Amateurs Are Using GPT-6 to Crack Decades-Old Ciphers, Overwhelming Cryptographers
ChuckDBrooks · x · 2026-10-19
Nature reports that LLMs have become a new tool for breaking historical codes — and are swamping cryptographers.
- Carter Leffen, with no cryptography background, used GPT-6 Astra to decrypt a German army message that had been sealed for 85 years: he asked the model to decode it, nudged it to keep thinking, and let it work overnight — a solution was waiting in the morning.
- A wave of non-experts is now using LLMs to crack historical ciphers, leveraging the models' ability to rapidly deploy decryption algorithms while searching the web for supporting historical documents.
- Cryptography specialists welcome the surge of interest but are flooded with requests to validate AI-derived solutions, which are frequently wrong (article text truncated at this point).
Source: Nature news, 16 October 2026.
More from AGI Musings
- François Chollet: believed futures deliver their consequences without ever happening — fchollet · 2026-10-19
- Paras Chopra reviews Sensing Qualia: a fresh "external" take on the hard problem — paraschopra · 2026-10-19
- "Where are the androids?" Reddit debates why AI automated art before housework — dev_ttangkong · 2026-10-19
- Robot inference cost on frontier models fell from $2.36 to 38 cents in 8 weeks — PeterDiamandis · 2026-10-19
- Diamandis releases new WTF episode on AI alignment and falling machine costs — PeterDiamandis · 2026-10-19
- Ethan Mollick: AI agents just made expert Stata coding skills largely obsolete — emollick · 2026-10-19