A European studio registers its catalogue with the US Copyright Office. It costs little, the certificates arrive, and everyone feels better. Two years later a competitor in Germany copies a design, and the certificates turn out to do almost nothing about it.
This is not a story about a mistake. Registering was reasonable. The gap is between what a registration is for and what people assume it is for.
What a US registration is actually for
It is a procedural instrument of one national system. Under US law, registration acts as a precondition for bringing an infringement action over a US work, and registering early enough opens up categories of remedy that are otherwise unavailable.
Those are real advantages and they are worth having if the United States is a market you care about. They are also, every one of them, advantages inside US courts under US law. The certificate is not a global title deed. It is a filing receipt with procedural consequences attached, in one jurisdiction.
Copyright is territorial
This is the part that surprises people, and it is the whole explanation.
There is no single worldwide copyright. What exists is a set of national copyrights, one per country, each governed by that country's own law. When a work is copied in Germany, German law decides what happened and what follows. A French court applies French law to acts on French territory. The Berne Convention coordinates between these systems. It does not replace them with one.
So the question is never "is my copyright valid in Germany". It is "what does German law say about this act", and the answer to that does not depend on a filing made in Washington.
What national treatment does and does not carry across
Berne's central mechanism is national treatment. A member country must give authors from other member countries the same protection it gives its own nationals. This is a strong protection and it is why your work is protected in Germany without you doing anything.
What national treatment carries is the right. What it does not carry is the American procedural record.
A German court applying German law to a German act does not inherit US registration requirements, because German law has none. Nor does it treat a US certificate as having the procedural effect it would have in the United States, because that effect is a feature of US statute rather than of the document itself. The certificate is still evidence of something. It is simply evidence of a filing on a date, weighed like any other evidence, not a key that unlocks anything.
The right crossed the border automatically. The receipt did not.
The practical consequence for a European rights holder
The registration cost little and did nothing wrong. The problem is what it displaced: the belief that the evidence question was handled meant nobody built any other record.
When the German dispute arrives, the questions are the ordinary ones. Which version of the design existed, when did it exist, who made it, and can any of that be shown to somebody who was not there. A US certificate speaks to a date of filing, which is usually well after the date of creation, and says nothing about which file, which version, or which member of the team.
Registering in every country is not a serious answer either. Most do not run a register to register in, the ones that do are not coordinated, and the cost scales with a catalogue that grows weekly.
What travels with the work instead
Something attached to the file rather than to a jurisdiction.
A cryptographic hash binds a record to one exact version, so the thing being proved is a specific artefact rather than a title or description. A qualified electronic timestamp from a Qualified Trust Service Provider fixes when that record existed, independently of any date field you control. Under eIDAS, a qualified timestamp carries a legal presumption as to the date and time it indicates and the integrity of the data it is bound to.
Be precise about what that gets you, because this is the point where the claim is easiest to overstate. A sealed record is technically verifiable anywhere: anyone, in any country, with no account and no contact with you, can check that this file matches this hash and that this timestamp was issued when it says. That is a property of the cryptography and it genuinely does not stop at a border.
What that is not is automatic legal admissibility everywhere. How a given court treats a given piece of evidence is a matter for that court under its own procedure, and no technology changes that. The eIDAS presumption is an EU instrument and operates as such.
The honest framing is narrower than a slogan and more useful than one. The record travels with the file and can be checked wherever it lands. What each forum makes of it is a question for that forum. That is still a substantially better position than a certificate that is inert the moment it leaves one country, or than a folder of screenshots.
What this means for a first register outside the United States
It suggests a register organised around the wrong thing would repeat the mistake. A second national register produces a second national record, useful in one place and inert elsewhere, which is precisely the limitation being complained about.
The useful design is not national at all. Bind to the work by hash rather than to a description. Take the time from a third party rather than from the rights holder. Make the result checkable by anyone without an account. Then the record is portable by construction, and its usefulness does not depend on which country the dispute turns up in.
For teams whose work is already crossing borders, that is the more practical place to spend attention. Our page for IP legal teams covers how this fits an existing enforcement workflow, and the Article 50 page covers the same evidence mechanics from the AI-disclosure side.
