The one figure that cannot be broken
The disposable mask
Since the weakness of the Vigenère cipher stems from the repetition of the key, let’s take this idea to its logical conclusion: what if the key were as long as the message, and were never reused?
This is the one-time pad.
The three conditions
The encryption is identical to Vigenère, but the key must adhere to three absolute rules:
- it is as long as the message;
- it is perfectly random;
- it is never reused.
Why it is unbreakable
Claude Shannon demonstrated this in 1949: this system offers perfect secrecy. It is not a question of computing power; it is an impossibility in principle.
The argument boils down to a single observation. Let’s take a five-letter ciphertext. There is a key that decrypts it to ATTAQ, another that decrypts it to PAIX!, and yet another for any five-letter text.
An attacker, even with infinite computing power, would obtain all possible messages. The ciphertext therefore contains no information about the plaintext. There is nothing to break.
Why it is almost never used
The problem is practical, and it is a deal-breaker:
- to exchange 1 GB of data, you first need to exchange 1 GB of key — securely;
- if you know how to exchange 1 GB securely, why not exchange the message directly?
- the key must be truly random; a pseudo-random number generator undermines the security guarantee;
- reusing a key even once brings the whole system down.
The disposable key was used for the Moscow–Washington ‘red telephone’ and for certain espionage networks, where key books could be physically transported.
Key takeaway: the only perfect cipher exists; this has been proven; yet it is unusable in practice. All modern cryptography involves accepting security that is purely computational, but manageable.

