Chrome has decided that post-quantum HTTPS certificates will not arrive as bigger X.509 certificates in its existing root store. Instead, it is building a separate root programme and trust store for a different certificate format. That decision splits the post-quantum path for public web PKI from the path most enterprises will follow for their own internal PKI. This article sets out what the primary sources say, then what we think it means for organisations that run PKI and HSMs.
How to read this article. “Established facts” restates what named primary sources say, with numbered references. “SafeCipher analysis” is our own opinion. Everything here is based on documents that are drafts or work in progress, and may change. Sources were read on 16 November 2026.
Established facts
What Chrome announced
On 27 February 2026, Chrome’s Secure Web and Networking Team announced a new programme. It said Chrome has “no immediate plan” to add traditional X.509 certificates containing post-quantum cryptography to the Chrome Root Store. Instead Chrome is developing an evolution of HTTPS certificates based on Merkle Tree Certificates (MTCs), being worked on in the IETF PKI, Logs, And Tree Signatures (PLANTS) working group.1
In an MTC, a certification authority (CA) signs a single “tree head” covering potentially millions of certificates, and the certificate presented to the browser is a compact proof that it is included in that tree. Chrome says this decouples the strength of the algorithm from the size of data sent in the TLS handshake, and makes transparency a property of issuance rather than an add-on.1
Chrome’s stated rollout has three phases. Phase 1, a feasibility study with Cloudflare in which every MTC connection is backed by a traditional X.509 certificate, was underway at announcement. Phase 2, targeted at Q1 2027, invites operators of Chrome-usable Certificate Transparency logs to bootstrap public MTCs. Phase 3, targeted at Q3 2027, launches the Chrome Quantum-resistant Root Store (CQRS) and onboards further CAs under a separate Chrome Quantum-resistant Root Program. That programme will run alongside the existing Chrome Root Program.12
What the draft policy says (version 0.4.0)
The Chrome Quantum-resistant Root Program Policy exists as a draft. The version published is 0.4.0, labelled Release Candidate 1.0.0, dated 28 October 2026. Earlier drafts were 0.1.0 (14 May 2026), 0.2.0 (17 June) and 0.3.0 (14 August). Its change history says 0.4.0 aligns with draft-ietf-plants-merkle-tree-certs-07.3 Points relevant to PKI and HSM teams include the following.
- Scope. The policy applies to MTC CAs trusted by default in Chrome. It states that it does not apply to CAs that only issue to an enterprise organisation (private or locally trusted CAs), and that for private PKI Chrome supports X.509 certificates and trust anchors using ML-DSA in TLS 1.3 as of Chrome 150.32
- Eligibility. Initial eligibility is restricted to organisations that operated a “usable” Certificate Transparency log before 1 February 2026. A later update is expected to open applications to others, who would first operate a mirroring cosigner.3
- Automation and lifetime. MTC CAs must support ACME and ACME Renewal Information. Subscriber certificates are limited to 7 days or 47 days depending on key configuration, and reuse of domain control validation data is capped at 10 days.3
- Key protection. CA cosigner private keys must be generated, maintained and used within an HSM validated to FIPS 140-3 Level 3 or Common Criteria EAL 4+. The policy allows encrypted backups and wrapped-key transfers for disaster recovery or HSM migration, provided plaintext key material never exists outside a validated HSM boundary and operations are under multi-person control. From 1 January 2029 the HSM must hold a CMVP certification explicitly covering ML-DSA and ML-KEM, and a future update may require key attestation.3
- Structure. Each operator has one required active cosigner key, up to three optional further active keys (four at most), and two required offline reserve keys that need multi-person authorisation to activate. MTC CAs must not issue subordinate CA certificates of any kind.3
The standards behind it
The IETF draft is draft-ietf-plants-merkle-tree-certs, now at version 08, dated 2 November 2026 (the Chrome policy v0.4.0 aligns to version 07), an active working-group draft on the Standards Track. It describes an MTC as an X.509 certificate that integrates public logging with issuance. The draft’s own status note says Internet-Drafts may be updated or obsoleted at any time and should be cited only as work in progress.4 The draft gives the motivation in numbers: ML-DSA-44 uses 1,312 bytes per public key and 2,420 bytes per signature, and two signed certificate timestamps plus a leaf signature would add 7,260 bytes of authentication overhead even with a directly trusted intermediate.4
Timing
The PKI Consortium’s Post-Quantum Cryptography Conference runs from 1 to 3 December 2026 in Amsterdam and online. Attendance is free.5
Public and enterprise PKI side by side
The table below is our summary of the sources above. “Public” means MTC CAs that Chrome trusts by default under the draft policy. “Enterprise” means private or locally trusted CAs, which the policy says are out of its scope.
| Topic | Public (Chrome draft policy) | Enterprise (private PKI) |
|---|---|---|
| Certificate format | Merkle Tree Certificate: an inclusion proof plus cosignatures | X.509 with ML-DSA supported in TLS 1.3 from Chrome 150 2 |
| Trust store | Separate Chrome Quantum-resistant Root Store, proposed target Q3 2027 1 | Your own trust anchors, distributed by you |
| Hierarchy | Flat: no subordinate CAs 3 | Your design: offline roots and issuing CAs remain possible |
| Lifetime | 7 or 47 days maximum, by key configuration 3 | Your policy |
| Automation | ACME and ACME Renewal Information required 3 | Your policy |
| HSM | FIPS 140-3 Level 3 or CC EAL 4+; CMVP cover for ML-DSA and ML-KEM from 1 Jan 2029 3 | Your risk assessment |
| Status | Draft policy v0.4.0; IETF draft -08 34 | Depends on your clients and libraries |
SafeCipher analysis
The points below are our interpretation. They are not statements of what any standard or browser requires of you.
