Headline: Anthropic’s Claude Mythos Reveals Major Flaw in Post-Quantum Signature Candidate TL;DR — An unreleased version of Anthropic’s AI, Claude Mythos Preview, discovered two previously unknown attacks on cryptographic algorithms. One targets HAWK, a post-quantum digital-signature scheme that had advanced to NIST’s third round and was the last lattice-based contender. Claude found a mathematical symmetry that cuts the cost of stealing HAWK’s smallest key from 2^64 operations to 2^38 — about 67 million times less work. Fixing the flaw essentially doubles HAWK’s key sizes and erodes the advantages that made it attractive: compact keys and fast signing. What happened - Anthropic says Claude Mythos Preview identified two novel vulnerabilities in crypto schemes; the most consequential hit HAWK, a signature system designed to prove a transaction came from a user without revealing their private key and to resist quantum attacks. - NIST had advanced HAWK to Round 3 of its post-quantum standardization process in May. Claude found a previously unused symmetry in HAWK’s math that dramatically lowers the computational effort needed to break its smallest keys (from 2^64 to 2^38). Why it matters - The practical fallout: patching HAWK would require roughly doubling key sizes. That undermines HAWK’s core selling points — small signatures and fast signing — which directly affect block-space usage and transaction fees on blockchains. - Important caveat: HAWK has never been deployed and Bitcoin still uses ECDSA (a pre-quantum scheme). No live networks have been broken by this discovery. - Still, the event highlights a central risk for the industry: post-quantum plans (from consortiums like BlackRock-Coinbase-Strategy, to Vitalik’s roadmap and projects at NEAR and Zcash) hinge on the soundness of replacement primitives. Claude’s finding shows a leading candidate can be compromised before shipping. What’s next - There’s a silver lining: the same AI that finds attacks can be used to audit and harden cryptography. Firms and standards groups are racing to apply advanced models to their own code and designs to surface vulnerabilities proactively. - The key question now is whether AI tools will break cryptography faster than they help rebuild it — and how quickly the community can adapt and re-evaluate post-quantum candidates. Source: Morning Minute by Tyler Warner.
Anthropic's Claude Mythos Discovers Major Flaw in NIST Post-Quantum Signature Candidate HAWK
ChainGPTShare
Summary
Anthropic’s Claude Mythos Preview uncovered a major flaw in HAWK, a post-quantum digital signatures candidate in NIST’s third round. The AI found a mathematical symmetry that cuts the effort to break HAWK’s smallest keys from 2^64 to 2^38 operations. Developers now face the need to double key sizes, which harms the algorithm’s compact design. Though HAWK is not yet deployed, the flaw raises concerns about post-quantum security. The incident shows how AI can help strengthen cryptographic systems. Layer 2 scaling solution developers may need to reassess their cryptographic dependencies.
Source:Show original
Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information.
Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.