only cipher secure in all three model variants. Going beyond Terrapin, we show that Encrypt-and-MAC (EaM) with a CBC cipher is secure, even in the KPA model. In particular, we describe a novel BEAST-like chosen-plaintext attack on the channel integrity of EaM-CBC, which separates the KPA and CPA models for this scheme.

A randomized encoding for a promise problem is a randomized reduction whose outcome distribution on input $x$ can be simulated within some distance $d$, called privacy, using only one bit of information about $x$: whether it is a YES or NO instance. The encoding is one-sided if this property holds only for YES instances. Our main contribution is showing that (classical and quantum) one-sided randomized encodings have privacy and correctness amplification: any problem with a randomized encoding with privacy $1-1/poly(n)$ and error $1/2-1/poly(n)$, also has an encoding with negligible privacy and error. We then use this result to obtain the following statements: - We show that NISZK, the class of problems that have a non-interactive (statistical) zero-knowledge proof, has strong zero-knowledge amplification, i.e., every problem with a non-interactive zero-knowledge proof with zero-knowledge error $1-1/poly(n)$, also has a non-interactive zero-knowledge proof with negligible zero-knowledge error. This solves a problem open since the work of Goldreich, Sahai, and Vadhan (CRYPTO '99). - The worst-case hardness of a problem that has a perfect (i.e., zero error) one-sided encoding implies one-way functions (OWFs). And if the encoding is quantum, it implies one-way state generators (OWSGs). We then conclude that by removing the error from one-sided randomized encodings, the worst-case hardness of the complexity class SZK becomes sufficient for the existence of OWFs. - Weak and imperfect indistinguishability obfuscation (iO) implies OWFs assuming that the Polynomial Hierarchy does not collapse to its third level. Here, weakness means that the computational distance between the obfuscated circuit and the original circuit is $1-1/poly(n)$, and imperfectness means that the error is $1/2-1/poly(n)$. To achieve this privacy and correctness amplification, we study randomized encodings through the lens of lossy reductions (Ball et al. ITCS 2020). In particular, we introduce an extended and flexible notion of lossy reductions, and show that it is equivalent to randomized encodings. This equivalence serves as the base of our main results and might be of independent interest.

Proton is a popular security and privacy service provider with a large user base spanning both organizations and individuals. Proton Docs/Sheets support real-time collaborative document editing while claiming to provide end-to-end security. We analyze the cryptographic design and the collaborative editing protocol of Proton Docs/Sheets inspecting the open-source Web client and webpage code. We demonstrate three distinct ``integrity'' attacks against Proton Docs/Sheets that can cause history rewriting, context manipulation, and censorship, all of which can, in fact, evade detection. The first two can be launched even when the Proton server acts honestly, and the third is mounted by a corrupted Proton server. We also present the corresponding mitigation methods. Our attacks highlight the subtleties of end-to-end security in collaborative settings involving multiple users and constant updates. This state of affairs naturally calls for systematic formal treatment (i.e., design and/or analysis) of the security of such systems.

At Asiacrypt'25, Lapiha and Prest proposed a lattice-based IND-CCA threshold key-encapsulation mechanism (TKEM) obtained from a threshold identity-based encryption (TIBE) and a signature scheme. Their construction relies on a variant of the Boneh-Canetti-Halevi-Katz (BCHK) transform, instantiated with a lattice-based TIBE. However it suffers from large ciphertexts at 540 KiB for $\kappa = 128$ bits of security. We present substantial improvements to their TIBE, resulting in the first concretely efficient lattice-based IND-CCA TKEM, with ciphertexts just under 30 KiB for a threshold $T = 32$, $Q = 2^{45}$ queries, and the same $\kappa$. Our design simplifies the original framework by leveraging the power of random oracles already present in their construction. We further enhance efficiency by adopting approximate computations where appropriate and by replacing module-NTRU trapdoors with NTRU trapdoors, achieving a remarkable eighteenfold reduction in ciphertext size. Finally, leveraging recent developments in secret sharing, we ensure the verifiability of key-extraction shares even in the presence of malicious parties.

Goldreich's pseudorandom generators (PRGs) expand a secret seed by evaluating a fixed low-locality predicate on random subsets of its bits. We study the concrete and asymptotic security of constructions whose local predicate combines majority with XOR. We introduce Deterministic Pooled Recovery (DPR), which guesses a set of seed positions and pools the outputs whose majority inputs are forced by the guess. The selected outputs yield noiseless sparse linear equations. For the proposed \(\operatorname{MAJ}_7\oplus\operatorname{XOR}_4\) challenge published in ToSC 2025, with seed length \(n=1024\) and output length \(m=n^2\), DPR has an estimated recovery cost of \(2^{109.71}\) operations using \(\omega=2.38\) for the linear-algebra exponent. This is below the claimed 128-bit security level. We also refine the common-bias analysis by evaluating its advantage at each fixed seed weight. This viewpoint motivates Bias-Amplified Pooled Recovery (BAPR), which uses partial majority bias to construct pooled noisy linear systems. For \(m=n^s\) with fixed \(s>1\), odd majority locality \(a\in\omega(1)\cap O(\log n)\), and fixed XOR locality \(b\ge3\), we prove that BAPR admits a worst-case \(2^{o(n)}\)-time implementation that recovers any fixed seed with high probability. This rules out exponential security in this intermediate-locality regime. In addition, we identify exploitable affine structure and residue constraints in proposed alternative predicates.

Nation State Agencies go to great lengths to obtain signal intelligence, including backdooring cryptographic standards and equipment. Although the existence of these backdoor efforts is common knowledge, only few of known backdoored systems have been publicly analysed. In this paper we present the first detailed forensic cryptanalysis of the backdoored UA-8295 message terminal and we try to answer questions how the UA-8295’s backdoor was designed and for which attack. Towards a better understanding of real-world backdoor design, we propose a methodology based on the Backdoor Conjecture that provides handles to reason about the design of backdoors and the attacks they are designed for.

Adaptor signatures enable conditional payments on blockchain networks: a pre-signature is bound to a public instance $Y$ and can be completed into a valid signature by any party knowing a witness $y$ with $\mathcal{R}(Y,y)=1$, while the pre-signature and full signature together allow the signer to extract $y$. Existing constructions only support specific algebraically-structured relations, lose witness privacy, or have significant overhead both in signature size and computation. We present a practical adaptor signature framework for arbitrary NP relations, built from a new primitive we call Online/Offline NIZK whose proof is separated into online and offline parts. The key property is that the witness can be extracted from the online part knowing the randomness used in the offline part. We give two generic constructions: a basic scheme achieving witness hiding, and an instance-hiding scheme achieving a new, stronger forward-secure privacy notion we introduce. We instantiate our framework for the AES-128 relation using the VOLE-in-the-Head paradigm and the FAEST post-quantum signature scheme. The resulting complete signature is $11.6$ kB, roughly $60\%$ smaller than the state-of-the-art adaptor signature for NP (Ciampi et al., CT-RSA '25), and our implementation is an order of magnitude faster in overall signature generation, verification, and extraction.

We explore the use of level structures to generalize the SQIsign signature scheme. We give a general framework where, given the public key and the commitment, the challenge is to exhibit an isogeny between them with an additional requirement, namely to map a chosen level structure to nother. We then instantiate the framework using 1-dimensional and 2-dimensional isogenies. In doing that we provide a new explicit Deuring correspondence for supersingular elliptic curves with level structures and solve new constrained norm equations.

The Round-3 SNOVA signer samples vinegar variables, which are elements of $\mathbb{F}_q$, by reducing uniform byte strings modulo a power of $q$. We show that the resulting bias leaks secret-key information for the six odd-characteristic alternative parameter sets. Even though the leakage is small (the most leaky variables leak at most $0.086$ bits of information), this still leads to efficient key-recovery attacks which we demonstrate in practice. Key recovery becomes a $q$-ary LPN problem of dimension $o\ell$ with a known, full-support error distribution. A naive algorithm needs some ten thousand signatures and $2^{90}$ to $2^{130}$ operations, which is already far below the $170$ to $331$ bits claimed for the six affected sets. Using more signatures makes the attack practical: using standard LPN machinery - BKW reduction with Fourier hypothesis testing - we recover the secret key for four parameter sets in practice using between $9$ and $180$ million signatures, and at most 16 minutes of wall clock time. Fixing SNOVA to sample the vinegar variables uniformly would prevent the attack completely.

