about
An AI Approach to Verified Production Cryptographic Libraries (arxiv.org)
2 points by sbulaev 60 days ago | hide | past | pdf | discuss on HN

In plain words: A tool turns plain API contracts into internal specs and machine-checked proofs for real cryptographic code, leaving the code untouched. It verified a widely used encryption routine and rebuilt a proof that took a human team months, finishing in 11.4 hours.

Abstract

Cryptographic code is critical infrastructure that must be correct, yet formally verifying production libraries remains difficult. Existing language-model proof systems solve isolated obligations with specifications and premises already given, leaving production-library verification unresolved. We present CryptoProver, an AI-based system that synthesizes internal specifications and Verus-checked proofs from high-level API contracts. Without changing executable code, CryptoProver constructs a new independent proof of curve25519-dalek and verifies RustCrypto's previously unverified chacha20 implementation against an RFC 8439 specification. These cryptographic lineages underpin deployed systems including Signal and Shadowsocks; Signal has an estimated 218M global downloads. The independent, human-led curve25519-dalek verification was developed publicly over eight months by five main contributors. Given the API contracts and a fixed trusted library of field specifications, arithmetic facts, axioms, and vstd, CryptoProver synthesizes the internal specifications and proofs in 11.4 hours with USD 466.99 in recorded API cost. CryptoProver follows a trust-first design principle: mechanical gates reject specification weakening, invented axioms, and cross-module breakage, while isolation blocks reference proof retrieval, including from git history.

Chuyue Sun, Su Fong, Zhiyi Kuang, Yizheng Jiao, Nina Narodytska, Haoze Wu, David L. Dill, Clark Barrett
arXiv:2608.00965 · cs.CR, cs.AI · submitted Aug 2, 2026
abstract · pdf · html

add comment on HN