Patent application publication happens at eighteen months from the earliest priority date, and it is not optional in most cases.
It ends secrecy permanently. Whether the application grants, gets rejected, or is abandoned the following week, the disclosure is public forever.
It also creates the blind spot that makes freedom-to-operate analysis inherently incomplete: anything filed in the last eighteen months is invisible to everyone.
Which cuts both ways. Your competitors cannot see your filings either.
Why the rule exists
| Purpose | Detail |
|---|---|
| Public notice | Others learn what is being claimed |
| Prior art creation | Prevents duplicate effort |
| International harmonisation | Most countries publish at 18 months |
| The bargain | Disclosure in exchange for a limited monopoly |
Publication is the disclosure half of the patent bargain, delivered whether or not the monopoly half ever arrives.
The rule
| Timing | 18 months from the earliest priority date |
| Measured from | A provisional, a foreign filing, or the non-provisional |
| Depends on examination | No |
| Depends on the claims being allowed | No |
| Reversible | No |
| Applies to provisionals alone | No — only once a non-provisional follows |
A provisional never followed up stays unpublished. That is one of the quieter advantages of the provisional route — twelve months to decide, with secrecy intact if you abandon.
What publication does
| Effect | Detail |
|---|---|
| Ends trade secret protection | Permanently and irreversibly |
| Makes the disclosure prior art | Against later applications by others |
| Makes the file history public | Office actions, responses, cited art |
| Creates provisional rights | Narrow — see below |
| Signals your direction | Competitors can read it |
| Enables third-party submissions | Others can cite art against you |
The trade secret point is the one that cannot be undone. Anything you want to keep secret must be kept out of the application entirely, and that decision has to be made before filing.
Abandoning does not restore secrecy. The publication stands.
Third-party submissions
| Who may submit | Anyone |
| What | Prior art relevant to a published application |
| Window | Defined, relatively early in prosecution |
| Effect | The examiner must consider it |
| Cost to the submitter | Low |
Publication opens your application to this. A competitor watching your filings can submit art they believe defeats your claims, at very little cost.
It works in both directions. Monitoring competitor publications lets you do the same.
It is automatic
| Requires a request | No |
| Requires examination first | No |
| Can be withdrawn | No |
| If abandoned afterwards | The publication stands |
Provisional rights
| Requirement | Detail |
|---|---|
| Period covered | Publication to grant |
| Remedy | A reasonable royalty only |
| Actual notice | