Nobody is going to finish a post-quantum cryptography migration in the next seven weeks.
That is fine. The first EU milestone is not asking for a migration.
It is asking whether you know where your cryptography lives, what depends on it, and who you would have to call to change it. Most security teams cannot answer that today. You can fix that before the year ends.
What does the EU PQC roadmap expect by 31 December 2026, and who is it actually addressed to?
The document is the first release of the Coordinated Implementation Roadmap from the NIS Cooperation Group work stream on PQC. Its own scope section is clear about the audience. It is "a first high-level concept paper on the transition to PQC aimed at the Member States." It adds that "with some adaptions this document will also be useful for government organisations and other entities including the ones that have to comply to the NIS2 Directive."
So the roadmap is a recommendation to governments. It is not a regulation that binds your company directly.
The date is real, though. The roadmap's timeline box reads "Milestone 1: 31.12.2026." The text behind it: "Member States should ensure the measures listed in First Steps are accomplished before the end of 2026."
From today, that is 48 days.
Two later dates frame the work. The second milestone is 31.12.2030, when "the PQC transition for high-risk use cases has been completed." Then 31.12.2035, for medium-risk use cases.
The roadmap is also explicit that the clock is not sequential. Member States should "not wait until the end of 2026 to start implementing the Next Steps," and "waiting for the completion of Milestone 1 cannot be afforded."
If you are one of the NIS2 entities, expect your national authority to start asking the questions the First Steps describe. Four of the eight First Steps translate directly into work your team can own: cryptographic asset management, dependency maps, quantum risk analysis, and the supply chain. The rest of this piece turns those four into tickets.
Why is a cryptographic inventory the "no-regret" first move for every entity?
Because it pays off even if quantum computers never arrive on schedule.
The roadmap puts it this way: "A first essential step and 'no-regret' move for every entity is to create and maintain current inventories of assets that perform cryptographic operations and of assets that have cryptographic operations performed on them as part of asset management."
It then names the payoff outside PQC. These inventories "support incident and vulnerability management as part of the general IT controls, not only related to the quantum threat." And they apply "both on premise and in the cloud."
That is your budget argument. You are closing an asset management gap that already hurts every time a library vulnerability lands and nobody knows which services use it.
There are two asset types:
→ Assets that perform cryptographic operations: TLS terminators, HSMs, VPN concentrators, code-signing services, key management systems, libraries inside your applications.
→ Assets that have cryptographic operations performed on them: the databases, backups, documents, firmware images and message streams that get encrypted or signed.
Most first-pass inventories capture only the first list. The second list is where "store now, decrypt later" risk actually sits.
What should each inventory record capture, and why use a CBOM format?
The roadmap defines a cryptographic inventory as "a structured overview of cryptographic assets." On format, it is direct: "Using a standardised format for a cryptographic inventory, like CBOM (Cryptographic Bill of Materials, an extension of the SBOM standard), is recommended."
The roadmap does not prescribe fields. Here is the record we would put on the ticket, built so it feeds the later steps without rework:
- Asset and owner. System name, business service, named technical owner, named risk owner.
- Role. Performs cryptography, has cryptography performed on it, or both.
- Mechanism. Algorithm family, key size, protocol and library. Flag anything RSA or discrete-log based as quantum-vulnerable.
- Location. On premise, cloud tenant, SaaS, or embedded in a device.
- Data lifetime. How long the protected data must stay confidential. This drives the risk level later.
- Changeability. Can the mechanism be swapped by configuration, by code change, or only by vendor release?
- Supplier. Who ships the component, and whether you have a contact who can answer PQC questions.
- Discovery source. Scanner, CMDB, manual interview. Auditors will ask how you know.
Field 6 is your crypto-agility signal. The roadmap defines cryptographic agility as "the design of cryptographic protocols and systems in a modular way that enables replacing the cryptographic components." If a record says "vendor release only," you have found a constraint, not just an asset.
What breaks: scanners see network endpoints and code repositories well. They miss cryptography inside appliances, OT devices, partner-managed SaaS and archived backups. Plan a short interview round with system owners for anything the scanner cannot reach, and record that gap in field 8 rather than pretending it is closed.
How do you build dependency maps that expose internal and third-party migration constraints?
The roadmap says "entities should do a mapping of dependencies for applications, products, platforms, operations, looking at both internal dependencies and third party dependencies."
It is candid about why this matters: "Essential information for prioritisation and planning can only be provided after this step."
The inventory tells you what exists. The dependency map tells you what has to move together:
- Start from your top business services, not from the crypto list. Pick the ones carrying long-lived confidential data or signing trust.
- For each service, trace every inventory record it touches. Your PKI will show up in almost all of them.
- Mark each edge as internal or third-party.
- For every third-party edge, copy the supplier and changeability fields from the inventory.
- Highlight chains where one vendor-only component blocks several services.
The roadmap expects this to tilt outward: "This dependency mapping will eventually be driven by supply chain dependencies."
