DeFi Intel

Morning Minute: Claude Mythos Breaks Post-Quantum Cryptography

Plus markets are mixed ahead of FOMC; new Wall St. giants are backing the Clarity Act; and Zcash officially launched its Ironwood upgrade

TL;DR

Anthropic’s Claude Mythos AI discovered a key‑recovery attack that halves HAWK’s security, leading to its withdrawal from NIST’s post‑quantum round. The finding adds AI‑driven risk to blockchain projects like Stellar and Tron, which are racing to deploy quantum‑resistant upgrades.

Key facts

  • Claude Mythos found a symmetry‑based key‑recovery attack on HAWK‑512, reducing security from 2^150 to 2^108.
  • HAWK was withdrawn from the NIST post‑quantum standardization round in July 2026.
  • AI analyses suggest Classic McEliece’s key‑recovery complexity may be as low as 2^94 operations.
  • Germany’s BSI advised against deploying Classic McEliece by early October 2026.
  • Stellar’s Bell upgrade will hit testnet on November 3 and may launch on mainnet on November 23, adding native post‑quantum signatures for smart contracts.

On July 29, 2026, the crypto community learned that Anthropic’s AI model Claude Mythos had uncovered a symmetry‑based key‑recovery attack against the post‑quantum signature scheme HAWK. The breakthrough, reported by Project Eleven co‑founder Conor Deegan, showed that the attack could halve HAWK‑512’s effective security from 2^150 to 2^108. The same AI‑driven research also highlighted weaknesses in the Classic McEliece scheme, prompting Germany’s BSI to advise against its deployment. These findings arrived as major blockchains such as Stellar, Tron and Starknet were racing to roll out quantum‑resistant upgrades, underscoring a new, AI‑focused threat to post‑quantum cryptography.

Background

Post‑quantum cryptography has become a strategic priority for blockchain platforms because powerful quantum computers could eventually break the elliptic‑curve algorithms that protect most digital assets today. Google has set an internal 2029 deadline for quantum‑secure systems and urged the broader industry to follow suit. In response, Stellar announced its Bell upgrade, slated for November 2026, to introduce native post‑quantum signatures for smart contracts and to prepare enterprise wallets for quantum‑secure signing by the end of the year, with a broader rollout to standard wallets in 2027. Tron’s founder Justin Sun similarly declared that the network’s quantum‑resistant cryptography was live on testnet and ready for mainnet deployment at any time. Meanwhile, a consortium of blockchain projects, including Bitcoin, Ethereum and Algorand, have each published proposals or pledged funding to develop quantum‑resistant solutions. The emergence of AI‑driven attacks, exemplified by Claude Mythos’s discovery, adds a new dimension to the security landscape: rather than waiting for quantum hardware, adversaries can now use sophisticated machine‑learning models to find mathematical flaws in candidate post‑quantum schemes, potentially accelerating the timeline for necessary upgrades.

What happened

Project Eleven co‑founder Conor Deegan reported on October 19, 2026 that Anthropic’s AI model Claude Mythos had identified a symmetry‑based key‑recovery attack on the HAWK digital signature scheme, a candidate in the NIST post‑quantum standardization round. The attack reduced HAWK‑512’s effective security from 2^150 to 2^108, prompting the withdrawal of HAWK from the NIST round in July 2026, as noted in the cryptobriefing report. Deegan also highlighted that AI analyses had lowered the estimated key‑recovery complexity for Classic McEliece to as low as 2^94 operations, leading Germany’s Federal Office for Information Security (BSI) to advise against its deployment by early October 2026. In parallel, Stellar announced its Bell upgrade (Protocol 30) will hit testnet on November 3 and, if successful, move to mainnet on November 23, introducing native post‑quantum signatures for smart contracts and targeting enterprise wallet security by the end of 2026, according to Ambcrypto. Tron’s founder Justin Sun posted on X that Tron’s post‑quantum cryptography was live on testnet and ready for mainnet at any time, as reported by Cointelegraph. Together, these developments illustrate a simultaneous push by blockchain projects to adopt quantum‑resistant technologies while AI research exposes vulnerabilities in the very schemes they plan to use.

