How to prepare your post-quantum roadmap: a 7-step method and a template
Most regulated organisations now accept that they will have to move away from RSA and elliptic-curve cryptography. Far fewer have a document that says how. The EU’s Coordinated Implementation Roadmap expects the transition to have started by the end of 2026 and high-risk use cases to be migrated by the end of 2030 (what each milestone asks for). This guide covers the work in between: turning a policy deadline into a plan your own teams can carry out.
A good post-quantum roadmap is not a list of RSA findings, and it is not a copy of the EU calendar. It answers five questions:
- What cryptography do we have, and where is it?
- Which of it is at risk, and by when must it be fixed?
- What do we do first, next and last?
- Who owns each part, and who pays for it?
- How do we show, every quarter, that our exposure is going down?
The seven steps below produce those answers in order. Each step ends with a deliverable that goes into the roadmap document described at the end of this guide.
Step 1: Get a mandate and name an owner
A post-quantum migration touches applications, infrastructure, PKI, suppliers and budgets over almost a decade. Without an executive sponsor, it stalls at the first contract that has to be renegotiated.
- Sponsor: a member of the executive committee, usually the CIO or COO, who approves the budget and settles disputes between business lines.
- Owner: the CISO’s office, which writes and maintains the roadmap.
- Core team: enterprise architecture, PKI and key management, application development, network, procurement and legal, plus a business owner for each critical service.
Anchor the mandate in obligations the board already recognises. NIS2 Article 21(2)(h) requires policies on the use of cryptography. Under DORA, Article 6 of the ICT risk management RTS (Delegated Regulation (EU) 2024/1774) requires the encryption policy to provide for updating cryptographic technology as cryptanalysis develops. The EU roadmap supplies the dates against which those policies will be read.
Step 2: Scope from business services, not systems
Starting from a list of servers produces an inventory nobody can prioritise. Start instead from the services the organisation must keep running and protected: the critical or important functions you already report under DORA, or the services in scope of NIS2. Then list the cryptographic use cases behind each one.
| Use case | Typical cryptography | Why it matters |
|---|---|---|
| External TLS (websites, APIs) | ECDHE key exchange, RSA or ECDSA certificates | Traffic can be recorded today and decrypted later |
| Internal TLS, mTLS, service mesh | The same, often with private CAs | High volume and rarely inventoried |
| VPNs and site-to-site links | IKEv2 with (EC)DH, RSA or ECDSA authentication | Long-lived links carrying sensitive data |
| SSH and administration | RSA or ECDSA keys, ECDH | Privileged access paths |
| PKI, code and firmware signing | RSA or ECDSA signatures, long-lived roots | Hardest to change; signatures must outlive the device |
| Data at rest and archives | Symmetric encryption, keys wrapped with RSA | Long confidentiality requirements |
| Email and document signing | S/MIME, CMS, RSA or ECDSA | Legal value that lasts for years |
| HSM, KMS and cloud key management | Depends on the vendor | Limited by the supplier’s own roadmap |
| OT and embedded devices | Fixed in firmware | Lifecycles of 15 years or more |
Step 3: Build an inventory you can trust
The roadmap is only as good as the inventory beneath it. For each use case, find where the cryptography really is: in source code (library calls, algorithms, key sizes), in certificates, in live TLS and SSH configurations, in keystores and cloud KMS, and in the products you buy. Three rules keep the inventory usable:
- Scan first, then ask. Questionnaires find what people remember; scans find what is deployed. Keep questionnaires for suppliers and for systems you cannot scan.
- Keep the provenance. Every entry should point to a file and line, a certificate or an endpoint, so an engineer can act on it and an auditor can check it.
- Make it repeatable. A one-off spreadsheet is out of date within a quarter. Export the result as a CBOM in a standard format such as CycloneDX 1.6, so one scan can be compared with the next.
Our guides on what a CBOM looks like and how to estimate the blast radius of a migration cover this step in detail.
Step 4: Classify risk by use case
Not everything has to move by 2030. The EU roadmap separates high-, medium- and low-risk use cases, and your own classification decides which group each one falls into. Three questions drive it:
- How long must the data stay confidential? If the answer is ten years or more, traffic recorded today is already exposed. This is the harvest-now-decrypt-later risk.
- How long will the migration take? Large estates, embedded systems and supplier dependencies can take years.
- How long must a signature be trusted? Root certificates, firmware signatures and signed archives must remain verifiable for their whole life, so they need a post-quantum path well before the device or document is retired.
