ENISA's 22 October 2026 webinar on how Estonia is approaching the post-quantum cryptography (PQC) transition included a slide deck on X-Road, the data exchange layer that links Estonian public and private registers. According to the hook material for this piece, the deck sets out an 11-step action plan. RIA, the Estonian Information System Authority, owns and operates X-Road. NIIS, the Nordic Institute for Interoperability Solutions, develops it. Cybernetica is running a 2025-2026 register-migration pilot for RIA. We could not open the slides directly (the server returned HTTP 429), so we confirmed the event, its agenda and its speakers from ENISA's event page, which links the deck. The step-by-step detail below rests on the hook description and on NIIS's own published study.
The case for moving early
Start with the strongest argument for urgency. Encrypted traffic can be recorded today and decrypted later, once a large fault-tolerant quantum computer exists. NIIS's public report on migrating X-Road to PQC, dated 13 January 2026, calls this the "Harvest Now, Decrypt Later" threat. It says the threat is most serious for data with retention requirements beyond 10 years. Population, tax and health records exchanged over X-Road fit that description. Cryptographic migrations also take years of planning, testing and deployment. Waiting for a quantum computer to appear before starting would be a bet that the migration takes less time than the threat does.
That argument holds up, and Estonia has acted on it. The NIIS study catalogued 27 cryptographic assets in the X-Road data exchange layer, including keys, digital signatures and hashes. It mapped 10 quantum-enabled threats onto them. It recommended 11 security controls and a hybrid (dual-stack) migration plan that starts with high-risk, long-confidentiality-life assets.
Why this is a good model of proportionate policy
The notable feature of the Estonian approach is sequencing. ENISA's event page describes the national roadmap in three phases: a cryptographic inventory of existing systems, detailed transition plans with timelines and priorities, and implementation across digital infrastructure. The NIIS study follows the same order, starting with an inventory and a risk assessment before any algorithm swap.
This matters for innovation policy. A blanket mandate to "be quantum-safe by date X" forces organisations to guess where the risk sits. An inventory-first approach lets scarce engineering time go to assets where the harm from a delay is greatest. It also keeps the door open to better algorithms as standards mature. Hybrid schemes let operators run classical and post-quantum cryptography side by side. That means a flaw found later in a new algorithm does not leave the platform exposed.
The division of labour looks sensible too. X-Road is shared infrastructure, and NIIS's mandate covers development shared with Finland and Iceland. A single technical owner that tests changes once is far cheaper than hundreds of agencies each improvising. Cybernetica's register-migration pilot, a vendor-run test with RIA as the customer, is a good way to learn where migrations break before they reach production registers. The company has said it also won three Estonian government procurements on PQC, including a quantum-risk assessment for the Population Register, in a press release dated 10 November 2025. That is a vendor's own statement, not an official one.
Where the plan needs scrutiny
The NIIS report is candid about the costs. It expects larger keys and signatures to add performance overhead, and it flags "major logistical challenges" in rebuilding the platform's public key infrastructure. On a data exchange layer that handles very large numbers of requests, those costs are real. They are the reason a pilot matters, and the reason its findings should be published.
Three points are worth watching:
- Pilot transparency. The slides, as described, give no pilot results. A 2025-2026 pilot that ends without public findings on latency, signature size and rollback would waste most of the lesson for other X-Road operators and for other governments.
- Timeline credibility. The European Commission and Member States published a coordinated implementation roadmap for the transition to PQC on 23 June 2025 (the page was last updated on 3 September 2026). It asks Member States for a synchronised transition. We could not confirm the specific milestone years from that page, so we do not restate them here. Estonia's plan will only be credible if its steps map onto whichever milestones the roadmap sets.
- Inconsistent public descriptions. ENISA's page says Estonia has already made its National PQC Roadmap available. Cybernetica's 2025 release says it was contracted to develop that roadmap. Both can be true if the roadmap was completed after the release, but readers deserve a clear public version and date.
There is also a fair caution from within Estonia's own research community. At a 2024 conference, cryptographer Jan Willemson argued that classical cryptography can remain acceptable where the confidentiality horizon is short, as reported by The Quantum Insider. He also warned that you do not always know the future value of your signatures. That is the right way to think about triage. It argues for risk-based prioritisation, not for waiting.
What other governments should take from it
The lesson is not that every country should copy an 11-step plan. It is that a regulator should earn trust in a deadline by showing its working: a published inventory, a threat model, named owners, a pilot, and results that others can reuse. Governments that skip straight to compliance dates will produce paperwork. Governments that publish their inventories and pilot findings will produce migration knowledge, and the open-source, multi-country X-Road model makes that knowledge easy to share.
Estonia is also showing that a small state can lead on infrastructure security without heavy new regulation. The work is being done through technical bodies, procurement and public reports. If the pilot findings are published, this will be a template worth copying.