Market impact

Following the announcement of Stellar’s Bell upgrade, XLM’s price moved about 5% lower before rebounding 7% over the week, with bulls defending a September low near $0.17 and market structure remaining bearish unless the token clears $0.19, according to Ambcrypto. While the Claude Mythos breakthrough itself did not trigger immediate price swings, the broader narrative of post‑quantum readiness has kept investor attention on projects like Tron and Starknet, which are also positioning themselves as quantum‑resistant. No specific price data for Tron or other assets were provided in the sources, but the reported market reaction to Stellar’s news suggests that security upgrades continue to influence short‑term trading dynamics in the crypto space.

Why it matters

The AI‑driven attack on HAWK demonstrates that post‑quantum cryptographic schemes can be compromised without a quantum computer, compressing the timeline for identifying and mitigating weaknesses. This challenges the assumption that quantum hardware is the sole future threat and forces the industry to consider AI as a parallel adversary. For blockchain platforms, the findings raise urgent questions about the robustness of their planned quantum‑resistant upgrades, such as Stellar’s Bell upgrade and Tron’s testnet deployment. If AI can uncover flaws in candidate algorithms, projects may need to adopt hybrid signing approaches, like Project Eleven’s Strongpoint platform, which pairs classical and post‑quantum signatures to hedge against both quantum and AI‑based attacks. The BSI’s advisory against Classic McEliece further signals regulatory caution, potentially influencing compliance decisions for custodians and exchanges. Overall, the Claude Mythos discovery underscores the need for continuous cryptographic auditing and modular algorithm designs to stay ahead of rapidly evolving attack vectors.

DeFi Intel’s take

Editorial analysis, not a statement of fact.

DeFi Intel's take: the Claude Mythos breakthrough suggests that the post‑quantum security race is now a two‑front battle, with AI tools capable of exposing mathematical flaws faster than traditional analysis. We read this as a warning that blockchain projects cannot rely solely on the promise of quantum‑resistant algorithms; they must also build flexibility into their cryptographic stacks. Hybrid signing solutions and modular algorithm swaps, as demonstrated by Project Eleven’s Strongpoint, appear to be prudent paths forward. The market’s modest reaction to Stellar’s upgrade indicates that investors are already pricing in some security risk, but the lack of broader price turbulence may reflect confidence that existing upgrades remain safe for now. Continued vigilance and rapid response to AI‑found vulnerabilities will likely become a core component of blockchain security strategies.

Sources used for this analysis

DeFi Intel wrote this analysis from the reporting above. Any figure or quotation can be checked against those pages.

Frequently asked questions

What specific attack did Claude Mythos discover?

Claude Mythos identified a symmetry‑based key‑recovery attack on the HAWK digital signature scheme, which halved HAWK‑512’s effective security from 2^150 to 2^108, according to Project Eleven’s co‑founder Conor Deegan.

How is Stellar planning to address post‑quantum security?

Stellar’s Bell upgrade, scheduled for November 2026, will introduce native post‑quantum signatures for smart contracts and aims to have enterprise wallets quantum‑secure by the end of 2026, with broader wallet and validator upgrades planned for 2027.

What regulatory advice has been issued regarding Classic McEliece?

Germany’s Federal Office for Information Security (BSI) advised against deploying Classic McEliece by early October 2026 after AI analyses suggested its key‑recovery complexity could be as low as 2^94 operations.

Direct answers

Short answers, written to be quoted.

What did Claude Mythos discover about HAWK?
Claude Mythos discovered a symmetry‑based key‑recovery attack that halves HAWK‑512’s effective security from 2^150 to 2^108, prompting its withdrawal from NIST’s post‑quantum round.
How is Stellar preparing for quantum threats?
Stellar’s Bell upgrade, scheduled for November 2026, will introduce native post‑quantum signatures for smart contracts and aims to secure enterprise wallets by the end of the year.
What warning did Germany’s BSI issue?
Germany’s Federal Office for Information Security warned against deploying the Classic McEliece scheme because AI research indicated its key‑recovery complexity could be far lower than previously thought.
Read full article on Decrypt →

DeFi Intel is an entity-graph aggregator: we curate, tag and link crypto news to a typed knowledge graph of protocols, tokens, people and incidents. We do not republish the full article body. Use the link above to read the original report at Decrypt.

Discussion

Comments are reviewed before publication.

No comments yet.

Want the full article?

Continue reading on Decrypt →

Stay current on entity-tagged crypto news

Get the weekly DeFi Intel brief — entity-graph intelligence covering protocols, tokens and incidents, free to your inbox.