--- title: Discovering Cryptographic Weaknesses with Claude | daily.dev description: Anthropic researchers used Claude Mythos Preview to discover significant cryptographic weaknesses in two algorithms. The first attack improves on the... canonical: https://daily.dev/posts/discovering-cryptographic-weaknesses-with-claude-2toibqngb twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Discovering Cryptographic Weaknesses with Claude | daily.dev og:description: Anthropic researchers used Claude Mythos Preview to discover significant cryptographic weaknesses in two algorithms. The first attack improves on the... og:url: https://daily.dev/posts/discovering-cryptographic-weaknesses-with-claude-2toibqngb og:image: https://api.daily.dev/og/posts/2toiBQNgB.png og:image:alt: Discovering Cryptographic Weaknesses with Claude og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Discovering Cryptographic Weaknesses with Claude **[Hacker News](https://daily.dev/sources/hn)** · 19 min read · 1 upvotes · 0 comments ## Summary Anthropic researchers used Claude Mythos Preview to discover significant cryptographic weaknesses in two algorithms. The first attack improves on the best-known attack against HAWK, a post-quantum digital signature scheme under NIST review, effectively halving its key strength after just 60 hours of semi-autonomous work. The second introduces a novel 'Möbius Bridge' fingerprinting technique that improves attacks on 7-round AES by 200–800×. Neither attack affects production systems today, but both demonstrate that frontier AI models can now perform expert-level cryptanalysis autonomously. Additional preliminary attacks were found against LEA and Serpent-128. Anthropic also released CryptanalysisBench, a benchmark for evaluating LLM cryptanalytic capabilities, and plans to host an academic workshop on AI's role in security research. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Community take How the wider developer community reacted, aggregated from 1 discussion and 121 comments across hackernews (as of 2026-08-02). **TL;DR:** The community is genuinely impressed by Claude's cryptanalytic discoveries (especially the HAWK attack) but tempers enthusiasm with important caveats: the HAWK weakness was theoretically known beforehand, the $100k cost is eyebrow-raising, and the AES result targets a weakened variant under unrealistic attack conditions. **Sentiment:** 35% positive · 45% mixed · 20% skeptical **The case for** - The HAWK automorphism discovery concretely instantiated a previously only-theoretical attack, demonstrating LLMs are good at finding mathematical counterexamples and concrete constructions. - Multi-agent workflows showed interesting emergent dynamics where one agent's rejected idea was picked up and exploited by another, suggesting stochastic diversity has real value. - Commenters note that sloppy, typo-laden prompts still work well, validating that RLHF/fine-tuning has made models robust to imperfect human input. - AI augmenting expert researchers (rather than replacing them) is seen as a genuine and meaningful capability demonstrated here. **The pushback** - The HAWK 'weakness' was already known to be theoretically exploitable; finding the concrete automorphism is impressive but not discovering an unknown vulnerability as the PR implies. - The AES result targets only 7 of 10 rounds under a chosen-plaintext attack model that is no longer realistic in modern cipher modes like AES-GCM. - $100k in compute for results that may be reproducible on cheaper models is seen as wasteful and reflects a 'tech aristocracy' dynamic. - Massive publication bias: all the failed sessions are hidden, making the headline results look more impressive than the overall process warrants. - 'Prompt engineering' skepticism: the messy prompts used by Anthropic's own researchers undercut the idea that careful prompting is a real skill. **By community** - hackernews (mixed): Commenters are split between genuine excitement about AI-assisted cryptanalysis and pointed skepticism about overstated PR, unrealistic attack conditions, and the enormous compute cost. **Hottest debate:** Whether the HAWK result represents a truly novel AI discovery or merely the concrete instantiation of an already-known theoretical vulnerability, and how much credit Mythos deserves. **Open questions** - Could cheaper models (e.g. Chinese frontier models) replicate these results at a fraction of the cost? - How reproducible are these discoveries — what fraction of autonomous sessions actually find something useful? - Will similar AI-assisted analysis find weaknesses in SHA-256 reduced-round variants or other widely deployed primitives? - Should there be a coordinated disclosure framework specifically for AI-discovered cryptographic vulnerabilities with real-world impact? **Highlights** > TL;DR: They marginally improved on the best known academic attack on 7-round AES-128 (which normally uses 10 rounds - you do not need to worry about AES being broken). The attack on HAWK is perhaps more interesting - they were able to halve the effective key length. HAWK is a candidate for NIST standardisation. It has been studied academically, but isn't really deployed anywhere (because it hasn't been standardised!) > — [Retr0id on hackernews · 3 comments](https://news.ycombinator.com/item?id=49087623) > This seems like a bit of an overstatement: Despite HAWK having survived two rounds of expert human review over a period of two years, Mythos was able to improve the best-known attack on it in just 60 hours of work—effectively cutting its key strength in half. since, later: Mythos’s attack works by finding a specific, previously unexploited symmetry called a nontrivial automorphism in the lattice used by HAWK. Prior work proved that efficiently finding such an automorphism