A longstanding cryptographic mystery dating back to the 17th century has been solved by an AI language model named Claude Fable 5.1. The puzzle, known as Sir Thomas Urquhart’s Cyphral Distich, consists of two lines of 32 numbers each, encoding a secret message that had remained unsolved for centuries. Previous attempts by historians and cryptographers failed because they overlooked a crucial clue embedded within the original text.

The Cyphral Distich appears at the end of Urquhart’s work Logopandecteision and was first posed as an open problem in 1899. It resurfaced in cryptography literature and was listed among the top 50 unsolved historical ciphers. Traditional methods such as frequency analysis and substitution ciphers did not yield results.

Claude Fable 5.1 approached the problem by recognizing two key hints: the number 32, which corresponds to both the number of Proquiritations (short poetic passages) preceding the cipher, and the repeated theme of "wishes" in the text. The AI deduced that each number in the cipher refers to a word index within the corresponding Proquiritation, and the first letter of that word forms the plaintext.

Using this method, the AI revealed a royalist prayer: "O GOD UPHOLD KING CHARLS THE SECOND AND MAKE HIM THE SUPREME RULER OF THIS LAND." The message aligns with Urquhart’s political sympathies and fits the expected poetic structure.

The AI also tackled a larger related cipher, the Cyphral Octastich from Urquhart’s 1652 work The Jewel. This more complex cryptogram uses page numbers as indices rather than paragraphs. Claude Fable 5.1 decoded nearly the entire message, which is another royalist prayer written in verse.

While some minor uncertainties remain due to historical text variations and missing original copies, the AI’s solution is largely verified and consistent with the source material.

This achievement demonstrates AI’s growing capability to solve not only mathematical problems but also historical and archival puzzles that have traditionally required extensive human effort and domain knowledge. The AI’s success came from persistence and creative analysis rather than advanced cryptanalysis techniques, suggesting new possibilities for applying AI to uncover lost knowledge and solve longstanding mysteries.