Viral Claim: "GPT-6 Astra" Cracked a WWII Enigma Message; Unverified

Starting Sept 17, a viral post claimed "GPT-6 Astra" autonomously decrypted a 1941 unsolved German Army Enigma message in 10 hours, reportedly a march-route request for the Rosenow area. The post described a fully autonomous workflow: searching historical archives, comparing uncertain letters, finding contextual clues, building an Enigma simulator, writing cryptanalysis code, testing candidate keys in parallel, and cross-validating the recovered plaintext. The original source traces to @carterleffen, amplified by @tdhopper, @IgorCarron and @npew; on Sept 18 Polymarket's official account also posted the claim.

Confirmed

Unconfirmed

Why it matters

If true, it would mark fully autonomous long-horizon AI in archival research and classical cryptanalysis; but an "unreleased product achieves a hard decryption" narrative is a classic high-spread, low-credibility story, and betting platforms have incentives to drive topic interest. Absent official or independent third-party verification, this is best treated as a case study in AI rumor propagation and marketing manipulation.

2026-09-17 ~ 2026-09-18 · 5 related posts

Primary sources

1 near-duplicate retellings: tdhopper