SentinelOne engineer uses GPT-6 Astra to decode 217-year-old Napoleonic cipher in six hours
Carter Church, an engineer at SentinelOne, says he used OpenAI's GPT-6 Astra model to decipher a 1809 coded military dispatch that had remained unread on Cryptiana's list of unsolved historical ciphers. The message, sent from the headquarters of Eugène de Beauharnais to General Auguste de Marmont relaying Napoleon's orders, was reportedly decoded from a single image and prompt within about six hours. Church has published a solution package and script online so others can verify the result.
GoKawiil's interpretation of the reporting above, not reported fact.
If verified by historians and cryptographers, the feat would suggest advanced multimodal AI models can now perform complex, multi-step research tasks—reading handwriting, inferring cipher structure, and cross-checking historical context—without specialized training for the task. Church notes the decoded text appears to complete a sentence left incomplete in Napoleon's own 1865 memoirs, which could be of genuine interest to historians studying the Napoleonic Wars. The claim currently rests on one engineer's account and self-published materials, so independent verification would be needed to confirm its historical accuracy.
- GPT-6 Astra reportedly deciphered a 217-year-old unsolved Napoleonic-era cipher in roughly six hours from one image and prompt.
- The decoded message was a troop briefing from Eugène de Beauharnais's headquarters to General Auguste de Marmont relaying Napoleon's orders.
- The decoded text reportedly fills a gap in Napoleon's 1865 memoirs, completing a previously incomplete sentence.
Kryptos Cipher Puzzle Book — If AI cracking a 217-year-old Napoleonic cipher has you fascinated by codebreaking, this puzzle book lets you try your own hand at classic and modern ciphers. It's a fun, hands-on way to engage with the same kind of cryptographic challenges that stumped historians for centuries before AI stepped in.
See Kryptos Cipher Puzzle Book on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.Source: tomshardware.com — Mark Tyson, 2026-10-04
Published there as: “ChatGPT-6 Astra cracks 217-year-old Napoleonic code in just six hours”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.