would permit an attack, but did not answer if such an automorphism was accessible in the lattice used by HAWK. The automorphism discovered by Mythos allows a faster enumeration attack that, while still exponential, means that one needs to double the size of HAWK keys to achieve the same level of security. Not downplaying Mythos's contribution here[1], but that first paragraph strongly hinted (at least to me) that there were no known weaknesses. "Discovering a weakness that had previously been only theoretical" is vastly different from "discovering an unknown weakness." Again: very cool Mythos was able to do this. It just seems like another case of "LLMs are good at finding concrete mathematical (counter)examples" - which is also cool! But the PR here is cynical. ...and it is kind of incredible to think that they spent $100,000 over 3 days looking for an automorphism. Not the possibility of an automorphism, that was already known. Man. [1] ... or focusing too hard on the strange use of mathematical language... > — [Diogenesian on hackernews · 2 comments](https://news.ycombinator.com/item?id=49088289) > > The multi-agent workflow led to interesting dynamics. For example, the key idea in producing this attack was discovered by a pair of workers working together. Both started investigating the idea; the first worker prematurely rejected the idea as infeasible, but the second found a way to fully exploit it. The pair kept exchanging messages, and eventually both agreed they had found an effective attack. this is pretty interesting. the way it is written doesn't make it sound like the collaboration actually led to the discovery, but rather just the stochastic nature of each thread in the search. it would be interesting to replay and repeat the search (possibly with prior/context pertubations) to get a sense for how often it finds or misses the known working path. > — [a-dub on hackernews · 3 comments](https://news.ycombinator.com/item?id=49087236) > “Each of the results cost roughly $100,000 in API cost to develop.” And “Over the course of a week, one Anthropic researcher worked together with Claude to develop the HAWK attack, and another researcher built a scaffold4 that allowed Claude to fully autonomously discover the AES attack.” Spending $100k in tokens in a week is an impressive feat even with massive parallelization. I suspect the TPS their internal folks have access to is far higher than their bulk public endpoints. There’s a tech aristocracy rapidly emerging in our society and it’s going to tear us apart. > — [mmaunder on hackernews · 3 comments](https://news.ycombinator.com/item?id=49087869) > When high quality effort is applied to a tool, such as AES or the linux kernel, we intuit that it "hardens" the tool. That is, it makes the tool more correct, more resilient, less assailable, etc. Similarly, when effort is applied to an open problem, such as the Riemann hypothesis or P v NP, without progress, it "hardens" the problem: it makes the problem feel more daunting to whoever takes a stab at it next. Andrew Wiles, whose interview also hit the homepage today (https://news.ycombinator.com/item?id=49075264), couldn't just tackle Fermat's Last Theorem head on, he had to wait until a different, modern problem reduced to it, because FLT had gathered this mystique of unassailability through its 300 years of existence. A thing I worry about is that as AI transmutes tokens into effort, it'll split the world into two: some problems will yield, making human effort entirely unnecessary, and others will harden to the point where human effort will feel increasingly less worthwhile, because "even AI couldn't solve it". I don't like this. AI is spiky, so I suspect it'll continue having major blind spots, and yet its mere presence will probably have a chilling effect on what would have otherwise been useful human effort. > — [staticshock on hackernews · 3 comments](https://news.ycombinator.com/item?id=49089200) **Source threads** - [hackernews](https://news.ycombinator.com/item?id=49087091) · 81 points · 121 comments ## Similar posts on daily.dev - [Claude Mythos Preview Exposes Hidden Code Flaws Fast](https://daily.dev/posts/claude-mythos-preview-exposes-hidden-code-flaws-fast-wdwyjdo3x) · IEEE Spectrum · 0 upvotes · 0 comments - [Claude Mythos Preview \\ red.anthropic.com](https://daily.dev/posts/claude-mythos-preview-red-anthropic-com-wkjczmbgu) · Hacker News · 5 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#quantum-computing](https://daily.dev/tags/quantum-computing), [#claude](https://daily.dev/tags/claude), [#cryptography](https://daily.dev/tags/cryptography) [View this post on daily.dev](https://daily.dev/posts/discovering-cryptographic-weaknesses-with-claude-2toibqngb) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Discovering Cryptographic Weaknesses with Claude","url":"https://daily.dev/posts/discovering-cryptographic-weaknesses-with-claude-2toibqngb","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/discovering-cryptographic-weaknesses-with-claude-2toibqngb"},"datePublished":"2026-07-28T18:08:38.188Z","dateModified":"2026-08-02T19:34:01.754Z","description":"Anthropic researchers used Claude Mythos Preview to discover significant cryptographic weaknesses in two algorithms. The first attack improves on the...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/451e184fd71d31aa7ccda69de188fca6?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/451e184fd71d31aa7ccda69de188fca6?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/discovering-cryptographic-weaknesses-with-claude-2toibqngb","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,quantum-computing,claude,cryptography","timeRequired":"PT19M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"Discovering Cryptographic Weaknesses with Claude"}]} ```