No-Regret First Moves: Hybrid TLS And Cleaning Up Cryptographic Antipatterns
Some post-quantum work can start now at low cost. This article explains hybrid key exchange in TLS 1.3, where it is already available and which cleanup tasks pay off regardless.
Checked against primary sources and independently reviewed on . Sources are listed at the end.
Most of a post-quantum programme is planning and coordination that will run for years. A few tasks, though, can start almost immediately, cost little and reduce risk whatever happens next. Practitioners call these no-regret moves.
This article covers the two most widely recommended: turning on hybrid post-quantum key exchange for web traffic, and fixing the cryptographic habits that make every later change harder. It also covers where official positions on hybrid cryptography differ.
How Hybrid Key Exchange Works In TLS 1.3
In a hybrid key exchange, the two sides run a classical key agreement and a post-quantum one at the same time and combine the results. An attacker would have to break both to recover the session keys. OMB’s memorandum M-26-15 outlines how this works in TLS 1.3.1
- ClientHello
The client opens the handshake offering a hybrid option, with a public value for a classical method such as X25519 and one for a post-quantum key encapsulation mechanism such as ML-KEM-768.
- ServerHello
A server that accepts the hybrid option replies with its own X25519 value and an ML-KEM ciphertext, which wraps a fresh secret using the client public key.
- Two Shared Secrets
Client and server now each hold the same pair of secrets: one from X25519 and one from ML-KEM.
- Combined Key Schedule
Both secrets go into the TLS 1.3 key derivation together, so the session keys depend on both of them.
The post-quantum half is not a mirror of the classical one. ML-KEM, defined in FIPS 203, is a key encapsulation mechanism: the server does not send a public value of its own but uses the client’s public key to wrap a secret, which only the client can unwrap.2
Because the recorded handshake can only be decrypted by breaking both components, hybrid key exchange protects traffic against harvest now, decrypt later attacks while keeping a classical safety net in case a flaw is found in the newer algorithm. OMB also notes that TLS 1.3’s redesigned handshake is well suited to hybrid exchange, and requires US federal agencies to support TLS 1.3 no later than 2 January 2030.1
Where It Is Already Available
For public websites, the hybrid scheme X25519MLKEM768 is the obvious first candidate. Europol’s financial services report, published in 2026, says it is already the default in the latest versions of most major cryptographic libraries, that most web browsers have supported it since November 2024, and that the largest content delivery networks support it or have short-term plans to.3 The report also cites F5 Labs research, published in June 2025, which found hybrid post-quantum key exchange on five of the world’s ten most visited websites at that time.4 Europol scores public websites as low migration time and calls them the earliest practical opportunity to run post-quantum protection in production. It also sees a side benefit: an early, visible deployment helps technical teams, decision-makers and customers understand that preparing now is feasible.3
SSH has moved in the same direction. OpenSSH 10.0, released on 9 April 2025, made the hybrid algorithm mlkem768x25519-sha256 its default for key agreement.5
CISA’s product category list, published on 23 January 2026, shows how far support has spread. It lists cloud platforms (infrastructure and platform as a service), chat and messaging software, web browsers and servers, and full disk encryption as categories where post-quantum capable products are widely available, and says organisations, federal civilian agencies included, should buy only post-quantum capable products in those categories.6 CISA notes that most of these products protect key establishment with post-quantum algorithms but not yet signatures.6 Networking equipment, software as a service (SaaS), operating systems, storage, identity and access management, and hardware security modules are listed as still transitioning.6
Official Positions On Hybrid Differ
Agencies do not agree on how much weight to give hybrid cryptography. OMB says it can be useful during migration but warns that it is “an intricate and resource-intensive stopgap” and asks agencies to evaluate the trade-offs carefully.1 The joint statement from France, Germany, the Netherlands and other EU member states, by contrast, says deploying hybrid post-quantum solutions is becoming urgent.7
In our view the difference matters less than it first appears for most organisations. On the web, hybrid key exchange is usually an option that suppliers have already built and tested, so much of the effort OMB warns about falls on systems an organisation builds itself. Where you do build your own cryptography, the trade-off deserves a proper assessment.
Clean Up The Antipatterns
An antipattern is a common practice that looks convenient but creates risk and technical debt. Europol’s report argues that removing cryptographic antipatterns early improves security now and speeds later migration, even for systems that are not ready for post-quantum algorithms.3 For public websites it lists five.
| Antipattern | Remediation | Why it helps migration |
|---|---|---|
| Manual TLS certificate management | Automate issuance, renewal, rotation and revocation | Certificate changes become routine rather than projects |
| Inconsistent TLS configurations | Set a standard baseline, manage it automatically, target full TLS 1.3 | One change can be applied everywhere |
| Keeping TLS 1.0, 1.1 or weak ciphers for compatibility | Monitor negotiated versions and cipher suites, then retire legacy options | Reveals hidden dependencies before they block migration |
| Pinning to certificates you do not control | Use modern certificate management and monitoring, such as Certificate Transparency and CAA records | Reduces outage risk when certificates or algorithms change |
| Wildcard certificates | Use hostname-specific certificates | Limits the damage if one key is compromised |
The same report recommends that cryptography leaders catalogue known antipatterns with architects and operations teams, write them into security policy and development guidelines, and track remediation with clear metrics.3 Hard-coded keys and parameters in code, covered in the crypto-agility article, belong on the same list.
Footnotes
-
Office of Management and Budget, “Execution of the Migration to Post-Quantum Cryptography” (M-26-15), 24 June 2026. whitehouse.gov ↩ ↩2 ↩3
-
NIST, FIPS 203, “Module-Lattice-Based Key-Encapsulation Mechanism Standard”, August 2024. nvlpubs.nist.gov ↩
-
Europol, “Prioritising Post-Quantum Cryptography Migration Activities in Financial Services”, Publications Office of the European Union, 2026. Contributing authors: Oscar Covers, Thibaud Ecarot, Rebecca Gibergues, Jaime Gómez García, Francis Gorman, Imran Khan, Ivan Makarov, Sarah McCarthy, Michele Mosca, Iván Soto, Leila Taghizadeh and Jelena Zelenovic. Licensed under CC BY 4.0. The antipattern table is adapted from it, with changes. europol.europa.eu ↩ ↩2 ↩3 ↩4 ↩5
-
F5 Labs (David Warburton), “The 2025 State of Post-Quantum Cryptography (PQC) on the Web”, 26 June 2025. f5.com ↩
-
OpenSSH, “OpenSSH 10.0 release notes”, 9 April 2025. openssh.org ↩
-
CISA, “Product Categories for Technologies That Use Post-Quantum Cryptography Standards”, 23 January 2026. cisa.gov ↩ ↩2 ↩3
-
ANSSI, BSI, the Netherlands and 15 other EU member states, “Securing Tomorrow, Today: Transitioning to Post-Quantum Cryptography”, 30 November 2024. cyber.gouv.fr ↩
Knowledge Hub content is general information. It is not legal advice, a compliance certification, a guarantee of security or a substitute for an assessment of your own systems. Standards and rules change; check the sources for the latest position.