Computer Science · Cryptography
Cryptography & The Enigma
From ancient Rome to the Allied codebreakers at Bletchley Park, encryption has shaped history. Step through three eras of cryptography: shift ciphers, the Enigma machine, and modern RSA public-key encryption.
Three Eras
- Caesar Cipher — shift every letter by a fixed amount
- Enigma Machine — polyalphabetic substitution with rotors
- RSA — public-key encryption using prime factorization
- Compare what the key controls and what an interceptor can learn
One key. One fixed shift.
Try it: Encrypt MEET AT NOON with shift 3. Which repeated letters stay repeated? To recover the original, paste the ciphertext into the input and use shift 23.
What would an interceptor need?
Compare the three experiments: a Caesar shift, Enigma’s starting settings, and an RSA private key. Explain why a complicated-looking message is not proof of security.
Check your explanation
Caesar has very few keys to try. Enigma hides repeated-letter patterns, but settings and procedures matter. RSA allows a public encryption key, yet only large, correctly used keys offer practical protection.
Caesar’s shift cipher
A fixed shift replaces each letter with another. With only 25 nonzero shifts, an interceptor can try every key; repeated letters also preserve patterns.
The Enigma Machine (1930s)
Rotating wheels changed the substitution with each keypress. Rotor order, starting positions, ring settings, and plugboard connections were parts of the key.
Bletchley Park (1939–45)
Polish mathematicians first broke military Enigma. Building on their work, Allied teams used likely plaintext, operating mistakes, and machines such as the British Bombe to test possible settings.
RSA (1977)
Rivest, Shamir, and Adleman introduced RSA, a public-key cryptosystem. Its security depends on large keys and careful protocols. This tiny, unpadded classroom model is not secure encryption.
AI Codebreaking (2026)
In September 2026, developer Carter Leffen set OpenAI's GPT-6 Astra loose on MVUEH, an 82-character German Army Enigma message from July 10, 1941, that had sat unbroken in public archives since 2005. Over roughly 10 hours, the AI searched archival scans, wrote its own Enigma simulator, and tested millions of settings. The break came from a crib: "Rosenow," a place name repeated in a related message solved the same day. The recovered plaintext asks for a march route and reports the sender's position — "Bitte um Angabe des Marschweges. Befinde mich in Rosenow, Rosenow. Sofort Funkantwort." Veteran researcher Frode Weierud verified the solution and logged it on his CryptoCellar archive, calling it the first hard Enigma problem broken with AI. It's the same idea as the Bombe to its left — mechanically testing candidate settings against a guessed piece of plaintext — just running on a language model instead of rotating drums.
Primary Sources
What the machines actually looked like
Every photo below is a real object or person from this history, not an illustration. Click any photo to enlarge it.
Keep Exploring
Related STEAM Labs
Ciphers, codebreakers, and the math behind them show up across the CS and math lessons.
Explore the evidence: Crypto Museum: how Enigma works · Science Museum: the codebreaking teams.