But now, more than two centuries later, the mystery has finally been cracked. And the unlikely codebreaker? Artificial intelligence.
The encrypted letter was published in a French army journal in 1969, but its code remained unsolved until a cybersecurity engineer used OpenAI’s GPT-6 Astra model to decode it. Within six hours, the AI produced a readable translation, uncovering a partial decoding key compiled by French cryptology historian Daniel Tant.
AI Cracks 217-Year-Old Code in Six Hours
By analyzing a low-resolution image of the letter, the AI identified 33 symbols linked to Tant’s research, accounting for roughly one-third of the message. It then cracked the remaining two-thirds in six hours. The decrypted text revealed details about French military positioning during a critical period, as Austria prepared to launch a major offensive against Napoleon’s empire.
Historian Praises AI’s Role in Decoding History
Dr. Zack White, a history research fellow at the University of Portsmouth, called the achievement “a really interesting use of Artificial Intelligence,” noting its potential to crack coded messages and languages yet to be deciphered. The breakthrough highlights the growing role of AI in historical research, according to the sources.
Message Sent Amid European Tensions
The letter was sent during a time of heightened military activity, with French forces spread across Europe and significant numbers of troops tied up in Spain. The encryption was critical to preventing sensitive information from falling into enemy hands, as physical communication relied on couriers traversing vast and potentially hostile territories.
More on Napoleon Bonaparte
Keep reading