Skip to content
Back home
Tech watch

Post-quantum cryptography

In August 2024, NIST turned a decade of theoretical research into concrete federal standards: FIPS 203, 204, and 205. What used to be an academic conference topic became, overnight, a compliance obligation for anyone protecting long-lived data.

The stakes fit in one phrase: “harvest now, decrypt later.” Traffic encrypted today with classical algorithms (RSA, ECC) can be intercepted and stored as-is, waiting for a sufficiently powerful quantum computer to decrypt it retroactively. Migrating before that day isn’t optional: it’s a race against a threat with no confirmed deadline.

The algorithms

ML-KEM

Encapsulation

Establishes a shared secret between two parties, the post-quantum equivalent of Diffie-Hellman key exchange. Lattice-based, and the fastest of the three, which is why it’s already the one deployed in TLS and encrypted messaging.

FIPS 203

ML-DSA

Signature

A digital signature scheme derived from CRYSTALS-Dilithium, also lattice-based. Larger signatures than RSA (up to 4.6 KB), but the default candidate for document and certificate authentication.

FIPS 204

SLH-DSA

Hash-based

A hash-based signature scheme rather than a lattice-based one: a deliberately different mathematical assumption, kept as a safety net in case lattices ever turn out to be breakable. Slower and heavier (up to 50 KB), reserved for cases where caution outweighs performance.

FIPS 205

Timeline
2016

NIST launches its international post-quantum standardization competition.

2022

Kyber, Dilithium, SPHINCS+, and Falcon are selected as finalists.

Aug 2024

FIPS 203 (ML-KEM), FIPS 204 (ML-DSA), and FIPS 205 (SLH-DSA) are published as final standards.

Mar 2025

HQC is selected as a fifth algorithm: a second key-exchange mechanism based on error-correcting codes rather than lattices, so the standard doesn’t rest on one mathematical family alone.

2027

NSA deadline (CNSA 2.0): every new U.S. government system must support post-quantum cryptography.

2030

Transition deadline for existing systems.

2035

Target for full quantum resistance across all U.S. National Security Systems.

Real-world adoption

Chrome, Edge, and Firefox now enable a hybrid key exchange by default (X25519MLKEM768, combining the old and new standards); Signal and iMessage have already switched their protocols over. At Cloudflare, the share of TLS traffic encrypted with post-quantum key exchange went from roughly 2% in early 2024 to over 67% by April 2026, one of the fastest cryptographic migrations ever measured. The gap sits on the enterprise side: most are testing, few have shipped it to production.

2% → 67%

Cloudflare post-quantum TLS share, 2024 → 2026

87% / 7%

Enterprises testing vs. fully deployed

Sources