In this paper we propose a constant time lattice reduction algorithm for integral dimension-4 lattices. Motivated by its application in the SQIsign post-quantum signature scheme, we provide for the first time a constant time LLL-like algorithm with guarantees on the length of the shortest output vector. We implemented our algorithm and ensured through various tools that it indeed operates in constant time. Our experiments suggest that in practice our implementation outputs a Minkowski reduced basis and thus can replace a non constant time lattice reduction subroutine in SQIsign.

Many isogeny-based schemes rely on the Deuring correspondance and thus require computations with quaternion ideals. In this paper, we generalize the recent results of Leroux on the inert representation of quaternion ideals, yielding a normalized representation together with a set of simple and efficient algorithms for quaternion ideals in the context of the Deuring correspondence when the prime characteristic $p$ is equal to $3 \bmod 4$. One of the main benefit of our new algorithms is that the size of the integers involved in the computations is both easy to bound, and smaller than previous work. In particular, when applied to the latest version of SQIsign, our algorithms yield a concrete integer bound of $O(p^4)$ in the worst case, proven under an experimentally verified assumption on the size of the integers required to perform the lattice reduction part of the response sampling. This bound improves upon all known previous bound by a factor at least $O(p^2)$. Our new algorithms are used in the latest update of SQIsign's implementation submitted to the third round of the NIST call for additional post-quantum signatures. In particular, their low integer bounds facilitated the removal of a dependency on the GMP library in favor of a custom library of fixed-size integers without performance overhead. We stress that our algorithms are not limited to SQIsign, but apply to any isogeny-based cryptographic protocol that relies on the Deuring correspondence such as the PRISM signature scheme for instance.

The field of Fully Homomorphic Encryption (FHE) has seen many theoretical and computational advances in recent years, bringing the technology closer to practicality than ever before. For this reason, practitioners in related fields, such as machine learning, are increasingly interested in using FHE to provide privacy to their applications. Despite this progress, selecting secure and efficient parameters for FHE remains a complex and challenging task due to the intricate interdependencies between parameters. In this work, we address this issue by providing a rigorous theoretical foundation for parameter selection for any LWE-based schemes, with a specific focus on FHE. Our approach starts with an in-depth analysis of lattice attacks on the LWE problem, deriving precise expressions for the most effective ones. Building on this, we introduce closed-form formulas that establish the relations among the LWE parameters. In addition, we introduce a numerical method to enable the accurate selection of any configurable parameter to meet a desired security level. Finally, we use our results to build a practical and efficient tool for researchers and practitioners deploying FHE and other LWE-based schemes in real-world applications, ensuring that our approach is both rigorous and efficient.

Folding protocols cross public-randomness boundaries: input-dependent information must be fixed before a challenge is known, while the required output may depend on it. We ask how much state must be retained after accounting for information already available at the boundary. With unrestricted prechallenge preprocessing and linear postchallenge readout, a retained representation is feasible exactly when its coordinates span, modulo this available information, every output direction induced by the supported challenges. Thus the minimum retained width is the dimension of this residual output space. The width depends on both the admissible source domain and the actual challenge support. Coordinatewise multiplication of length-\(n^2\) vectors has width \(n^2\) on the full product domain but \(2n-1\) when both inputs lie in the dimension-\(n\) Reed–Solomon code evaluated at \(n^2\) distinct points. Below the exact-width threshold, for independent uniform inputs to a bilinear map, we characterize the optimal exact-reconstruction probability by a rank-weighted subspace optimization. Bilinear side information couples the missing directions through a common linear correction, so optimizing them independently can fail. Applying the framework to direct homogeneous quadratic folding on a full product domain, an algebraic separation condition on the challenge family forces width at least \(rk\), where \(r\) is the dimension of the span of mixed bilinear outputs. Scaled-Cauchy challenges attain this bound on any support of at least \(2k+1\) distinct non-pole values and yield an explicit matching \(rk\)-coordinate polynomial representation.

Middle-Product LWE (MP-LWE) can inherit hardness from Ring-LWE over many number fields. Peikert and Pepin developed a direct reduction using field traces. Njah Nchiwo and Pellet-Mary (NP26) proved polynomial loss for the route through Polynomial-LWE under an additional condition on the defining polynomial. Recently, Pellet-Mary and Xia (PX26) unified both routes and proved polynomial loss for defining polynomials with polynomially bounded coefficients under suitable modulus conditions. We analyze Peikert and Pepin's trace reduction for monic irreducible integer polynomials $f$ of degree $n\ge2$ with coefficients of absolute value at most a fixed $B\ge1$. First, when the prime modulus satisfies $q\nmid[\mathcal O_{\mathbb{Q}(\theta)}:\mathbb{Z}[\theta]]$, where $\theta$ is a root of $f$, we prove the existence of multipliers with primal loss $O_B(n\sqrt{\log n/\log\log n})$ and dual loss $O_B(n^2)$. The multipliers have polynomial-size descriptions and serve as advice. Second, whenever the polynomial class is nonempty, every $Q\ge21$ admits a prime $q\in[Q,2Q]$ for which the index condition holds for at least a $1-O_B(n\log n/Q)$ fraction of the polynomials. At the same prime, we prove the same lower bound when counting the distinct number fields defined by these polynomials. Both coverage bounds tend to one as $n\to\infty$ if $Q/(n\log n)\to\infty$. Third, we prove that the trace reduction can match the linear noise bounds achieved by NP26's reduction through Polynomial-LWE. For the same defining polynomial, Ring-LWE variant, and MP-LWE parameters, a suitable trace multiplier exists whose linear noise amplification is no larger than that achieved by NP26's multipliers.

Recent works have studied Byzantine agreement (BA) in the mixed-fault model with resilience threshold $n>2t+r+s$, where $n$ is the number of parties, $t$ is the number of Byzantine parties, $r$ is the number of receive-faulty parties, and $s$ is the number of send-faulty parties. In particular, Feng et al.~(ASIACRYPT~2025) proposed a BA protocol with optimal resilience, expected asymptotically quadratic communication complexity, and expected constant round complexity. However, existing BA protocols guarantee a meaningful output $v$ only when all parties hold the same input $v$. Such a guarantee is insufficient for efficient construction of atomic broadcast protocols, leaving efficient atomic broadcast in the mixed-fault setting as an open problem. In this paper, we present the first atomic broadcast protocol (UABC) with optimal resilience in the mixed-fault model. In addition, our protocol achieves communication complexity $\mathcal{O}(n\ell+n^2\lambda^2)$ when $t=\Theta(n)$, where $\ell$ is the input size and $\lambda$ is the security parameter. This complexity is asymptotically optimal for moderately large $\ell$, namely when $\ell\ge n\lambda^2$. At the core of our design is an undead verifiable information dispersal (UVID) protocol, which allows an input to be efficiently dispersed and later reconstructed in the mixed-fault setting. Leveraging UVID together with the undead multi-valued consensus (ASIACRYPT~2025), we further construct a new primitive called weak undead multi-valued validated Byzantine agreement (wUMVBA) and its optimized variant (Opt-wUMVBA). Unlike standard multi-valued validated Byzantine agreement, wUMVBA and Opt-wUMVBA guarantee that the output is either a meaningful value satisfying the validity predicate or $\bot$. Moreover, the probability of outputting $\bot$ is at most $(t+s)/n$. Finally, by leveraging Opt-wUMVBA, we obtain our optimal atomic broadcast protocol for the mixed-fault setting.

