Copyright, trade secrets, designs, patents and trademarks are five different intellectual property rights, built on different statutes, with different tests and different terms. Every one of them is decided the same way: by whoever can prove what existed, and when.
Intellectual property is usually taught and sold as a set of separate boxes. Copyright protects expression. Trade secrets protect confidentiality. Design rights protect appearance. Patents protect inventions. Trademarks protect the use of a mark in trade. Each has its own statute, test, and term.
The separation is real at the level of the law. It disappears at the level of the dispute.
A court resolving any of these claims is not applying five different evidential doctrines — it is asking the same practical question, dressed in different statutory language.
A copyright claim turns on whether the claimant's work existed before the defendant's. A trade secret claim turns on whether the secret was defined and protected before it left the building. A design right turns on when it was first shown. A patent turns on a filing date, and on what existed beforehand. A trademark survives revocation only if its use can be dated inside a five-year window. Different statutes, same question, asked five different ways: what did you have, and on what date.
This matters because the businesses that lose these disputes rarely lose on the law. They lose because the record that would answer the question does not exist, or exists in a form the other side can dismiss as self-produced.
A well-drafted contract, a correctly filed application and a properly registered mark all matter, and none of them substitutes for the missing record. The rest of this post covers where each right's version of that question arises, and what closes the gap. Each section links to a full treatment, so this post stays the map rather than all eight territories at once.
Right | How It Arises | What's Contested | Full Treatment |
Copyright | Automatic, on fixation | Who created it first | How to Protect Your Copyright |
Trade Secrets | Secrecy, value, reasonable protection steps | Whether it was defined and protected before it left | How to Prove a Trade Secret Existed |
Design Rights | First public disclosure | When and where it was first shown | How to Prove You Designed It First |
Patents | The filing date | What existed before that date | The Dates a Patent Filing Does Not Record |
Trademarks | Registration, revocable after 5 years unused | Whether use is dated inside the window | Trademark Proof of Use |
Copyright is the starting point because it needs nothing to exist. It arises automatically the moment a work is fixed in tangible form, with no application, no fee, no register.
That is also its evidential weakness. There is no file number to point to, no certificate confirming a date. When authorship or priority is disputed — a claim that one party copied another, a dispute with a departing collaborator — everything rests on records the creator produced themselves, dated by systems the creator controls. For writing, photography, music composition and other creative work, what that record needs to look like is covered in full on how to protect your copyright.
Software is a specific version of the same problem, with its own evidence. A Git commit timestamp comes from the machine that made the commit, and is rewritable by a rebase or a force push. That is not a hypothetical weakness; it is the first thing opposing counsel raises. What resolves it is a record of the code deposited by a party with no stake in the outcome, covered in full on how to prove you wrote your source code.
A trade secret is not a right that exists automatically in the way copyright does. Article 2(1) of Directive (EU) 2016/943, mirrored by the UK Trade Secrets Regulations and the US Defend Trade Secrets Act, sets three conditions: secrecy, commercial value from that secrecy, and reasonable steps to keep it secret.
The intuition is that trade secret claims turn on whether the information was taken. More often the deciding point comes earlier: whether the holder can show the secret was defined with enough precision to distinguish it from general know-how, and that dated evidence of concrete measures backed the claim. Neither is a question about the taking; both burdens sit with the holder from the outset.
This is the right most exposed by the departure of a single person. An employee leaves, a competing product appears, and the case turns on whether the employer can show what the secret was, distinct from general skill and knowledge, before they left. Records a company produced on its own systems answer that weakly, because the other side will argue they were written to fit the litigation. The full mechanics are set out on how to prove a trade secret existed.
Unregistered design protection inverts the usual pattern entirely. Instead of registration creating the right, disclosure does. The moment a design is first made available to the public, an unregistered EU design and a UK supplementary design both spring into existence, and both run for three years from that date, with nothing marking when the clock started.
That makes disclosure do two jobs at once. It starts the clock and fixes the territory the right exists in — a question left open since Article 110a(5) of the Community Design Regulation, which required first disclosure inside the EU, was deleted outright on 1 May 2025 with no reasoning given. Firms still advise simultaneous disclosure in multiple territories until a court rules on what it means.
A French court has already engaged with this category of evidence. In AZ Factory v Valeria Moda, decided by the Tribunal judiciaire de Marseille on 20 March 2025, blockchain-anchored cryptographic timestamps were admitted alongside corroborating evidence, and the court went on to find infringement and award damages. The blockchain evidence in that case was not produced by Provlyn, and we have no involvement in the matter.
It was a first-instance decision and does not bind other courts, but it is a demonstration that this category of evidence is not theoretical. The full account, including what the judgment did and did not establish, is on how to prove you designed it first.
A patent is the outlier in this list, because one of its critical dates is settled beyond argument. The filing date is fixed by the patent office, an independent third party with nothing at stake, and neither side ever contests the record.