A practical rule set, aligned with the EU milestones:
| Class | Typical profile | EU target |
|---|---|---|
| High | Data that must stay confidential for 10+ years, long migrations, long-lived roots of trust, internet-facing key exchange | Migrated by the end of 2030 |
| Medium | Internal services, shorter-lived data, moderate migration effort | Migrated by the end of 2035 |
| Low | Short-lived data, little exposure, systems due for retirement | As far as feasible by 2035, or retired |
Write the criteria down. A supervisor will want to know why a use case is classed as medium rather than high, and written criteria are what make the answers consistent.
Step 5: Define the target state
The roadmap has to say what “migrated” means. For most EU organisations, the answer follows the national agencies:
- Hybrid during the transition. ANSSI and BSI both recommend combining a classical algorithm with a post-quantum one rather than relying on post-quantum alone. For TLS 1.3, the hybrid group X25519MLKEM768 is standardised in RFC 10024.
- The NIST standards. ML-KEM (FIPS 203) for key establishment, ML-DSA (FIPS 204) for signatures, and SLH-DSA (FIPS 205) where a conservative hash-based signature suits long-lived roots or firmware.
- Nothing built on drafts. FN-DSA (the future FIPS 206) and HQC are not yet published as standards. Keep the algorithm choice configurable rather than waiting for them.
- Crypto-agility as a requirement. Algorithms should be set in configuration or a shared library, not hard-coded in each application, so the next change costs less than this one.
Record these choices in the cryptography policy itself, so the policy and the roadmap say the same thing. See crypto-agility explained and our comparison of BSI, ENISA, NCSC and ANSSI guidance.
Step 6: Sequence the work in dated waves
Order the work by risk class, then by effort and dependencies: a use case that relies on a PKI change cannot move before the PKI does. Tie each wave to the external milestones, so the plan can be read against them.
| Period | Focus | Example deliverables |
|---|---|---|
| Now to end-2026 | Evidence and pilots | Scoped inventory, risk register, signed roadmap, one or two pilots on high-risk use cases, such as hybrid TLS on an external service |
| 2027–2028 | High risk, quick to move | Hybrid key exchange on external TLS and VPNs, post-quantum requirements in every new purchase, a PKI and HSM upgrade plan |
| 2029–2030 | High risk, hard to move | Remaining high-risk use cases migrated, post-quantum or hybrid signatures for code signing and new roots |
| 2031–2035 | Medium and low risk | Internal services, OT and embedded systems at their refresh, retirement of legacy cryptography |
Then check the plan against the national dates that apply to you:
- France: ANSSI has announced that post-quantum protection will be required for product qualification from 2027, and that buying products without it after 2030 will no longer be reasonable.
- Germany: BSI TR-02102-1 recommends classical-only key agreement only until the end of 2031, and classical signatures until the end of 2035.
- Financial groups: the G7 Cyber Expert Group suggests moving the most important systems around 2030–2032, with 2035 as the overall target.
- UK operations: the NCSC expects discovery and an initial plan by 2028, priority migrations by 2031 and completion by 2035.
Use natural refresh cycles. Hardware renewals, certificate rotations and contract renewals are the cheapest moments to migrate, so map them onto the waves instead of funding a separate project for each change.
Step 7: Bring in suppliers and the budget
Much of your cryptography sits in products you buy: HSMs, VPN appliances, load balancers, PKI software, SaaS and OT equipment. Their roadmaps set the limits of yours. Ask each critical supplier, in writing:
- Which post-quantum algorithms and hybrid modes do you support today, and from which version?
- When will you support hybrid TLS key exchange (RFC 10024) and ML-DSA certificates (RFC 9881)?
- Can algorithms be changed by configuration, without a new release?
- Can you provide a CBOM, or at least a list of the cryptography in the product?
Add the same points to contract templates and renewals, so that every new purchase reduces your exposure instead of extending it.
For the budget, estimate effort wave by wave rather than for the whole programme at once. The inventory gives the number of affected components, and the classification and blast-radius analysis give the effort for each. Keep the one-off project costs (inventory, pilots, PKI changes) separate from running costs (continuous discovery, supplier follow-up, policy reviews).
Measure progress every quarter
A roadmap that is never measured turns into a slide. Choose a few indicators that come straight from the inventory, so they can be recomputed after each scan:
- Inventory coverage: the share of in-scope services scanned in the last quarter.