The rapid expansion of quantum computing threatens the security foundations of classical public-key cryptography, including RSA and ECC, both vulnerable to Shor's algorithm. For the automotive industry, where secure communication underpins critical functions such as over-the-air (OTA) updates, secure boot, firmware signing, PKI validation, and vehicle-to-everything (V2X) connectivity, this threat demands early and informed migration planning. This paper presents a cross-platform performance evaluation of NIST-selected and candidate Post-Quantum Cryptography (PQC) algorithms integrated into the Transport Layer Security (TLS) protocol, in both standalone and hybrid (classical + PQC) configurations. Signature schemes CROSS, MAYO, SPHINCS+ (SHA and SHAKE variants), ML-DSA, and FALCON are evaluated alongside key encapsulation mechanisms including FrodoKEM, BIKE, and ML-KEM. Testing spans two representative platforms: a resource-constrained Raspberry Pi 4B (Broadcom BCM2711, 4GB RAM) representing embedded automotive ECUs, and an Ubuntu-based Intel Core i5 system representing backend or gateway infrastructure. Execution time, memory consumption, and CPU core utilization are measured across algorithm-KEM combinations to characterize real-world feasibility. Results indicate lattice-based schemes offer a strong balance of performance, interoperability, and long-term assurance suited to V2X and PKI-heavy communication, while compact signature schemes such as MAYO and FALCON better serve resource-constrained operations like secure boot and firmware signing, where verification speed and low memory footprint are prioritized over frequent key exchange. The study further identifies a critical dependency of PQC feasibility on 64-bit architecture, highlighting the limitations of legacy 32-bit automotive chipsets that lack dedicated cryptographic accelerators. These findings offer practical, use-case-specific guidance for automotive OEMs and suppliers navigating the transition from classical to quantum-resistant cryptography, supporting risk-informed decisions across onboard and offboard vehicle communication architectures.

Updatable Encryption (UE) allows ciphertexts to be updated under new keys without decryption, enabling efficient key rotation. Constructing post-quantum UE with strong security guarantees is challenging: the only known CCA-secure scheme, COM-UE, uses bitwise encryption, resulting in large ciphertexts and high computational costs. We introduce $\mathsf{DINE}_\ell$, a family of compact and efficient CCA-secure isogeny-based post-quantum UE schemes parametrized by an integer $\ell$, achieving increasing security levels for $\ell = 3,4,8$. Each encryption, decryption, or update requires only a few power-of-2 isogeny computations in dimension 2. At NIST security level 1, $\mathsf{DINE}_\ell$ supports 28B messages, 480B ciphertexts, and $\ell\times336$B update tokens, significantly smaller than prior constructions. Our C implementation demonstrates practical performance, with updates in $\ell \times 15$ms, encryptions in 65ms, and decryptions in 155ms. Our design combines high-dimensional isogeny representations with the Deuring correspondence. We also introduce new algorithms for the Deuring correspondence which may be of independent interest. Moreover, the security of our scheme relies on new problems that might open interesting perspectives in isogeny-based cryptography.

We construct the first non-trivial weighted threshold encryption scheme with silent setup. The CRS and each public key consists of O(W) group elements, where W is the maximum committee weight. Both the weight assignments and the threshold for decryption can be dynamically chosen at encryption time. Each partial decryption is a single G1 element computed non-interactively with one scalar multiplication, independent of the party's weight. The CPA ciphertext is two G_1 elements, improving upon the shortest known unweighted scheme, and CCA security costs two additional field elements. We prove security in the generic bilinear group model.

Range proofs are a fundamental building block of lattice-based proof systems. Existing approaches relying on standard Johnson-Lindenstrauss (JL) struggle to provide succinctness: unstructured JL incurs linear verifier complexity, while structured JL introduced in RoK and Roll [ASIACRYPT'25] produce large projection vectors that need to be sent in costly committed form. For a witness of dimension $m$ and a security parameter $\lambda$, the JL projection vector is $O(\lambda),$ but the projection matrix is $O(\lambda m)$. Structured JL can be represented in $\widetilde O(\rho \lambda^2)$, but the resulting projection length grows to $\widetilde O(m / \rho)$ for any trade-off parameter $\rho > 1$. \hspace{5mm} In this paper, we introduce Tensor Train Random Projection (TTRP) to lattice-based cryptography, offering a ``best of both world" structured JL variant. The Tensor Train (TT) matrix representation drops to $\widetilde O(\lambda)$, while maintaining the projection vector at $O(\lambda)$. \hspace{5mm} We show that TTRP can be seamlessly integrated into proof systems by linearizing the well-formedness of the projection with the sum-check protocol. Concretely, using TTRP + sum-check as a drop-in replacement for existing range proof techniques yields an approximate shortness check with $O(\lambda \log m (\log\log m)^2)$ verifier complexity and proof size around $4$ KB for relevant parameter settings.

We give a sharper bound on the error introduced by replacing a random permutation oracle with Carolan’s compressed permutation oracle. Building on Rosmanis’s representation-theoretic approach, we represent exact permutation states in the compressed oracle’s database space, allowing a direct comparison of the two query operations. For a uniform permutation on N points, this comparison bounds the soundness error by 4q/√N after q ≤ N/4 queries. As the main application, we obtain tight query complexity for constant success probability in sponge preimage and collision search by substituting this bound into Carolan’s search reductions. The same analysis improves bounds for the one-more problem and cycle finding, and gives tight query complexity for constant success probability in keyless Davies–Meyer collision search

We formalize and construct $sparse$ arguments of knowledge, where given a length $N$ witness that contains $n$ non-zero values the prover time is $O(n) \circ o(N)$. In particular, we achieve a sparse argument of knowledge for the customizable constraint system relation, which generalizes circuit-satisfiability. We achieve this by first utilizing $sublinear$ lookup arguments to provably select only the subset of the constraint system that touches the non-zero entries of the variable assignment, and then running a standard argument of knowledge with mild qualifications. Our resulting sparse argument of knowledge is publicly verifiable and supports a universal and updatable setup. It features an $O(N \log N)$ preprocessing phase, an $O(n \log n)$ prover time, and an $O(1)$ verifier time, improving on the prior $O(n \log^2 n)$ and $O(n \log N)$ prover time.

Oracle services enable dApps to access real-world data and respond to events happening outside the blockchain ecosystem. A large number of academic and industrial oracles have emerged, capturing various designs, capabilities, and security assumptions/guarantees. This rapid development makes it challenging to comprehend existing solutions, understand their trade-offs, and build on them. In this paper, we introduce a systematization of knowledge for blockchain oracles. We develop a multi-dimensional systematization framework assessing existing solutions based on their capabilities, trust and security assumptions/guarantees, and their design architectures. To further aid in this assessment, we go on to distill a number of insights and gaps, and provide a guide for practitioners (on the use of these oracles) and researchers (by highlighting future research avenues).

We propose an efficient threshold mechanism that is well suited for certain hash-based signatures. Our proposal has a number of compelling features: • The final signature assembled from a quorum of signature shares has the same format as a non-threshold hash-based signature. In particular, the verifier is unaware that the signer is thresholdized. Moreover, we can add and remove parties and change the threshold without changing the public key. • Our framework can be used to thresholdize one-time signature schemes, such as Winternitz, and a few-time schemes, such as XMSS, HORS, FORS, and PORS+FP. • The public key and the signatures reveal nothing about the number of parties $n$ or the threshold $t$. • An instantiation of our framework scales well to a large number of parties $n$ and a large threshold $t$. One limitation of the framework is that it does not support an efficient distributed key generation protocol (DKG). Another limitation is that the framework does not directly apply to SLH-DSA. It is primarily designed for the Winternitz, XMSS, and FORS schemes.

Adaptive reprogramming is a central tool for proving security in the quantum random oracle model (QROM). Recently, it has been extended to the setting when an oracle-dependent advice is in presence. However, the existing result is restricted to classical advice, and addresses only a single reprogramming. Extending it to n reprogrammings through naive hybridization yields a loss that is linear in the number of reprogrammings. In this paper, we establish an adaptive reprogramming theorem that accommodates quantum advice, and has a tighter overall security loss that only scales in the cubic-root of n, with an attack that matches the number of queries required in order to reach constant advantage. Using our reprogramming theorem, we provide improved time-space tradeoff bounds for PRG security of random oracles, and an IND-CPA security proof of salted pseudo one-time pad, in the quantum advice setting. Along the way, we also identify and prove a folklore lemma about indistinguishability of two oracles, which may be of independent interest. As another application, we provide a generic security analysis of Fiat-Shamir and hash-and-sign signature schemes, in QROM with quantum advice. We do so via a two-step modular analysis, where in the first step we perform a generic UF-CMA to UF-NMA reduction using our reprogramming theorem, and in the second step we provide a characterization of when such a scheme satisfies UF-NMA security, which we then use to discuss security of several signature schemes in this model.

