Fortifying NIST-PQC Against An Attacker Smarter Than Expected

The nightmare of every cryptograher is an attacker using. surprise mathmatical insight to crack the code. The NIST-PQC cryptographer is not exempt. A NIST-PQC collapse would be so catasrophic that extra protection is warranted, and can be applied through "Decoy Tolerant Cryptography".

It works as follows: The protected message M₀ is NIST-PQC encrypted to ciphertext C₀ using key K₀. An AI engine generates n 'fake messages' that are contextually as plausible as M₀. (M₁, M₂, ..... M_n), encrypting each with the same cipher using keys K₁, K₂, ..... K_n respectively to generate n ciphertexts: C₁, C₂, ..... C_n. The resultant (n+1) ciphertexts are mixed into a composite ciphertext CC, and sent to the recipient. The recipient is in possession of K₀, it treates C₁, C₂, ..... C_n as noise, and decrypts C₀ to M₀. The class of ciphers that enables this act of "unmixing" is called "Decoy Tolerant Ciphers". It is a characteristic of "Pattern Devoid Cryptography" that they are decoy tolerant.

Hopefully the NIST-PQC used will do its job, but in the event that the attacker is smarter than expected then they will be able to extract t ≤ n messages from CC, and will not be able to point to M₀, not even be sure that M₀ is among the t message candidates they extracted.

Since the history of cryptography is replete with cases of cryptanalysts that are smarter than expected, and since life in cyber space depends on cryptographic integrity, it is hard to find a good reason not to arm ourselves this straight forward protection offered by decoy tolerant ciphers.