Cover photo

ZK Insights | 16 Oct 2026

#133

This week we cover Aztec's post-quantum roadmap, cryptographic package integrity, Lean proof guarantees, useful proof of work, reusable execution proofs, post-quantum proof verification and account keys, private stablecoins, distributed key generation, Magnetar, asynchronous threshold FHE, Bitcoin vaults, private face recognition, post-quantum secure aggregation, efficient edabits, inclusion proofs, and private Transformer inference benchmarks.

Blogs

Post-Quantum Aztec

  • Zac Williamson, Mike Connor

  • Recommend:

  • Distinguishes protecting today's private data against harvest-now-decrypt-later attacks from a full post-quantum migration, with a 2030 migration target spanning proving, authentication, encryption, and Ethereum settlement.

  • https://aztec.network/blog/post-quantum-aztec

Package managers have an integrity problem

When the Natural-Language Paper and the Lean Proof Disagree

How LLMs might make crypto mining useful

Proof of useful work for verifying AI treaties

Research Discussions

Rapidblocks with split proofs: execution proofs that outlive their state

ERC-8403: an account as one hash, and post-quantum keys

Private Stablecoin on Ethereum

Exposing an ISA for post-quantum proofs on Ethereum

Talks and Resources

TORTL DKG: Two-Round Trustless Distributed Key Generation

  • Florian Le Mouël, NIST TCPT3

  • An October 13 presentation covering elliptic-curve and module-lattice DKG variants, using secret sharing and zero-knowledge proofs to check correctness. The official page provides an abstract and a technical writeup; a recording link has not been confirmed.

  • https://csrc.nist.gov/presentations/2026/tcpt3-4b1

Papers

Magnetar: High-speed post-quantum RV64 zkVM from standard hashes

Tulva: Practical Asynchronous Threshold FHE

  • Valerio Cini, Antonio Guimarães, Russell W. F. Lai, Michał Osadnik, Ivy K. Y. Woo

  • Recommend:

  • Enables threshold decryption without requiring participants to coordinate or be online simultaneously, with an implementation and evaluations for privacy-preserving EVM use cases.

  • https://eprint.iacr.org/2026/2485

Trustless Bitcoin Vaults

  • Lukas Aumayr, Xinshu Dong, Runchao Han, Srivatsan Sridhar, David Tse, Nikhil Vanjani

  • Recommend:

  • Constructs vaults for collateral-based applications where withdrawals follow application contract logic without trusting a third party, including an Aave v4 lending example. First submitted October 14 and revised October 15.

  • https://eprint.iacr.org/2026/2466

GhostFaceMPC: Privacy-Preserving Face Recognition with Multi-Party Computation

  • Davide Carnemolla, Dario Catalano, Mario Di Raimondo

  • Covers facial feature extraction and matching under MPC without exposing raw images or biometric embeddings, using model quantization and circuit optimizations with open-source code.

  • https://eprint.iacr.org/2026/2499

SecAgg++: A Post-Quantum Privacy Preserving Federated Learning Protocol

  • Alessandro Perez, Giulia Cavicchioni, Silvio Cretti, Riccardo Longo, Alessandro Tomasi, Silvio Ranise

  • Integrates post-quantum primitives into single-server secure aggregation while preserving efficiency and supporting recovery when clients drop out.

  • https://eprint.iacr.org/2026/2497

Edabits Generation with Quasi-Linear Communication Complexity

  • Alexander Bienstock, Daniel Escudero, Qijia Fan, Fengrun Liu, Yifan Song

  • Reduces edabit generation communication from quadratic to quasi-linear when a constant fraction of parties are honest, improving building blocks for MPC comparison, truncation, and bit decomposition.

  • https://eprint.iacr.org/2026/2482

FlatMerkle: Logarithmic-Space Inclusion Proofs over a Single Linear Hash Stream

  • Sergio Demian Lerner

  • Supports logarithmic-size inclusion proofs over a single linear hash stream, with applications to Bitcoin transaction annexes, data availability, and cross-chain cost comparisons.

  • https://eprint.iacr.org/2026/2468

A Graph-Based Analysis of Comparability, Scale, and Attribution in Private Transformer Inference

  • Souvik Das, Ariful Azad, Nitesh Saxena

  • Analyzes 62 private Transformer inference systems; only one of 1,891 candidate system pairs is directly comparable under shared assumptions, highlighting pitfalls in performance comparisons.

  • https://eprint.iacr.org/2026/2464

Heterogeneous CPU–NPU Implementation of the AIMer Post-Quantum Signature for Low Latency on a Mobile SoC

Robust ECDSA Threshold Signing at Scale

Speeding up Sumcheck using Algebraic Geometry

Issuer-Independent Decentralized Exchange for E-Cash

TenseUp: Doubly-Efficient Lookup Arguments from Tensor Codes and Applications to Dynamic SNARKs

Censorship Resistance in Credential Systems and Application to the EUDI Wallet

A Black-Box Compiler from Static to Dynamic SNARKs

CUTS: Chained Uniform-sum Tree Signatures

Secret Sharing from Non-Extractable Randomness

Witness-Succinct UC-Secure SNARKs with Weaker Trust Assumptions