Solving multivariate polynomial systems is a fundamental problem in cryptanalysis, with increasing relevance in algebraic attacks on arithmetization-oriented (AO) primitives. Current approaches include Gröbner bases and the Sylvester resultant. However, Gröbner basis methods typically rely on FGLM to change the order, which requires a zero-dimensional ideal, whereas the Sylvester resultant eliminates only one variable at a time, limiting its flexibility in multivariate elimination. We revisit the Dixon resultant as an efficient and flexible tool for eliminating several variables simultaneously. We derive refined upper bounds on the Dixon matrix size via lattice-path counting and analyze the complexity under several determinant computation models, yielding explicit complexity estimates. For well-determined systems, the Dixon resultant is a viable alternative to Gröbner basis methods. Moreover, given their respective limitations, Dixon resultants are particularly suitable for parametric elimination in underdetermined systems, while Gröobner basis methods are generally preferable for solving overdetermined systems. We present an efficient open-source C implementation, DRSolve, with multiple determinant methods and a degree-aware submatrix selection strategy to mitigate the impact of extraneous factors. On the randomly generated well-determined systems tested, our implementation is competitive with the state-of-the-art Gröbner basis solvers Magma and msolve, with DRSolve performing better in the low-variable/highdegree regime, while Magma and msolve remain preferable in the high-variable/low-degree regime. Finally, we formulate three elimination strategies for polynomial systems arising from AO primitives: direct elimination, iterative elimination, and reduction-based hybrid elimination. We demonstrate these strategies on Poseidon, Vision, and XHash12, obtaining complexity reductions in many cases.

The NIST SP 800-208 eXtended Merkle Signature Scheme (XMSS) profile XMSS-SHAKE256_10_256 repeatedly invokes SHAKE256 primitives, making field-programmable gate array (FPGA) accelerator latency sensitive to Keccak permutation time and absorb/finalization-control overhead. Established round unrolling reduces permutation-side latency but leaves residual controller work. We propose SAFE Final-Block Construction (SAFE FBC), combining guarded SHAKE256-rate-aware eligibility, direct eligible final-block construction, and sequential-finalization bypass without modifying Keccak. A controlled 12-configuration study crosses four controller/finalization organizations with one-, two-, and three-round-per-cycle datapaths over nine XMSS-oriented cases: 96-byte F and pseudorandom-function (PRF) inputs, 128-byte H and PRFkeygen inputs, and selected message-hash cases around the SHAKE256 rate boundary. At every fixed depth, SAFE FBC reduces aggregate L2-final latency by 13 cycles versus the Structural Controller Decoupling (SCD) baseline; L2-final spans the first accepted AXI input beat to the final accepted AXI output beat. At three rounds per cycle, aggregate latency falls from 305 to 292 cycles, with no case regressing. F and ordinary PRF show the largest per-call gain, from 30 to 25 cycles (16.67%). Component measurements localize the saving to absorb/finalization control; Keccak permutation latency, squeeze latency, and input-transfer span remain unchanged. SAFE FBC complements unrolling, trading latency for substantial lookup-table/slice growth with very small flip-flop overhead.

Constructing a post-quantum signature scheme that is simultaneously compact and efficient remains a central challenge in code based cryptography. Existing schemes based on turning a zero-knowledge identification scheme into a signature exhibit either large signatures or slow verification procedures, while the design of hash-and-sign schemes has led to constructions such as Wave and MIRANDA, having small signatures at the cost of massive public key sizes and time-consuming signature algorithms. In this work, we present ASTRA-Sign: a quASi cyclic code-based full-distance decoding TRApdoor Signature Scheme, combining the hash-and-sign paradigm with quasi-cyclic moderate density generator matrix codes to obtain small signatures and small public keys. The security of our scheme is based on the hardness of finding low weight codewords in quasi-cyclic codes, and on the hardness of finding a codeword that has full Hamming distance from a given random vector. We analyse key recovery and signature forgery attacks against ASTRA, and we propose several parameter sets achieving 128-, 192- and 256-bit security. Our scheme exhibits public keys and signatures below 1 kiB for 128 bits of security, with 40 µs verification times.

Robust asynchronous threshold Schnorr signatures enable honest parties to complete signing despite malicious behavior and network delays. To reduce online latency, these schemes typically follow an offline–online paradigm. The parties prepare batches of shared nonces in the offline phase and consume them to generate signatures online. In high-throughput constructions such as SPRINT (Benhamouda et al.~Eurocrypt 2024), super-invertible matrices are used to extract multiple nonces and their corresponding commitments from a single execution of the nonce-generation protocol. However, the associated group operations incur substantial computational overhead. We identify several computational bottlenecks in SPRINT and present $\mathsf{DASH}$, a suite of robust asynchronous threshold Schnorr signature protocols that substantially improve computational efficiency while retaining SPRINT’s amortized linear communication cost. Our contributions are threefold. First, we propose distributing the group computations involved in nonce extraction among the parties. This reduces the amortized computational cost from $\mathcal{O}(\log n)$ to $\mathcal{O}(1)$ group operations per signature per party. To enable this distributed computation, we employ a batching technique that allows each dealer to distribute $L$ nonce polynomials simultaneously. Second, we introduce an efficient batched share-verification procedure that jointly checks shares of all $L$ polynomials. For concrete parameters, this can accelerate share verification by a factor of $40-60$. We also introduce a simple agreement protocol via player elimination, which could be of independent interest. Finally, under a stronger honest-majority assumption, we develop a packed two-stage online reconstruction protocol that enables parties to reconstruct signatures through error correction. This eliminates the need for signature-share verification, and consequently, avoids the expensive group operations performed by the aggregator in SPRINT. Surprisingly, even in the worst case, our $\mathsf{DASH}$ protocols preserve the same asymptotic amortized computation and communication costs, provided that $L \gg n\log n$. %We prove the security of all $\mathsf{DASH}$ variants in the random oracle model under the standard discrete logarithm assumption.

Fully Homomorphic Encryption (FHE) allows computation on encrypted data without decrypting it. In theory, FHE makes privacy-preserving machine learning possible. In practice, however, it remains impractically slow for real workloads. A major source of slowdown is bootstrap operations; in CKKS, a popular FHE scheme for tensor workloads, the slowdown is compounded by scale management and rescale operations. FHE compilers for machine learning inference aim to make bootstrap placement and scale management efficient and easy by compiling high-level tensor programs into optimized CKKS computations. Unfortunately, existing approaches miss crucial optimization opportunities because they overlook a key property of CKKS programs: bootstrap and rescale placement are fundamentally coupled through the level budget. In this paper, we present Orbit, an FHE compiler that jointly optimizes bootstrap and rescale placement through a novel Integer Linear Programming (ILP) formulation that reasons about both ciphertext level and scale constraints. To make this formulation tractable for structured tensor workloads, particularly convolutional neural networks, we introduce three techniques that reduce ILP complexity while preserving optimality. Across five workloads and multiple cryptographic parameter configurations, Orbit achieves a geometric mean speedup of $19\%$ over DaCapo, $73\%$ over Orion, and $52\%$ over ReSBM, keeps compilation under 6 minutes, and retains model accuracy within $0.3\%$ of plaintext execution.

Threshold ring signatures (TRS) enable a quorum of $T$ users to jointly sign a message while hiding which $T$ of the $N$ ring members participated, supporting privacy-preserving endorsement in ad-hoc settings. That said, many deployments do not need anonymity over every $T$-subset of a ring: when the approval pattern is already public, a structured ring can be sufficient. In this work, we first formalize this setting as a structured threshold ring signature (sTRS) and introduce $\mathsf{LoTRS}$, a lattice-based sTRS that avoids a dedicated leader and keeps interaction to the optimal number of two rounds by separating the threshold signing relation from the anonymity mechanism. To the best of our knowledge, $\mathsf{LoTRS}$ is the first construction in which a TRS variant is obtained by combining: (i) an aggregated signing layer: a two-round lattice-based multisignature protocol producing an aggregated signature relation, with (ii) a selection-hiding layer: a $1$-out-of-$N$ proof that hides the chosen ring element supporting that relation. While it is natural to use a $T$-out-of-$N$ proof to build a TRS, our $\mathsf{LoTRS}$ exploits a $1$-out-of-$N$ proof to significantly improve efficiency. $\mathsf{LoTRS}$ concretely instantiates the aggregated signing layer using $\mathsf{DualMS}$ (Crypto'23) and the selection-hiding layer arising from Esgin et al.'s lattice-based one-out-of-many proof (IEEE S&P'22). Our $(T, N\!\cdot\!T)$-$\mathsf{LoTRS}$ construction achieves $\mathsf{polylog}(N, T)$ signature size and outperforms $(T, N)$-TRS schemes significantly. For example, for $N=100$ and $T=50$, our analytic signature size is $52.66$ KiB, which is $\approx2.3 \times$ smaller than the previously best performing lattice-based scheme $\mathsf{LastRings}$ by Jeon et al (ISC'25). Our Rust reference implementation further supports practicality: for $T=16$ and $N=32$, i.e., structured ring size $T\cdot N = 512$, it produces $42.72$ KiB signatures, with mean signing time $151.4$ ms and verification time $54.56$ ms in a release build on a Ryzen AI 9 HX 370 laptop.

