Skip to content

Understanding post-quantum cryptography

Why security researchers are preparing for quantum attacks before the computers needed to carry them out are available.

planted June 19, 2026 · last tended November 9, 2026

A captured spark. Unverified, unpolished, possibly wrong.


The part of post-quantum cryptography that caught my attention is a timing problem. Someone could collect encrypted information now and keep it in the hope that a future computer will make it readable.

That is the concern behind “harvest now, decrypt later”. As NIST explains, it matters especially for information that needs to remain confidential for many years.

What would change?

RSA is one familiar public-key system. Its security depends in part on how difficult it is to factor a large number into its prime factors. Shor’s algorithm provides a way for a sufficiently capable quantum computer to solve that problem efficiently.

The qualifications matter: the threat requires a large, reliable quantum computer, and the timing remains uncertain. Quantum computing also doesn’t mean that every kind of encryption becomes useless at once.

I find it easier to understand the concern by focusing on which mathematical assumptions a system depends on, rather than picturing quantum computers as machines that can instantly solve any problem.

What the replacement looks like

Post-quantum cryptography uses algorithms designed to resist both classical and quantum attacks. They run on ordinary computers; using them doesn’t require a quantum computer of your own.

NIST published its first three final standards in August 2024:

  • ML-KEM, derived from CRYSTALS-Kyber, lets two parties establish a shared secret.
  • ML-DSA, derived from CRYSTALS-Dilithium, provides digital signatures.
  • SLH-DSA, derived from SPHINCS+, provides signatures using a different, hash-based approach.

The first two use lattice-based mathematics. Having different approaches is useful because confidence in a cryptographic scheme can change as researchers study it. These are security assumptions supported by analysis, rather than promises that no future attack is possible. NIST’s standards announcement

Why prepare ahead of time?

Changing cryptography across software, devices, and services takes time. If the information needs to stay private for longer than the current protection is expected to hold, waiting for a demonstrated attack may leave too little time to respond.

That’s what connects this to thinking about an attacker’s options. I need to consider how long the information matters, as well as what an attacker can do with it today.

Paths that lead here

No paths yet. This note is still off the beaten track.

Where this note points

  • Thinking like an attacker · I'm interested in the gap between how a system is supposed to work and what its parts allow it to do.

More from these beds