Inventors are told, correctly, that describing an invention in public before filing can destroy the right to patent it. What they are told less often is that the severity of that rule varies enormously by jurisdiction, and that the mildest version is the one most commonly quoted. Planning a disclosure strategy around the mildest version is how an invention ends up validly protected in one place and permanently unprotectable everywhere else.
The strict test is easy to state and unforgiving in application. Anything made available to the public, anywhere in the world, in any language, by any person — including the inventor — before the filing date forms part of the prior art. There is no allowance for good faith, no allowance for a small audience, and no fee that repairs the position afterwards. Rights that existed on a Tuesday afternoon are gone by Wednesday morning, and nothing in the system brings them back.
Standard · The Test
What the Strict Rule Actually Asks
Under absolute novelty, the examiner is not asking whether the applicant behaved reasonably. The question is purely factual: on the day of filing, did a public disclosure exist that described this invention? If it did, the claim falls, and the identity of the discloser is irrelevant. An inventor's own conference paper is cited against them in exactly the same terms as a competitor's patent.
Systems with a grace period operate a narrower exception rather than a different test. Typically they disregard disclosures made by the inventor, or derived from the inventor, within twelve months before filing. That is a genuine safety net, but it is a local one. It does not travel, it does not bind other offices, and it does not create any right against a third party who files in the interval having seen the disclosure. Relying on it while intending to file more widely later is the specific mistake this page exists to describe.
Threshold · Public
The Bar Is Far Lower Than It Sounds
"Public" does not mean publicised. It does not require an audience, a sale, a press release, or anybody actually having read or understood the material. In general, a description is public once it has been made available to at least one person who was under no obligation to keep it confidential — and availability, not attention, is the operative word. A thesis catalogued on an open library shelf is public even if nobody borrows it. A page left live on a website for a weekend is public even if the analytics record no visitors.
- A crowdfunding page describing how the mechanism works.
- A stand at a trade fair where the working device is demonstrated.
- A detailed forum post, blog entry or social video showing the internals.
- A conference presentation, poster, abstract or published paper.
- An offer for sale, or a sale of a product from which the invention can be worked out.
- A conversation with a supplier or manufacturer without a confidentiality agreement signed first.
That last item accounts for a disproportionate share of losses, because it does not feel like publishing. An inventor sends drawings to three contract manufacturers for quotations. None of them is bound by anything. Each is free, legally speaking, to describe what they received to whoever they like — and even if none of them does, the disclosure has already occurred. Confidentiality has to precede the conversation; an agreement signed after the drawings were sent repairs the relationship, not the novelty.
Availability is the test, not attention. One recipient under no duty of confidence is enough to end the question.
Why "nobody saw it" is not a defence
Asymmetry · The Trap
How a Local Concession Becomes a Global Loss
The damaging sequence is almost always the same, and it is executed by careful people acting on partial information. An inventor learns that their home system allows a twelve-month grace period for their own disclosure. Reassured, they launch a funding campaign, publish the mechanism in detail, take pre-orders, and file a first application four months later — comfortably inside the local allowance. The home filing is fine.
Eleven months after that, with early sales encouraging, they decide to pursue protection in the territories where the product is being manufactured and copied. Those offices apply absolute novelty. Their own campaign page, published four months before their own first filing, is prior art against them in every one of those jurisdictions. The application is unpatentable there and always will be. The invention has a valid patent in one place and no prospect of one anywhere else, and the cause was a page the inventor wrote themselves.
Coverage of consumer inventions rarely dwells on this, because the interesting part of the story is the product. Accounts such as this report on a kitchen mixing device developed by two first-time inventors and a fuller account of the inventors behind it describe products that reached shelves — which necessarily means the filings preceded the publicity, in the correct order, quietly, before anybody wrote about them.
Exceptions · Narrow
The Few Disclosures That Are Excused
Strict systems are not entirely without mercy, but the exceptions are drawn tightly and are rarely available to the situation inventors actually find themselves in. Two recur. The first covers disclosures resulting from an evident abuse — material obtained in breach of confidence and published by someone who had no right to publish it. The second covers display at certain officially recognised international exhibitions, subject to formal notice and documentation at the time of filing. Both are commonly limited to the six months before filing.
Neither is a substitute for silence. The abuse exception requires proving both the confidential relationship and the breach, years later, in a forum designed to test such claims sceptically. The exhibition exception requires the event to qualify under specific criteria and the applicant to have completed the paperwork correctly at the outset. Any inventor who is currently relying on either has already made the mistake and is now hoping to be forgiven for it.
Practice · File First
The Only Policy That Works Everywhere
The reliable rule needs no knowledge of comparative law: file before disclosing, always, and treat every conversation outside a signed agreement as publication. A placeholder application filed for a few hundred in official fees converts an irreversible loss into a manageable timeline. It is not the cheapest thing an inventor can do — talking to people is free — but it is the cheapest insurance available against a category of loss that cannot be insured any other way.
Two operational habits follow from that. First, the filing has to describe what will actually be shown; a placeholder covering the January concept does not shelter a demonstration of the April mechanism. Second, the surrounding calendar has to be understood before the first disclosure rather than after it — the sequence of obligations that begins the moment an office stamps the first receipt is set out in this account of the twelve-month clock, and it is the frame within which every question of absolute novelty is answered in practice.
What makes this rule hard is not its complexity. It is that invention is exciting and secrecy is not, and the cultural script rewards announcing things. The entertainment industry's long fascination with the moment of innovation trains everybody to expect a reveal, and the reveal is precisely the wrong instinct at this stage. The practical alternative is unglamorous and well documented — a step-by-step account of moving an invention idea toward a real product puts the filing before the announcement for reasons that only become visible when it is too late to reorder them.
Where there is no grace period, there is no second draft of the disclosure. The first description of the invention is the only one that counts.
The rule that travels
File, then speak. In the strict territories that order is the whole of the law.