We present a simple leveled fully homomorphic encryption (FHE) scheme that achieves IND-CCA1 security in the standard model without relying on SNARKs. Our construction combines the gadget trapdoors introduced by Micciancio and Peikert [Eurocrypt 2012] with the GSW leveled FHE scheme [Gentry, Sahai, Waters, Crypto 2013]. We make only minimal changes to GSW to obtain security: ciphertext size remains similar and the homomorphic operations are unchanged. We prove IND-CCA1 security under a new delayed-target LWE assumption. We provide evidence for its hardness in two ways: under a lattice knowledge assumption, we give a reduction from standard LWE; for a restricted parameter regime, we give a quantum reduction from the one-more inhomogeneous short integer solution assumption. Finally, using the randomized evaluation technique from Bourse et al. [Crypto 2016], we obtain IND-CPA-D security for branching programs using only polynomial noise flooding and without requiring statistical correctness.

Key generation in ML-KEM (CRYSTALS-Kyber) samples a short secret from a centered binomial distribution (CBD) and immediately transforms it with the number-theoretic transform (NTT). Each execution draws fresh randomness, so an attacker obtains a single trace and cannot average. We show that one power trace of the optimized pqm4 implementation on an Arm Cortex-M4 suffices, and that the two operations leak complementary information. The CBD sampler stores each coefficient as a signed $16$-bit two's-complement word, which reveals its sign almost without error but too little about its magnitude to yield long error-free hint sets. The missing magnitude leaks in the subsequent NTT, where coefficients are multiplied by public twiddle factors. Used as hints in a primal lattice attack, the combined leakage reduces the estimated BKZ block size for ML-KEM-768 from $624$ to $129$, corresponding to $2^{182.2}$ and $2^{37.7}$ in the core-SVP model. A control experiment shows that the same NTT code leaks the Hamming weight of a near-uniform operand as strongly as that of the secret, yet reveals almost nothing about its value. What makes a coefficient vulnerable is how few values it can take, not how small it is. That points to a countermeasure inside the CBD sampler. Each sample $v$ is stored as a random representative $\widetilde v = v + rq$ of its residue class, with $r$ nonzero, so the NTT and all later code run unchanged. This $q$-randomization drives the sign leakage to the noise floor, reduces the NTT magnitude leakage by $4.5\times$, and raises the block size from $129$ back to $611$, about four bits of estimated security lost instead of more than $140$. The sampler grows by $937$ bytes of firmware.

Evaluation keys represent a primary memory and initialization bottleneck in matrix-native fully homomorphic encryption (FHE). In the Gentry–Lee (GL) framework, each Trace product yields a four component ciphertext whose BigSwitch procedure requires two extended-ring keys, dominated by a massive product-secret key (sXsY → sX). We present Trace-Factored BigSwitch (TFB), which structurally eliminates this product-secret evaluation key by exploiting the rank one tensor structure of the Trace-generated key, (1, sX, sY , sXsY ) = (1, sY ) ⊗ (1, sX). By routing the product secret through an sY → sX switch followed by standard base ring (s^2)X → sX relinearization, TFB achieves an exact evaluation-key memory saving of (n − 1)/(2n) (≈ 50%) without ever generating or storing evk_XY →X. To amortize TFB’s recurring base-ring relinearization cost in blocked matrix multiplication, we introduce FRee-L (Fused Relinearization with noise reduction). FRee-L accumulates K Trace products componentwise in the four-component domain and invokes BigSwitch only once per output tile, reducing post-processing noise injections from K to one and amortizing the latency overhead as O(1/K). In our OpenFHE-linked prototype (n = 256), TFB reduces the coefficient-domain BigSwitch key footprint from 768.0 to 385.5 MiB (49.80%), peak RSS by 15.31%, and cold key preparation by 51–63%. On real GPT-2 attention kernels (K = 4), FRee-L lowers TFB’s paired overhead to 5.90%, which further diminishes to 0.1933% (K = 160) and 0.0618% (K = 512) in deep accumulation workloads, confirming that substantial key savings are achieved with negligible workload-level compute penalty.

The Learning With Errors (LWE) problem provides a conservative and well-established security foundation for lattice-based cryptography, while Learning With Rounding (LWR) improves bandwidth efficiency through deterministic rounding. However, the rounding noise in LWR is inherently correlated with the hidden linear term, preventing tight and sample-preserving reductions to standard LWE for polynomial moduli. Prior work on the Learning With Quantization (LWQ) problem established a tight security reduction from LWE, where standard additive noise is intrinsically replaced by quantization error. We specialize this framework by explicitly instantiating the quantization lattice as a scaled integer lattice. Under this geometric choice, specifically utilizing aligned power-of-two moduli, we derive the exact finite-support error distribution and construct an explicit bijection between transmitted split samples and normal-form LWE samples. These ingredients yield \(\mathsf{MAMBA\text{-}Frost}\), a plain LWQ-based key encapsulation mechanism that seamlessly pairs unstructured LWE security with quantization-based compression. Because its effective error stems entirely from public dithered quantization, \(\mathsf{MAMBA\text{-}Frost}\) operates strictly with hardware-friendly power-of-two arithmetic. Furthermore, the construction employs an \(E_8\)-coded message embedding to optimize the correctness margin without altering the underlying LWQ hardness argument. We instantiate \(\mathsf{Frost}\) at NIST security levels 1, 3, and 5 and provide a complete software implementation. Its power-of-two design realizes the core arithmetic with shifts and masks, keeping the implementation simple and amenable to constant-time engineering. At level 1, \(\mathsf{Frost\text{-}128}\) reduces the combined public-key and ciphertext size by about \(40\%\) compared with \(\mathsf{FrodoKEM\text{-}640}\) and by about \(8\%\) compared with \(\mathsf{SCloud}^{+}\text{-}128\). On an AVX2 platform, \(\mathsf{Frost\text{-}128}\) achieves speedups of \(2.3\times\), \(1.8\times\), and \(1.5\times\) for key generation, encapsulation, and decapsulation over \(\mathsf{FrodoKEM\text{-}640}\), and \(2.1\times\), \(1.5\times\), and \(1.3\times\) over \(\mathsf{SCloud}^{+}\text{-}128\).

Secure/Multipurpose Internet Mail Extensions (S/MIME) is a standards-based mechanism for certificate-backed email signing and encryption. Its post-quantum migration is now technically actionable: ML-KEM public keys can be represented in X.509 certificates, and CMS can carry ML-KEM recipient information through \texttt{KEMRecipientInfo}. These standards solve an encoding problem, but they do not by themselves solve an assurance problem. A mailbox may possess a post-quantum-capable certificate while a generated CMS \texttt{EnvelopedData} object still protects the content-encryption key (CEK) through RSA or elliptic-curve key management; a multi-recipient message may mix ML-KEM and classical recipient paths for the same CEK; and archived mail may remain protected only by quantum-vulnerable mechanisms.This paper studies this assurance gap by shifting the unit of analysis from certificates to encrypted messages. We model encrypted S/MIME as a multi-recipient CMS object with certificate-bound paths to a shared CEK and show that post-quantum confidentiality is a universal message-level property: every valid path to the CEK must satisfy the active migration policy. Consequently, the presence of one ML-KEM recipient path is insufficient if another valid classical path can recover the same CEK. We then present \texttt{SMIME-PQCheck}, a standards-driven validation framework that combines X.509 profile checks, CMS recipient-structure analysis, policy-driven hybrid handling, and archive-level risk aggregation. The framework classifies S/MIME objects as \texttt{pqc-protected}, \texttt{hybrid-protected}, \texttt{classical-only}, \texttt{unsafe-mixed-mode}, \texttt{invalid}, or \texttt{unknown}. The result is a practical assurance layer for organizations migrating S/MIME deployments from certificate-level PQC readiness to auditable message-level protection.

