Artificial intelligence has cracked historical ciphers that remained unsolved for centuries, including a 200-year-old wartime message ordered by Napoleon Bonaparte. This breakthrough, achieved within hours by AI tools like OpenAI's GPT-6 Astra, marks a turning point in how historians and cryptographers approach encrypted documents.
- AI cracked 200-year-old Napoleonic cipher in hours
- Most ciphers on expert lists solved since September 2026
- AI reshapes historical research and manuscript analysis
What happened
A decades-old encoded message commissioned by Napoleon Bonaparte remained a mystery for over 200 years until artificial intelligence managed to decrypt it in just a matter of hours. AI engineer Carter Church used the advanced model GPT-6 Astra to break this 19th-century cipher, a task that would have taken weeks using traditional methods. Since mid-September 2026, AI models have solved most of the historical ciphers maintained on a well-known experts' list, signaling a rapid shift in decoding capabilities.
The success in cracking these codes is largely attributed to AI's ability to persistently analyze patterns and data without the manual toil previously required. AI’s impact extends beyond individual ciphers, enabling researchers to unlock hidden writings in ancient scrolls and to transcribe millions of historical manuscript pages, vastly speeding up work that was once painstaking and slow.
Why it matters
This AI-driven advancement represents a pivotal moment for historians, cryptographers, and researchers, as it lowers barriers to solving complex encrypted texts that hold vital historical insights. Previously, codebreaking was a specialized field reserved for experts and dedicated amateurs. Now, AI tools allow non-specialists to participate and contribute meaningfully, broadening the community engaged in examining historical data.
However, this rapid progress also brings mixed feelings among veteran codebreakers. Some express concern over losing the intellectual challenge and satisfaction of manual decryption since AI simplifies tasks once considered difficult. This sentiment highlights broader debates on AI’s impact on professional and creative fields, where human skill and effort may be diminished as machines take over more complex functions.
What to watch next
Future developments will focus on cracking remaining stubborn ciphers that lack sufficient data for AI to analyze effectively, keeping some mysteries alive for years. Additionally, researchers are expected to expand AI’s role in digitizing and interpreting vast archives of ancient manuscripts across multiple languages and regions, including fragile texts like the carbonized scrolls from Vesuvius and extensive collections of Hebrew and Latin documents.
As AI continues to evolve, plans to create large, searchable databases from historical sources could revolutionize the understanding of cultural and intellectual history. These efforts will likely create new digital tools resembling historical 'social networks,' enabling scholars to trace relationships and ideas from the past more comprehensively than ever before.