The 2024 revision of eIDAS created a new class of evidence: qualified electronic ledgers. The implementing rules published in December 2025 set out what these systems must do (Regulation 2024/1183, Implementing Regulation 2025/2531). A record entered on a certified ledger now carries a statutory presumption of integrity and chronological order in all 27 member states.
In any EU court, even in a dispute between companies from different member states, that presumption moves the burden of proof onto whoever challenges the record. So it is fair to ask what technical breakthrough earned it.
There isn't one.
The framework does not require decentralization, fault tolerance across independent parties, or open verification. It is perfectly comfortable with a closed system run by a single company. Read past the regulatory vocabulary and the required controls are signatures, hash chains and a protected signing key: the same toolkit engineers have bolted onto ordinary databases for decades.
The EU has not raised the technical bar. It has stamped the existing one and attached a legal privilege to the stamp. For IP owners and the lawyers who defend them, mistaking that stamp for stronger evidence is a costly error. It trades proof that anyone can verify for a label that only works where the label is recognized, and only if the vendor is still around to back it.
A Ledger in Name Only
"Qualified electronic ledger" sounds like distributed, tamper-proof infrastructure. The annex to Implementing Regulation 2025/2531 describes something much plainer:
- advanced electronic signatures or seals based on qualified certificates, to show where data came from;
- hash lists or hash trees (SHA-256 or stronger) to fix the order of records;
- signature-based integrity checks, so that any later change is detectable;
- certified hardware (Common Criteria EAL 4 or FIPS 140-3 Level 3) to hold the signing keys.
Nothing on that list needs a blockchain. Distributed ledgers are allowed, with some extra ISO requirements, but they are an option, not a condition. An append-only PostgreSQL table with signed entries, a Merkle root and a hardware security module (HSM, the tamper-resistant device banks use to guard their signing keys) clears the same functional bar.
Certified "qualified ledger"Equivalent system, uncertifiedTechnologyDatabase + PKI + hash chainDatabase + PKI + hash chain (or stronger)Legal weight in the EUStatutory presumption (eIDAS Art. 45k)Must prove its integrity from scratch
Ledger vendors will object that their systems are distributed, pointing to the number of nodes they run. But ten servers under one company's control are still one company. Whoever holds the root keys, the hosting contracts and the admin console can go bankrupt, be coerced or make a mistake, and all ten nodes go down with them.
The result is a legal hierarchy that ignores engineering. An audited database labelled "qualified" gets the presumption; an identical or better system without the label starts from zero. The law trusts the vendor's paperwork more than the mathematics of the evidence.
An Audit Is Not a Guarantee
Defenders point to the supervision regime. Qualified trust service providers are audited at least every two years and must keep a termination plan that preserves access to records if they shut down (Implementing Regulation 2025/2530, Article 3).
A termination plan is an intention, not a proof. It works only if, at the moment the provider collapses, there is still money, a willing successor and staff who follow the plan. Insolvency, distressed sales and restructuring are exactly the moments when none of that is certain.
An audit report says a provider passed a checklist on the day of the audit. It cannot bring back an archive that has gone offline, or rebuild a chain of custody halfway through a trial. Suing a bankrupt trust provider will not rescue a claim that has lost its proof.
Evidence Must Outlast the Vendor
Intellectual property can stay valuable, and contested, for decades. Take a company that keeps a process secret instead of patenting it. If a competitor later patents the same idea, the company can keep using its process only by proving it got there first, and that proof must hold for as long as the rival's patent can be enforced: up to twenty years.
Software companies rarely last that long. Products are discontinued, APIs deprecated, startups acquired or wound down. Meanwhile the disputes that matter (a former supplier reusing your drawings, a new patent covering a process you have run for years) often surface ten or fifteen years after the work was created.
At that point you must prove that the work existed, that it has not changed, and that it was yours. A ledger entry cannot do that on its own if checking it requires the vendor's closed API, a licence for discontinued software, or an escrow archive nobody has opened in a decade. When the provider goes, the evidence goes with it.
Every IP lawyer should therefore apply one test:
If the ledger vendor disappeared tomorrow, could an independent expert export the raw proofs and verify them in court using only open-source tools and public standards?
If the answer is no, you do not own your evidence. You rent it.
The Presumption Stops at the Border
IP is international. An invention developed in Munich is patented in the US, manufactured in Shenzhen and licensed in Tokyo. Infringers do not pick their jurisdiction to suit Brussels.
The presumption in Article 45k binds EU courts. A US federal judge, a Singapore arbitral panel or the High Court in London owes it nothing. Outside the EU:
- admissibility follows local procedure and expert-evidence rules, such as Daubert challenges in the US, where the opposing party asks the judge to exclude expert testimony whose method cannot be independently tested;
- an EU certificate does not replace a technical account of how the record was produced and checked;
- an expert who has to concede that verification depends on a vendor's undisclosed, closed-source code invites exclusion, or at best a heavy discount.
Compare an open standard such as RFC 3161 for timestamps. The format and the verification procedure are public, and any competent expert anywhere can check a token with free tools, without a licence and without the issuer's help. Evidence built that way travels. Evidence that relies on an EU certificate for its weight is strong in EU courts, but everywhere else it has to prove itself from scratch.
Use the Badge, Don't Build on It
None of this makes qualified ledgers useless. Inside the EU, a statutory presumption is a real procedural advantage, as long as the record stays verifiable for the life of the asset and does not depend on the vendor's survival. Outside the EU, it adds very little.
So IP owners do not have to choose between local recognition and global verifiability. They should layer the two. Anchor evidence first in open, global infrastructure that anyone can check without asking permission, such as RFC 3161 timestamps or the Bitcoin blockchain. Then, where a dispute is likely to be fought, add the proof that local courts recognize: a qualified ledger in the EU, an accredited timestamping authority in China, a government-accredited timestamp in Japan.
When you decide how your company records its core assets:
- Anchor globally first. Use open standards and public infrastructure that any expert, in any jurisdiction, can verify with free tools.
- Add local recognition where you may litigate. Treat nationally recognized proofs as a layer on top, not as the foundation.
- Keep the data and the proofs. A proof only certifies a hash; without the exact original file, it proves nothing. Keep the certified data, byte for byte, together with self-contained proof packages (raw signatures, hashes and certificate chains), all in your own custody.
- Verify without the vendor. Make sure your own team can check any historical record with open-source tools, with the vendor's platform switched off.
A statutory presumption can win you the argument in one jurisdiction. Proof that anyone can check lets you make the argument anywhere.