Folding is a core technique in building efficient code-based polynomial commitment schemes with polylogarithmic proof size and verification. To date, there are two families of linear error-correcting codes that have been known to allow folding, i.e., Reed-Solomon (RS) codes and foldable codes. However, neither admits optimal proximity gaps, which constitutes a fundamental bottleneck in the resulting proof sizes. In this work, we give the first folding scheme that enjoys optimal proximity gaps. Specifically, we construct IOPPs (interactive oracle proofs of proximity) for both folded RS (FRS) codes and univariate multiplicity (UM) codes, which achieve optimal proximity gaps $1-R-\varepsilon$ as shown in a recent work of Goyal and Guruswami (STOC 2026), where $R$ is the code rate and $0<\varepsilon<1-R$. Our key observation is that the typical even/odd folding commutes with polynomial reduction for suitably paired moduli, enabling consistency checks via the Chinese remainder theorem. This implies a new generic folding framework that recovers the FRI folding for RS codes and enables efficient folding for FRS and UM codes. We further refine the optimal proximity gap analysis for FRS and UM codes, yielding tighter concrete soundness bounds. Our IOPPs have $O(N)$ prover time and oracle proof size, together with $O(\lambda\log{N})$ query complexity and verification time, where $N$ is the block length and $\lambda$ is the security parameter. These bounds match the asymptotic complexity of FRI while attaining optimal proximity gaps and allowing smaller underlying fields at the same security level. E.g., a $128$-bit field suffices for $100$-bit security.

Incremental preprocessing in private information retrieval (PIR) schemes refers to handle insertions, modifications, and deletions to the database without requiring complete preprocessing after each update. This broadens the applicability of PIR in practical scenarios. However, two major issues remain: the concept of incremental preprocessing for the single-server PIR is still not established, and the row-level update strategy (iSimplePIR (Row-level)) introduces excessive unnecessary overhead. This paper aims to efficiently extend incremental preprocessing to the single-server setting. To our knowledge, we are the first to propose the formal definition of single-server incremental PIR. Besides, we construct an entry-level incremental scheme (iSimplePIR (Entry-level)) based on SimplePIR (USENIX ’23). iSimplePIR (Entry-level) supports real-time updates of individual entries, as well as optimization of communication for scenarios with certain update cycles by incorporating a row aggregation mechanism. For a 1\% column-major update in a 1GB database, iSimplePIR (Entry-level) achieves a 224$\times$ reduction in preprocessing computation overhead and a 4.2$\times$ reduction in both communication and monetary costs compared to iSimplePIR (Row-level). When applied to password breach detection with completely random entry updates, iSimplePIR (Entry-level) reduces preprocessing time by 86$\times$. Meanwhile, our method can be combined with various SimplePIR-based schemes to reduce preprocessing costs, such as DoublePIR, Authenticated PIR (based on the LWE assumption), VeriSimplePIR, and YPIR.

Secure merge considers the problem of combining two sorted lists into a single sorted secret-shared list. Merge is a fundamental building block for many real-world applications. For example, secure merge can implement a large number of SQL-like database joins, which are essential for almost any data processing task such as privacy-preserving fraud detection, ad conversion rates, data deduplication, and many more. We present two constructions with a communication bandwidth and rounds tradeoff. Logstar, our bandwidth-optimized construction, takes inspiration from Falk and Ostrovsky (ITC, 2021) and runs in $O(n\log^*n)$ time and communication with $O(\log n)$ rounds. In particular, for all conceivable $n$, the $\log^*n$ factor will be equal to the constant $2$, and therefore we achieve a near-linear running time. Median, our rounds-optimized construction, builds on the classic parallel medians-based insecure merge approach of Valiant (SIAM J. Comput., 1975), later explored in the secure setting by Blunk et al. (ITC, 2025), and requires $O(n \log^c n)$, $c \approx 1.71$, communication with $O(\log \log n)$ rounds. We introduce two additional constructions that merge input lists of different sizes. SquareRootMerge merges lists of sizes $n^{\frac{1}{2}}$ and $n$ and runs in $O(n)$ time and communication with $O(\log n)$ rounds. CubeRootMerge is closely inspired by Blunk et al.'s (ITC, 2025) construction and merges lists of sizes $n^{\frac{1}{3}}$ and $n$. It runs in $O(n)$ time and communication with $O(1)$ rounds. We optimize our constructions for concrete efficiency. Despite extensive research, efficient secure merge still relies on Batcher's merging network or generic sorting, with $O(n\log n)$ circuit size and $O(\log n)$ depth. Ours are the first constructions to lower their concrete costs through better asymptotics and small constants. Two-party implementations of all four protocols show that, for $n=2^{20}$ 32-bit elements per list, Logstar reduces online communication by $2.09\times$ versus Batcher and $3.69\times$ versus shuffled quicksort. Median reduces Batcher's online rounds by $19.0\%$, using $2.03\times$ its online communication. For unequal inputs, SquareRootMerge and CubeRootMerge reduce Batcher's online communication by $3.34\times$ and $4.69\times$, respectively.

Multiparty Private Set Union (MPSU) allows multiple parties to compute the union of their private sets while revealing no additional information beyond the prescribed output. Existing practical MPSU protocols typically manipulate individual elements and rely on a distributed shuffle to hide element ownership, introducing an $O(n)$ round-complexity bottleneck. We present United, a shuffle-free MPSU framework based on a new data-structure abstraction, Oblivious Insert-Only Invertible Sketches (OIIS). Rather than anonymizing an element-wise intermediate representation, the parties directly construct a protected symmetric sketch of the union. Conditional oblivious insertion ensures that each distinct union element is inserted exactly once while hiding the insertion condition and modified locations. We provide two realizations of OIIS. Our HE-based construction instantiates OIIS with a standard Invertible Bloom Lookup Table protected by multi-key partially homomorphic encryption, while our MPC-based construction uses additive secret sharing and introduces a new row-local power-sum OIIS technique for efficient conditional insertion. We implement and evaluate both constructions. Across the evaluated configurations, our MPC-based construction achieves up to $3.10\times$ lower runtime in LAN settings, while our HE-based construction achieves up to $2.73\times$ lower runtime in WAN settings compared with prior MPSU protocols.

The security of RSA is generally understood to be based on the complexity of factoring, and key size parameters are extrapolated from the general number field sieve (GNFS). However, this may not accurately represent RSA security in practical scenarios. An under-appreciated 2007 algorithm of Joux, Naccache, and Thomé allows an attacker to forge RSA signatures after temporary access to a raw RSA signing/decryption oracle in time close to the *special* number field sieve (SNFS) without factoring the key. We implement and run this algorithm for 1024-bit RSA. In total, the attack took 1380 CPU core-years over five calendar months, and made $2^{32}$ oracle queries. Most of this time is precomputation; after the precomputation the attacker can forge any signature of choice, offline, in 180 core-years. We carried out our attack using a hardware security module (HSM) as the signing oracle, thus demonstrating the ability to impersonate the HSM through black-box API interactions, without exfiltrating the key. Blind RSA schemes also provide such a signing oracle. Extrapolating our empirical running times to larger key sizes, we conclude that the concrete security of RSA with a signing oracle should be 15 to 30 bits lower than the factoring-based security estimates for the 1024-bit to 4096-bit RSA parameters that are common in practice. Even 4096-bit RSA does not appear to meet a 128-bit security level in this attack model. This highlights a gap in current RSA-type security assumptions, and gives classical cryptanalytic evidence in favor of moving away from RSA entirely during the current post-quantum transition.

Covert communication can be achieved without anamorphic encryption.

We analyze a recently published construction of redactable blockchains based on multivariate polynomial equations. We show that its public integrity conditions reveal sufficient algebraic structure to recover the relevant private evaluation points efficiently. As a consequence, the blockchain unauthorized redaction problem reduces to univariate root finding over a finite field followed by solving a linear system. We also discuss an alternative attack based on linearization.