- High-risk use cases with an owner, a target date and a funded plan.
- External TLS endpoints that negotiate hybrid key exchange.
- Critical suppliers with a written post-quantum roadmap.
- Quantum-vulnerable findings in high-risk use cases, trending down from one scan to the next.
- Open risk acceptances, each with an expiry date.
Review the roadmap with the steering body every quarter and reissue it once a year, or sooner if a standard, a national guideline or a supplier date changes. Our PQC standards tracker lists the sources to watch.
The roadmap document: a template
Bring the seven deliverables together in one document. This structure works for a board and for a supervisor:
- Executive summary: today’s exposure, the targets, the budget and the decisions needed.
- Mandate and governance: sponsor, owner, steering body, review cadence.
- Scope: services and cryptographic use cases, with their business owners.
- Inventory summary: coverage, method, known gaps, and a link to the CBOM.
- Risk classification: the criteria, results by class, and the reasoning for borderline cases.
- Target state: cryptography policy, hybrid choices, crypto-agility requirements.
- Migration waves: dates, dependencies, owners and pilots.
- Suppliers: status and committed dates for critical products.
- Budget and resources: per wave, split between project and running costs.
- Indicators and reporting: KPIs, baseline and targets.
- Exceptions: risk acceptances, each with an owner and an expiry date.
Keep the document short and link to the detail. The inventory and the risk register will change every quarter; the roadmap should not need rewriting each time they do.
Five mistakes to avoid
- Starting with the algorithm. Choosing ML-KEM is the easy part. Knowing where it has to go is the real work.
- An inventory built only from questionnaires. It misses vendored libraries, internal TLS and forgotten certificates.
- A one-off inventory. Code and certificates change every week, so last year’s snapshot cannot support this year’s plan.
- Treating 2035 as the deadline for everything. High-risk use cases are due by 2030, and data recorded today is already exposed.
- Leaving suppliers until last. Their timelines decide yours, and contracts take months to change.
A post-quantum roadmap is a management document built on technical evidence. Without the evidence it is a wish list; without the management, it is a scan report.
How CRYPTAGION fits
CRYPTAGION produces the inputs for steps 3, 4 and 6, and the indicators that follow them. It discovers cryptography across code, Git repositories, certificates, TLS endpoints, keystores and cloud KMS on a schedule, scores each finding from 0 to 100 including its harvest-now-decrypt-later exposure, and exports a CycloneDX 1.6 CBOM, a four-wave migration roadmap and a CISO report. It runs on-premises or air-gapped. The fixed-scope readiness pilot covers one critical domain and produces the end-2026 evidence for it.
The takeaway
Get a mandate, scope from services, inventory what is really deployed, classify by data lifetime and migration time, fix the target state, sequence the work in waves tied to 2026, 2030 and 2035, and bring suppliers in early. With two months left in 2026, the first three steps and a signed roadmap are a realistic goal for this year. The CISO readiness checklist is a quick way to see where you stand before you start.
Sources
- NIS Cooperation Group — A Coordinated Implementation Roadmap for the Transition to Post-Quantum Cryptography (June 2025)
- NIS Cooperation Group — EU roadmap on PQC: survey feedback and FAQ (September 2026)
- Directive (EU) 2022/2555 (NIS2), Article 21
- Commission Delegated Regulation (EU) 2024/1774 (DORA ICT risk management RTS), Article 6
- NIST FIPS 203 (ML-KEM), FIPS 204 (ML-DSA) and FIPS 205 (SLH-DSA)
- RFC 10024 — Hybrid ECDHE-MLKEM key agreement for TLS 1.3
- RFC 9881 — ML-DSA in X.509
- ANSSI — Post-quantum cryptography
- ANSSI — Follow-up position paper on post-quantum cryptography
- BSI — TR-02102-1, Cryptographic Mechanisms: Recommendations and Key Lengths (version 2026-01)
- G7 Cyber Expert Group — Roadmap for the transition to post-quantum cryptography in the financial sector (January 2026)
- G7 Cybersecurity Working Group — Preparing for the Post-Quantum Era: A Call to Action (September 2026)
- NCSC — Timelines for migration to post-quantum cryptography
Start with the inventory
Every step of the roadmap depends on knowing where your cryptography is. CRYPTAGION builds that inventory inside your perimeter, with the risk scores and CBOM the roadmap needs.
Book a live scan → Or take the 2-minute readiness assessment