Post-quantum for the rest of us
The post-quantum transition is half finished, and the finished half happened without you noticing. The other half is the part your certificates are in, and nobody has shipped it yet.
Two problems, not one
A quantum computer big enough to matter would break the mathematics behind both halves of TLS: the key exchange that protects a session, and the signatures that prove who you are talking to. The two halves are on very different schedules.
The finished half
NIST published the three standards in August 2024: ML-KEM for key exchange, ML-DSA for signatures, and SLH-DSA, a hash-based signature scheme.1 Chrome had already enabled a hybrid post-quantum key exchange by default on desktop in Chrome 124 in April 2024, and switched it to the standardized ML-KEM in Chrome 131 that November.2,3 When Apple turned it on by default in iOS 26 in September 2025, the share of post-quantum-protected traffic through Cloudflare jumped, and it reached 52 percent by early December 2025, up from 29 percent at the start of the year.4 None of this required changing a single certificate. Key exchange is negotiated per connection, so a server upgrade and a client upgrade were enough.
The unfinished half
Certificates are harder for a boring reason: size. A typical TLS handshake carries around five signatures and two public keys, and ML-DSA keys and signatures are roughly forty times the size of their elliptic-curve equivalents.2,5 Cloudflare estimates that switching every signature in a handshake to ML-DSA would add about fifteen kilobytes per connection.5 So, in Cloudflare's words from October 2025, "a lot of Internet traffic is protected by post-quantum key agreement, but not a single public post-quantum certificate is used", and while the first trials began in 2026, broad browser trust is unlikely before 2027.5,7 The most promising fix, Merkle Tree Certificates, replaces most of the signatures with a short proof; Chrome and Cloudflare are testing it, and Chrome plans a new root store for the post-quantum web starting in the third quarter of 2027.5
What to do now
- Inventory your algorithms. Every key you own has a type and a size, and you will need to know which are RSA, which are elliptic-curve, which are inside hardware, and which sign things that must stay verifiable for a decade. That list is the migration plan's input, and almost nobody has it.
- Turn on hybrid key exchange at your edges. Modern TLS stacks support it, browsers and phones already ask for it, and it costs nothing in certificates. This is the store-now-decrypt-later defense, and it is available today.
- Keep your certificates short and your renewal automatic. The 47-day schedule is doing you a favor here: an estate that renews everything every six weeks can switch signature algorithms the same way, in six weeks, when the time comes.
- Decide what your private roots should be. Inside your own walls you are not waiting for the browsers. A private root that must last ten years is a candidate for a hash-based signature, whose security rests only on the hash function, at the cost of much larger signatures.6
What not to do
Do not rip and replace. Do not issue post-quantum certificates into anything a browser needs to trust, because no root program trusts them for public websites yet.5 And do not buy a "quantum-safe" anything before you have the inventory in step one, because without it you cannot tell what the purchase would change.
trstctl's post-quantum support lives behind its single isolated cryptography path: ML-DSA certificates issued through standard enrollment and verified by a stock OpenSSL client, ML-KEM checked against a subset of NIST's test vectors, and SLH-DSA keys held in the isolated signer. Its own listeners negotiate classical key exchange today; adding the hybrid group there is next. Its crypto-agility inventory and migration campaigns exist for step one and step three.6
The lifecycle and post-quantum page covers crypto-agility and migration orchestration, and the limitations page lists which algorithms are available today.
Sources
- NIST, Announcing Approval of Three Federal Information Processing Standards for Post-Quantum Cryptography, 13 August 2024: FIPS 203, 204 and 205.
- Google, Advancing Our Amazing Bet on Asymmetric Cryptography, 23 May 2024: hybrid Kyber enabled by default on desktop as of Chrome 124; ML-DSA keys and signatures about 40 times the size of ECDSA.
- Google Security Blog, A new path for Kyber on the web, 13 September 2024: the switch to standardized ML-KEM in Chrome 131.
- Cloudflare, The 2025 Cloudflare Radar Year in Review: post-quantum encrypted traffic from 29% to 52% during 2025, driven by iOS 26.
- Cloudflare, State of the post-quantum Internet in 2025, 28 October 2025.
- trstctl, Crypto agility and post-quantum; documentation, Limitations, "Post-quantum cryptography (issuance algorithms)"; Lifecycle and PQC: crypto-agility (F16) and PQC migration orchestration (F57).
- Google, Cultivating a robust and efficient quantum-safe HTTPS, 27 February 2026: a Merkle Tree Certificate feasibility study with Cloudflare; a new root store for the post-quantum web from the third quarter of 2027.