Modern cryptocurrency wallets use Schnorr signatures together with public, deterministic key derivation: by repeatedly rerandomizing a single master public key, an unlimited number of public keys for incoming payments can be derived without access to secret data. Moreover, some wallets support multi-signature schemes such as MuSig2 for optionally aggregating derived public keys from $n$ distinct parties into "$n$-of-$n$" public keys, so that funds received on such aggregated keys can be spent only when authorized by all $n$ parties. However, these advanced key-management techniques stretch the underlying cryptographic schemes beyond the guarantees provided by existing security proofs. Although deterministic Schnorr wallets and MuSig2 have each been proven secure in isolation, the provable security of their composition---despite its deployment in wallets---has not been established thus far. The goal of this work is to narrow this gap between theory and practice. We provide the first formal security analysis of the joint use of single-signer Schnorr signatures and MuSig2, both under rerandomization and in deterministic wallets. First, we introduce the notion of a (rerandomizable) joint-signature scheme, in which an honest signer uses the same secret key for single-signer signatures and multi-signatures. Within this model, we prove tight security of Schnorr signatures plus MuSig2 under the algebraic one-more discrete logarithm (AOMDL) assumption in the combination of the algebraic group model (AGM) and the random oracle model (ROM). Finally, we provide a formal model of deterministic wallets with joint-signature schemes and show that Schnorr plus MuSig2 remains secure under public, deterministic key derivation.

We present the first polynomial-time attacks against certain parameter choices of the low-dimensional vector representation (LDVR) problem, introduced by Crites, Katz, Komlo, Tessaro, and Zhu (CRYPTO '25) to prove the adaptive security of FROST in the Algebraic Group Model. Building on work by Crites and Stewart (CRYPTO '25), they also show that every attack against LDVR directly yields an adaptive attack—i.e., one leveraging adaptive corruptions—against any threshold Schnorr signature that exposes to the attacker the values $g^{\mathrm{sk}_i}$ for the secret-key shares $\mathrm{sk}_i$ held by the signers, where $g$ is a generator of the underlying group. This class of protocols includes FROST (Komlo and Goldberg, SAC '20), Sparkle (Crites, Komlo, and Maller, CRYPTO '23), Lindell's three-round protocol (CiC '24), along with other masking-free protocols. Consequently, our results give the first polynomial-time adaptive attacks against a class of practical threshold signatures. Our attacks require a large, but still polynomial, number of signers: asymptotically, we require $n = \Theta(\log^2 p)$ signers and corrupt $t_c = \Theta(\log p)$ of them. For the prime-order subgroup of Curve25519, concrete and optimized instantiations of our attack require $n=5020$ signers and corrupt $204$ of them. This exceeds the signer counts commonly encountered in practice, but falls within the scenarios considered by NIST's call for threshold signatures. At the theoretical level, our results also yield a non-artificial separation between static and adaptive security based on a real-world multi-party protocol. Our results arise from a new reduction from LDVR to structured subset-sum problems determined by the evaluation points of the polynomial used to secret-share the signing key. We further show that allowing these evaluation points to be chosen adversarially can lead to substantially more efficient attacks.

Over the last fifteen years, side-channel analysis (SCA) against implementations of both classical and post quantum cryptography has undergone a quiet but fundamental change of paradigm: from treating leakage as evidence to be combined with an algorithm’s structure into a system of exact equations to be solved, to treating it as evidence to be combined into a joint probability distribution to be queried. This survey gives a systematic account of that evolution through four paradigms: Algebraic SCA (ASCA), which encodes leakage as hard Boolean or algebraic constraints and solves the resulting system with SAT or Gröbner-basis solvers; Tolerant/weighted Algebraic SCA (TASCA), which relaxes hard constraints into pseudo-Boolean costs and recovers a maximum-a-posteriori key via combinatorial optimization; Soft Analytical SCA (SASCA), which retains leakage as full likelihood functions attached to a factor graph and approximates the posterior with loopy belief propagation; and Exact SASCA (ExSASCA), which replaces the uncertified approximation of loopy belief propagation with exact inference on a knowledge-compiled tractable probabilistic circuit. Our central contribution is a single Bayesian factor-graph formulation under which all four paradigms are shown to be different representations of the same leakage-augmented joint distribution combined with different inference queries and solvers: we prove, in a precise but lightweight sense, that ASCA is the zero-temperature limit of TASCA, that TASCA is a maximum-a-posteriori restriction of SASCA’s model, and that ExSASCA is SASCA with its approximate solver replaced by an exact one on the identical graphical model. Building on this formulation, we develop a taxonomy that organizes the literature along a representation axis (hard, weighted, probabilistic) and a solver axis (satisfiability, combinatorial optimization, approximate message passing, exact tractable inference), survey the concrete attack literature in each paradigm on both AES and NTT-based post-quantum schemes (Kyber, Dilithium, Falcon/FN-DSA), and analyze the computational and security trade-offs of each approach, including its implications for what a security evaluator may soundly conclude from a failed attack. We close with open problems at the current frontier of tractable probabilistic inference for cryptographic implementation security.

Falcon offers compact signatures and well-studied mathematical security, but its side-channel security remains a critical challenge. In particular, its floating-point Gaussian sampler constitutes a major source of side-channel leakage, while protecting it efficiently is difficult because of its reliance on floating-point arithmetic. However, existing attacks targeting the sampler under realistic noisy-leakage conditions typically require thousands of traces and are evaluated on unoptimized implementations. In this paper, we present a new attack on Falcon's sampler, which reduces the number of required traces by more than two orders of magnitude compared with prior attacks under realistic noisy-leakage conditions and works on both unoptimized (\texttt{-O0}) and highly optimized (\texttt{-O3}) builds. Our attack targets the floating-point Gaussian centers $\mu$ used by the sampler at the recursion leaves of the \texttt{ffSampling} procedure. Specifically, it exploits signed-exponent leakage arising when $\lfloor\mu\rfloor$ is converted from an integer back to floating-point representation, as well as the weaker sign-only leakage. Signed-exponent leakage places the integer part $\lfloor\mu\rfloor$ of each center in one of $17$ signed power-of-two intervals, whereas sign-only leakage divides it into two intervals according to its sign. We propose CB-MLE, which uses gradient descent to obtain an estimate $f^\star$ whose computed Gaussian-center classifications are consistent with the observed classifications under the profiled confusion matrix. An ISD-style search tests subsets of the more reliable coefficients in $f^\star$ with LLL/BKZ on the corresponding reduced NTRU lattices to recover the full key. We collect power traces from the Falcon implementation in the PQClean library running on an ARM Cortex-M4 microprocessor. Our Gaussian-center classifier achieves single-trace classification accuracies of $0.9997$ and $0.884$ at the \texttt{-O0} and \texttt{-O3} optimization levels, respectively, confirming exploitable signed-exponent leakage from the Gaussian centers. Using these classification results, the full attack recovers Falcon-$512$ keys with success rates of $100\%$ from $20$ signatures at \texttt{-O0} and $100\%$ from $56$ signatures at \texttt{-O3}. For Falcon-$1024$, it reaches $100\%$ from $21$ signatures at \texttt{-O0} and $100\%$ from $100$ signatures at \texttt{-O3}.