Almost everything else around a patent is contested. When the invention was conceived. What existed at that point. What was shown to whom, under what agreement, before filing. Whether anything reached the public first, which under the European Patent Convention's absolute-novelty standard can permanently destroy the right, with no general grace period. Prior user rights in the UK under section 64 of the Patents Act 1977, and the narrower US equivalent under 35 U.S.C. 273, both turn on what a party was doing and when, and the US version demands the clear-and-convincing standard, not the ordinary civil balance.
None of those dates appear on any register. They live in private files, compared in a dispute against an opponent's files of the same kind, both self-produced. What changes that position, and where software prior art fits, is covered on the dates a patent filing does not record. Where the strategic choice is to publish rather than file, forfeiting patent rights to stop anyone else patenting the same thing, the same dating problem reappears, addressed on defensive publication.
A registered trademark is the one right on this list that can be actively taken away, not merely difficult to enforce. Section 46 of the Trade Marks Act 1994 and Article 58 of the EU Trade Mark Regulation both allow revocation after five years without genuine use, brought by any third party seeking to free the mark, and the burden falls on the proprietor to demonstrate use, not the challenger to prove its absence.
What defeats proprietors is rarely that the use did not happen. It is that the evidence cannot be dated inside the five-year window the application puts in issue. A website is the worst version of this problem: continuously overwritten, so the page that carried the mark three years ago no longer exists anywhere the owner controls, and a screenshot taken today proves nothing about then. The genuine use standard from Ansul, and where proprietors typically lose, is set out on trademark proof of use.
Every one of these rights is disclosed to somebody before a dispute ever starts. Investors, factories, contractors, collaborators, acquirers. Two problems recur across all of them regardless of which right applies.
The first is defining what an agreement covers. A confidentiality clause is squeezed between too broad to enforce and too narrow to cover what mattered, drafted before anyone knows what will be shared. Defining scope by the cryptographic fingerprint of each document, rather than by description, resolves it, and the case law on what courts require is set out on how to define what an NDA covers.
The second is proving what was sent, which a scope definition does not touch. A sent folder is held by the party relying on it as evidence; a recipient's download log is held by the other side. Neither is independent. This is the same problem in a specific commercial setting on who received which version of your deck, where a raise with twenty investors and twelve deck versions turns "did they see this" into a question nobody's outbox can answer.
Every section above describes a different right reaching the same evidential wall: a record that exists, but was produced and dated by the party who needs it to say what it says.
What closes that gap is the same mechanism regardless of which right is in question. A file is hashed with SHA-256, the hash is timestamped under RFC 3161 by an accredited trust service provider, and the timestamped hash is anchored to the Bitcoin blockchain. This is a single automated sequential process, each step depending on the output of the preceding one. The result is a portable certificate showing that a file with that exact fingerprint existed at that exact time, produced by a party with no stake in whichever way the dispute goes.
With optional eIDAS Article 41 qualification from an accredited Qualified Trust Service Provider, the timestamp carries a legal presumption of accuracy across EU member states, shifting the burden onto whoever disputes the timestamp. That presumption currently runs one way: the UK continues to recognise EU-qualified timestamps, but the EU does not automatically recognise a timestamp qualified only in the UK. Every certificate can be checked independently — SHA-256, OpenSSL, the European Commission's trusted lists, any OpenTimestamps client — because the party reading it in a dispute has no reason to take the issuer's word for anything.
What it does not do matters as much as what it does. A timestamp does not establish authorship, ownership, genuine commercial use, or that reasonable steps were taken to protect a secret. Those remain questions for the specific law governing each right, on the wider factual record. What a timestamp removes is the argument about dates — across every right described above, usually the argument that has to be won before the others can be reached.
None of this answers whether a given asset should be patented or kept as a trade secret, registered as a design or left unregistered, filed as a trademark or published defensively to deny it to a competitor. Those are strategic questions for a patent attorney or IP solicitor, turning on cost, duration, the ease of reverse-engineering, and how long an advantage needs to last.
What is common to every one of those strategic choices is that they are all one-way. A patent application publishes. A trade secret disclosed stops being one. A design shown destroys European registration novelty, with no general grace period. Whichever route is chosen, the choice is made once, and the evidence from before that point is the only version that will ever exist.
For a business holding several of these rights across the same body of work — a product that is patentable, whose source code is a trade secret, whose branding is a registered mark, whose packaging carries a design right — the practical answer is not eight different processes. It is the same deposit, made as each asset comes into being, feeding whichever right applies. The law treats copyright, trade secrets, designs, patents and trademarks as five different intellectual property subjects. The evidence they all depend on is one.
This post provides general information about intellectual property rights and the evidence supporting each one. It is not legal advice. For advice on a specific matter, consult a qualified lawyer or patent attorney in the relevant jurisdiction.
Related Reading
How to Prove You Wrote Your Source Code
How to Prove a Trade Secret Existed
How to Prove You Designed It First
The Dates a Patent Filing Does Not Record
Trademark Proof of Use: Evidence That Falls Inside the Window
James Snell is the founder of Provlyn, a platform providing cryptographic prior proof of IP ownership. provlyn.com