Astra and Opus just passed Turing's other test

TechCrunch
OpenAI's Astra and Anthropic's Claude Opus 5 LLMs have cracked two long-unsolved WWII Enigma messages, stunning cryptanalysts.

Summary

While Alan Turing is famous for the Turing Test, his wartime work cracking the Enigma code stands as arguably his greater achievement. Now, two large language models—OpenAI's GPT-6 Astra and Anthropic's Claude Opus 5—have independently cracked two long-unsolved Enigma messages that had baffled researchers for years. Developer Carter Leffen tasked Astra with finding and decoding an unbroken Enigma message; the model conducted its own archival research, identified context clues, built an Enigma machine simulator, and recovered plaintext of a message unresolved since 2005. Cryptology enthusiast Frode Weirerud, who maintains the Crypto Cellar database, validated the solution and praised Astra as behaving like a professional cryptanalyst. Separately, cryptanalyst Jack Willis used Claude Opus 5 with more explicit guidance to break a different unsolved message using a known officer's signature. Weirerud notes only seven unbroken Enigma messages remain (plus one with known plaintext but an unbroken code), and the pace at which AI is solving them suggests the last ones may soon fall.

(Source:TechCrunch)