OpenAI, Anthropic models crack unsolved coded messages used by Nazi Germany

OpenAI’s GPT-6 Astra and Anthropic’s Claude Opus 5 have successfully cracked long-unsolved messages encoded by Enigma machines used by Nazi Germany during World War II. Developer Carter Leffen used Astra to decode an unbroken message from a database of Enigma messages. Cybersecurity executive Jack Willis guided Claude Opus 5 to crack a second message using a German officer's signature.