Astra and Claude Opus 5 Decode Two Enigma Messages

Two cryptanalysts say they used large language models from OpenAI and Anthropic to decode two previously unsolved messages encrypted with Enigma machines.

Developer Carter Leffen asked OpenAI’s Astra model to find an unbroken message in a database and decode it. TechCrunch reports that Astra researched archival material, identified contextual clues, built an Enigma simulator and recovered the plaintext of a message that had stumped researchers since 2005.

Frode Weirerud, a retired electrical engineer who maintains the Crypto Cellar website and its message database, validated Leffen’s solution last week. Weirerud said he was unsure whether Astra had accessed archived messages from a “private collection” mentioned in its logs. He speculated they might have been shared online by another researcher or available in German government archives.

On September 21, cryptanalyst Jack Willis told Weirerud he had used Anthropic’s Claude Opus 5 to decode a different unsolved message. Willis gave the model more guidance, including a particular officer’s known signature, which helped it solve the message.

Weirerud says seven unbroken Enigma messages remain, plus one message whose plaintext is known but whose code remains unsolved. Some archival messages have resisted solutions because of transcription mistakes or errors by their original encoders.