Bootstrapping is the bottleneck of BGV/BFV homomorphic encryption, and for large plaintext primes $p$ its cost is dominated by digit extraction. Recently, this stage has been accelerated along two separate routes. The first lowers the degree of the digit-extraction polynomial: the bounded-support construction of Ma et al. (Eurocrypt'24) confines its support, and the order-four filter of Xiong et al. (to appear in Asiacrypt'26) removes three quarters of its monomials. The second lowers the depth of its evaluation: the Galois norm map of Okada et al. (Asiacrypt'23) and Zhao et al. (Crypto'26) evaluates a degree-$d$ factor in logarithmic depth. However, how the two routes relate and whether they compose has remained open. We propose a symmetry-graded framework that views the two evaluation routes as commuting group actions on the digit-extraction polynomial. Under this unified perspective, the existing evaluators correspond to different specializations of $\mathrm{cost}(D;r,d)$. For lower degree, we identify a novel rank-two lattice structure underlying digit extraction. The crystallographic restriction limits the filter order to $r\in\{1,2,3,4,6\}$. In particular, the methods of Ma et al. and Xiong et al. correspond to $r=2$ and $r=4$, respectively. Moreover, our framework derives a sparser order-six digit-extraction polynomial, enabling order-six symmetry for Mersenne primes. For lower evaluation cost, we construct a composed evaluator that first folds $P_A$ using the scalar filter and then evaluates the folded polynomial via the slot ring's Galois norm map. Since the rotation and Frobenius actions commute, the two optimizations can be combined within the same framework. Consequently, for any $(p,m)$ we select the optimal $(r,d)$ and evaluate digit extraction in $2\sqrt{D/(rd)}+O(\log D)$ non-scalar multiplications. On 13 general cyclotomic rings at $\geq 80$-bit verified security, our single-threaded HElib implementation improves upon the state-of-the-art evaluator of Ma et al., accelerating digit extraction by $2.4$–$4.9\times$ and thin bootstrapping by $1.3$–$2.8\times$, while modifying only the digit-extraction stage. In particular, on Ma et al.'s set IV with the Mersenne prime $p=8191$ and set V with $p=65537$, digit extraction is accelerated by $3.6\times$ and $3.8\times$, respectively, reducing the total bootstrapping time from 180.4 s to 85.5 s and from 217.6 s to 103.1 s. The implementation is publicly available, and the core analysis is machine-checked in Lean 4.

Let $C_0$ be the least constant such that, for every finite family of vectors $u_1,\ldots,u_n\in\mathbb C^d$ and independent finitely supported real random variables $\xi_i$, there are values $\varepsilon_i\in\operatorname{supp}(\xi_i)$ satisfying $\|\sum_{i=1}^n(\varepsilon_i-\mathbb{E}[\xi_i])u_i u_i^*\| \leq C_0\|\sum_{i=1}^n\mathrm{Var}[\xi_i](u_i u_i^*)^2\|^{1/2},$ where $\|\cdot\|$ denotes the operator norm. This rank-one matrix discrepancy formulation generalizes the signing formulation of the Kadison--Singer problem [KS59], resolved by Marcus, Spielman, and Srivastava [MSS15b]. Let $C_1$ be the least constant such that, for every $\epsilon>0$ and every finite family in $\mathbb C^d$, in every dimension $d$, satisfying $\sum_{i=1}^n u_i u_i^*=I$ and $\max_{i\in[n]}\|u_i\|^2\leq\epsilon$, there are signs $\varepsilon_1,\ldots,\varepsilon_n\in\{-1,1\}$ such that $\|\sum_{i=1}^n\varepsilon_i u_i u_i^*\|\leq C_1\sqrt\epsilon.$ Taking the $\xi_i$ to be independent symmetric signs gives $C_1\leq C_0$, since $\|\sum_{i=1}^n(u_i u_i^*)^2\|\leq\epsilon$ under these hypotheses. The general formulation allows arbitrary finite real supports and does not require $\sum_{i=1}^n u_i u_i^*=I$. Kyng, Luh, and Song [KLS20] proved $C_0\leq4$. We prove $2\leq C_1\leq C_0\leq2.176$. This improves the bounds $\sqrt2\leq C_0\leq3$ of Xie, Xu, and Zhu [XXZ21]. We conjecture that the optimal constant is $C_0=2$.

We present the first lattice-based, partially non-interactive threshold signature scheme that tolerates the adaptive corruption of up to T −1 signers, where T is the signing threshold. Our construction relies on the MSIS and MLWE assumptions, and has two rounds, of which only the second is message-dependent. We substantially improve upon prior adaptively secure lattice-based schemes (CRYPTO ’24 and EUROCRYPT ’26), which require at least two message-dependent rounds. In addition, compared to prior lattice-based partially non-interactive schemes (CRYPTO ’24, S&P ’25, CRYPTO ’25), we achieve better communication complexity.

Traditional secure multiparty computation (MPC) protocols assume a fixed set of participants throughout the computation. Dynamic MPC (DMPC) relaxes this assumption by permitting parties to join or leave during the process. However, practical deployment of DMPC in Internet settings favors tolerating asynchrony and achieving guaranteed output delivery (GOD). Recently, AD-MPC [CCS 2025] introduced an asynchronous DMPC (ADMPC) achieving this goal, but its reliance on per-epoch preprocessing and masking randomness incurs substantial concrete overhead. In this paper, we present Continuum, a concretely efficient ADMPC protocol with GOD. We introduce an aggregated hand-off protocol that ensures transfer consistency across epochs while eliminating per-epoch masking randomness, and a batch multiplication protocol based on the BGW paradigm that defers degree reduction to the hand-off phase and aggregates verification for all multiplications into succinct proofs. Together, these protocols enable a non-interactive computation phase with interaction concentrated in an efficient hand-off phase, overcoming the primary concrete inefficiencies of prior ADMPC designs. We implement Continuum and evaluate its performance on up to 128 geographically distributed servers. In experiments on 22 servers with circuits of depth 6 and width 100, Continuum reduces linear-gate transfer latency from 76.60s to 15.13s (80% reduction) and multiplication-gate evaluation time from 476.62s to 35.67s (93% reduction), compared with AD-MPC, while preserving strong security guarantees.

Low-entropy keys such as passwords and biometrics are pervasive, yet classical information-theoretic guarantees, such as perfect secrecy and entropic security, cannot generally be achieved in this regime. Honey encryption (HE) and homophonic ciphers (HC) tailor encryption to the message distribution to provide meaningful security, but their analyses do not establish optimal guarantees for arbitrary distributions. We ask: for messages and keys drawn independently from arbitrary distributions $(p_{\mathrm m},p_{\mathrm k})$, what is the optimal information-theoretic security achievable by any symmetric encryption scheme, and which conditions characterize its attainment? For key confidentiality (KC), the optimal guarantee is that the ciphertext $C$ reveals only negligible information about the key $K$, i.e., $I(K;C)=\operatorname{negl}(\ell)$. For uniform schemes, this is achieved if and only if decrypting under every fixed key yields a distribution statistically close to $p_{\mathrm m}$. HC and HE achieve this guarantee. For message-recovery security (MRS), the optimal guarantee is that no adversary guesses the message with probability above $p_{\max}+\operatorname{negl}(\ell)$, while a trivial adversary achieves at least $p_{\max}$ by guessing the most likely message or decrypting under the most likely key. We construct MRS-OE, a uniform scheme tailored to $(p_{\mathrm m},p_{\mathrm k})$ that attains $p_{\max}+O(2^{-\ell})$, and prove that $p_{\mathrm k}$-agnostic schemes, including HC and HE, cannot achieve optimal MRS in general. We also show that optimal KC and optimal MRS generally cannot be achieved simultaneously. Technically, we establish the KC characterization through flow and cut analysis on a weighted directed acyclic graph. We also introduce a continuous-ciphertext framework that separates structural constraints from discretization error.

We study the existence of bounded-degree supersingular isogenies under prescribed degree filters and exact torsion constraints. In characteristic $p$, a second-moment argument using the Brandt trace formula gives asymptotically positive coverage of independent uniform endpoints at $D=\lfloor c\sqrt{p}\rfloor$, for every fixed $c>0$ and every endpoint-independent binary filter retaining $\Omega(D^2)$ cyclic kernels. This avoids the logarithmic loss of a maximum-Petersson-norm estimate, but does not imply coverage tending to one. For curves equipped with a line in their $q$-torsion, where $q\neq p$ is prime, a cumulative Hecke estimate gives nonexistence $O\!\left(pq\ln^3(2pq)/D^2\right)$. Combined with a first-moment incidence bound, this separates two scalar-containing subgroups with the same upper-triangular span and additive Hom index $q$:their response-existence probabilities tend to one and zero when $q=p^{\beta+o(1)}$, $D=p^{1/2+\gamma+o(1)}$, and $0<\gamma<\beta<2\gamma<1$. The positive statement averages over a uniform source line and also yields an existential fixed-frame result; the exact challenge cardinalities differ. Determinant filters and finite odd-degree bounds at the SQIsign v3.0 parameters illustrate the results. All probabilities concern the stated arithmetic experiments, not the failure probability or security of a concrete signer.