Skip to content

PQC does not replace QRNG, and QRNG does not make you quantum-safe

Two errors, in opposite directions. Post-quantum algorithms do not remove the need for a good entropy source, and a quantum random number generator does not make a system quantum-safe. They address different problems and neither substitutes for the other.

Two errors pull in opposite directions, and the second one is the more flattering of the two to get wrong.

Error one · PQC replaces the need for good randomness

It does not. ML-KEM samples a fresh secret per encapsulation; ML-DSA in hedged mode consumes randomness per signature. Every one of those draws needs to be unpredictable, and a post-quantum algorithm seeded from a weak source produces weak keys with a modern name.

The migration actually increases entropy demand, because hybrid modes do two key exchanges and re-keying an estate generates keys in bulk. The dependency in detail.

Error two · a QRNG makes you quantum-safe

It does not, and this is the one our own category gets wrong most often.

A QRNG improves the quality of the randomness feeding your cryptography. It does nothing whatsoever about the fact that RSA and elliptic-curve algorithms are breakable by a quantum computer. A system using excellent quantum randomness with RSA-2048 is exactly as exposed to Shor's algorithm as one using any other source.

Being quantum-safe means the algorithms resist quantum attack. That requires algorithm migration and nothing else provides it. Any vendor implying their entropy source makes you quantum-safe is either confused or hoping you are.

What each one actually does

  • Post-quantum algorithms — protect against an adversary with a large quantum computer. Necessary. Not sufficient.
  • A strong, evidenced entropy source — protects against an adversary who can predict your key material. Necessary. Not sufficient.
  • Both — address two independent threats. Neither is a substitute and neither is optional if both threats apply to you.

Why the confusion persists

Both have "quantum" in the name and both are sold into the same anxiety. One uses quantum mechanics to generate randomness; the other defends against quantum computation. The word appears on both sides of a relationship it does not describe.

A useful test for any vendor claim: ask which of the two threats it addresses. If the answer is both, ask how — and then ask the same question about QKD, which is the third technology routinely folded into the same sentence.