1. There is no longer one post-quantum certificate path. For publicly trusted TLS, Chrome’s route is a new format, a separate trust store, a flat structure with no subordinate CAs, and very short lifetimes. For private PKI, the route that Chrome documents is ordinary X.509 with ML-DSA. An organisation with both public-facing and internal certificates will therefore have two migrations to plan, not one.
2. Start by finding out which of your certificates are public and which are private. Many estates mix them. A certificate inventory, ideally part of a cryptographic bill of materials, is what tells you which systems face the public web, which depend on private hierarchies, and which are neither. Our cryptographic audit work starts here.
3. Automation is the common prerequisite. The draft policy treats ACME and renewal signalling as mandatory for MTC CAs, and short lifetimes make manual renewal impractical. Whichever path you are on, certificate lifecycle management that can renew without human steps is the shared foundation. See our CLM services.
4. The HSM requirements are informative even if you never become an MTC CA. The policy’s key-protection section shows what a browser root programme expects of HSMs: validated hardware, multi-person control, offline reserve keys, key-generation ceremonies evidenced in writing by a qualified auditor, and, from 1 January 2029, a CMVP certification explicitly covering ML-DSA and ML-KEM. Many enterprises will not be held to that standard, but it is a useful benchmark when you review your own HSM architecture and vendor roadmaps. Ask each vendor which post-quantum algorithms are available in which firmware, and in which validated status. Our vendor-neutral HSM comparison lists the questions.
5. Enterprise PKI is not “behind”; it is on a different clock. Private PKI can adopt ML-DSA when its own clients support it. That makes client and library readiness the real constraint, not the browser root programme. Test early: our notes on ML-DSA certificate chains and dual private keys cover some of what to expect. Our post-quantum architecture page describes how we plan this.
6. Treat dates as targets, not commitments. The policy is a draft, the specification is an Internet-Draft, and Chrome describes phase dates as targets. Plan for the direction (shorter lifetimes, automation, transparency by default), and avoid committing budget to a specific cutover date.
Questions to put to your own teams and suppliers
- Which of our certificates are publicly trusted, and which are private? Who owns each list?
- Which clients, libraries and appliances cannot yet handle larger post-quantum signatures, and how would we find out?
- Which HSMs do we use for CA keys, which post-quantum algorithms does each support, and in what validated status?
- If a CA key ever had to move to a new HSM, what controls would apply to the migration? (See our article on HSM migration and PKI trust.)
- What would we do if a public CA we depend on changed its roots or its certificate format?
What to do now
- Inventory certificates and classify them as public, private or mixed.
- Confirm that renewal for every public certificate can be automated.
- Review which of your HSMs support ML-DSA, in which firmware and in which validated mode.
- Pilot ML-DSA in a private hierarchy where clients can be controlled.
- Follow the PLANTS working group and the Chrome programme’s change history.
What this article does not claim
We do not claim that MTCs will replace X.509 on the public web, that other browsers will follow Chrome, or that any date will be met. Apple, Mozilla and Microsoft root programme positions are not covered here. Cloudflare has said it intends to operate an MTC CA and is targeting early 2027 for inclusion in Chrome’s Quantum-resistant Root Store.6 Chrome’s own FAQ and announcement give Q3 2027 as the target for launching the root store, and we have not reconciled the two statements. Treat both as vendor targets.
Next step
If you want help working out which of your certificates face the public web, which sit in private PKI, and what a realistic post-quantum plan looks like, start with a cryptographic audit, or read our overview of enterprise PKI architecture.
References
All sources accessed on 16 November 2026.
- Chrome Secure Web and Networking Team, “Cultivating a robust and efficient quantum-safe HTTPS”, Google Security Blog, 27 Feb 2026: blog.google/security/cultivating-a-robust-and-efficient-quantum-safe-https
- Chrome Quantum-resistant Root Program, Frequently Asked Questions: googlechrome.github.io/chromerootprogram/cqrp/faq
- Chrome Quantum-resistant Root Program Policy, Version 0.4.0 (draft, last updated 28 Oct 2026): googlechrome.github.io/chromerootprogram/cqrp/draft-policy
- IETF, “Merkle Tree Certificates”, draft-ietf-plants-merkle-tree-certs-08, 2 Nov 2026: datatracker.ietf.org/doc/draft-ietf-plants-merkle-tree-certs
- PKI Consortium, Post-Quantum Cryptography Conference, 1 to 3 December 2026, Amsterdam: pkic.org/events/2026/pqc-conference-amsterdam-nl
- Cloudflare, “Building a post-quantum certificate authority with Merkle Tree Certificates”: blog.cloudflare.com/pq-ca-with-mtcs

Leave a Reply