--- title: Astra and Opus just passed Turing's other test | daily.dev description: Two independent cryptanalysts used frontier LLMs to crack previously unsolved WWII-era Enigma-encoded messages. Developer Carter Leffen prompted OpenAI's new... canonical: https://daily.dev/posts/astra-and-opus-just-passed-turing-s-other-test-nmqht3ysb twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Astra and Opus just passed Turing's other test | daily.dev og:description: Two independent cryptanalysts used frontier LLMs to crack previously unsolved WWII-era Enigma-encoded messages. Developer Carter Leffen prompted OpenAI's new... og:url: https://daily.dev/posts/astra-and-opus-just-passed-turing-s-other-test-nmqht3ysb og:image: https://api.daily.dev/og/posts/nmQhT3ysb.png og:image:alt: Astra and Opus just passed Turing's other test 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. # Astra and Opus just passed Turing's other test **[TechCrunch](https://daily.dev/sources/tc)** · 3 min read · 0 upvotes · 0 comments ## Summary Two independent cryptanalysts used frontier LLMs to crack previously unsolved WWII-era Enigma-encoded messages. Developer Carter Leffen prompted OpenAI's new Astra model to search an archive, do research, build an Enigma simulator, and recover plaintext of a message unsolved since 2005. Separately, Jack Willis used Anthropic's Claude Opus 5, with more guidance, to break another message by leveraging a known officer's signature. Cryptology researcher Frode Weirerud validated both solutions, noting Astra behaved like a professional archive researcher and completed in two days what would take a human weeks or months. Only seven unbroken Enigma messages remain. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How did someone use an AI model to break an unsolved Enigma-coded message from World War II? Developer Carter Leffen instructed OpenAI's new Astra model to search a database of Enigma messages for an unbroken one and decode it. The model performed its own archival research, found context clues, built a simulator of the Enigma machine, and recovered the plaintext of a message that had baffled researchers since 2005. _Developers exploring what autonomous AI agents can research and build follow cases like this on daily.dev._ ### How many Enigma messages from World War II remain unbroken? Seven Enigma messages remain completely unbroken, plus one additional message where the plaintext is known but the underlying code has not been solved. Cryptology researcher Frode Weirerud, who maintains the Crypto Cellar archive, tracks and validates attempts to solve these remaining messages. _Anyone tracking breakthroughs in applied AI research can follow developments like this on daily.dev._ ### Which AI model was used to break a WWII Enigma message using an officer's known signature? Anthropic's Claude Opus 5 was used by cryptanalyst Jack Willis to break a different unsolved Enigma message, with significantly more human guidance than the Astra case. Willis leveraged the known signature of a particular officer's name within the message to ultimately decode it. _Developers comparing model capabilities across tasks can keep up with results like this on daily.dev._ --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-agents](https://daily.dev/tags/ai-agents), [#cryptography](https://daily.dev/tags/cryptography), [#gpt](https://daily.dev/tags/gpt) [View this post on daily.dev](https://daily.dev/posts/astra-and-opus-just-passed-turing-s-other-test-nmqht3ysb) ```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":"Astra and Opus just passed Turing's other test","url":"https://daily.dev/posts/astra-and-opus-just-passed-turing-s-other-test-nmqht3ysb","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/astra-and-opus-just-passed-turing-s-other-test-nmqht3ysb"},"datePublished":"2026-09-25T17:29:35.855Z","dateModified":"2026-09-25T17:30:47.100Z","description":"Two independent cryptanalysts used frontier LLMs to crack previously unsolved WWII-era Enigma-encoded messages. Developer Carter Leffen prompted OpenAI's new...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/a867c6f1f345a9a963b4c86a7dc36b2e?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/a867c6f1f345a9a963b4c86a7dc36b2e?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"TechCrunch","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":"TechCrunch","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/tc","url":"https://daily.dev/sources/tc"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/astra-and-opus-just-passed-turing-s-other-test-nmqht3ysb","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ai-agents,cryptography,gpt","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"TechCrunch","item":"https://daily.dev/sources/tc"},{"@type":"ListItem","position":3,"name":"Astra and Opus just passed Turing's other test"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/astra-and-opus-just-passed-turing-s-other-test-nmqht3ysb#faq","mainEntity":[{"@type":"Question","name":"How did someone use an AI model to break an unsolved Enigma-coded message from World War II?","acceptedAnswer":{"@type":"Answer","text":"Developer Carter Leffen instructed OpenAI's new Astra model to search a database of Enigma messages for an unbroken one and decode it. The model performed its own archival research, found context clues, built a simulator of the Enigma machine, and recovered the plaintext of a message that had baffled researchers since 2005. Developers exploring what autonomous AI agents can research and build follow cases like this on daily.dev."}},{"@type":"Question","name":"How many Enigma messages from World War II remain unbroken?","acceptedAnswer":{"@type":"Answer","text":"Seven Enigma messages remain completely unbroken, plus one additional message where the plaintext is known but the underlying code has not been solved. Cryptology researcher Frode Weirerud, who maintains the Crypto Cellar archive, tracks and validates attempts to solve these remaining messages. Anyone tracking breakthroughs in applied AI research can follow developments like this on daily.dev."}},{"@type":"Question","name":"Which AI model was used to break a WWII Enigma message using an officer's known signature?","acceptedAnswer":{"@type":"Answer","text":"Anthropic's Claude Opus 5 was used by cryptanalyst Jack Willis to break a different unsolved Enigma message, with significantly more human guidance than the Astra case. Willis leveraged the known signature of a particular officer's name within the message to ultimately decode it. Developers comparing model capabilities across tasks can keep up with results like this on daily.dev."}}]} ```