What breaks: teams map dependencies once and file the diagram. A dependency map that is not tied to the inventory records goes stale the next time someone changes a load balancer. Keep the map as relationships between inventory records, not as a picture.
How do you put quantum risk into top-level risk management so the board evaluates it?
The roadmap's wording here is aimed at governments, so read it precisely. "Member States should encourage public administrations and other critical infrastructures, notably entities in scope of the NIS 2 Directive, to include the quantum threat to cryptography in their top-level risk management. This ensures that the threat and risk is being discussed and evaluated at the board level."
Your task is not a separate quantum register. It is a quantum risk analysis that lands in the register you already run.
The roadmap describes a model from the PQC Migration Handbook that it calls "compatible with the general IT risk management methodology as defined in ISO standard 27001." Three factors set a use case's quantum risk level:
→ "the quantum weakness of the cryptography used"
→ "the expected impact of the used cryptography being broken"
→ "the estimated time and effort required to migrate to PQC"
Your inventory fields 3, 5 and 6 feed those factors directly.
The roadmap gives a rough rule for the high bucket: "If confidentiality needs to be protected for a long time period (at least 10 years), and an attack after 10 years or more would still have significant impact, then the quantum risk level is high." Those are the use cases on the 2030 clock.
For the board: one risk entry per affected business service, scored with your existing method. One slide: how many high-risk use cases, which are blocked by suppliers, and what you are asking for.
How do you bring suppliers into the plan before their roadmaps are fixed?
The roadmap says "every organisation also needs to start the dialogue with product and service suppliers because the transition depends on them." Suppliers "need to integrate PQC and cryptographic agility" into their product and service roadmaps.
The timing point is the one to act on. A requirement you raise while a supplier is still planning is cheaper than one you raise after the product ships.
Run it as a short, structured exercise:
- Pull every supplier from the third-party edges of your dependency map.
- Rank them by how many high-risk use cases they block.
- Send the top tier a written question set: current algorithms, PQC plans, whether cryptography can be updated in the field, and whether software and firmware updates will use quantum-safe signatures.
- Record answers, or non-answers, against the inventory record.
- Feed the gaps into your third-party risk process and your next contract renewal.
The firmware question is not arbitrary. The roadmap's Next Steps call for "a quantum-safe upgrade path, i.e. even if a product is not fully transitioned to PQC, the software and firmware upgrade routines should use quantum-safe signatures." It also notes one may "consider to include cryptographic agility in evaluation of NIS2 conformity." Asking now is cheaper than asking after that becomes a procurement expectation.
What breaks: generic questionnaires get generic answers. Ask about the specific product version you run, and attach the inventory record so the supplier cannot answer about a different line.
What does a seven-week workplan look like, and what evidence shows it has started?
Forty-eight days is about seven weeks. That is not enough to finish. It is enough to start in a way you can show.
| Week | Work item | Evidence artefact |
|---|---|---|
| 1 | Name a PQC owner, a risk owner and a scope list of business services | Signed scope and ownership record |
| 2 | Run discovery tooling across on-premise and cloud | Raw discovery output, dated |
| 3 | Normalise results into CBOM-format inventory records; interview owners for gaps | Inventory with discovery source per record |
| 4 | Build dependency maps for the top services | Dependency relationships linked to inventory IDs |
| 5 | Score quantum risk level per use case | Risk entries in the existing register |
| 6 | Send supplier question sets to the top tier | Sent log and response tracker |
| 7 | Board or risk committee briefing; draft 2027 plan | Minutes and the approved plan |
Every row produces something a supervisor or auditor can read back. "We are working on it" is not evidence. A dated inventory with named owners is.
Do not stop at week 7. The roadmap's own instruction is that the Next Steps "should also be initiated as soon as possible but might take more time to complete."
How Aigis fits
Aigis treats the PQC roadmap like every other source: each First Step is traced to a verbatim citation in the document, mapped to the risks and controls it touches, and linked to the evidence that shows progress. Your inventory, dependency and supplier records sit on that same source-grounded record, next to your NIS2 and ISO 27001 work.
Your agents can do the discovery and chasing. Our engine keeps the record honest.
See how the platform connects risk management and third-party risk to the obligations that actually apply to you at agrc.ai.
FAQ
Is 31 December 2026 a legal deadline for my organisation? No. The roadmap says "Member States should ensure the measures listed in First Steps are accomplished before the end of 2026." It is a Member State milestone in a recommendation-based document. NIS2 entities are told the document will be "useful" to them, and your national authority may build on it.
What is a CBOM? The roadmap describes it as "Cryptographic Bill of Materials, an extension of the SBOM standard" and recommends a standardised format like it for cryptographic inventories.
Do we need a separate quantum risk register? No. The roadmap asks for the quantum threat to be included in "top-level risk management" and describes a model compatible with ISO 27001 risk methodology. Add it to the register you have.
Should we wait for Milestone 1 before starting supplier work? No. The roadmap states that Next Steps should start as soon as possible and that "waiting for the completion of Milestone 1 cannot be afforded."


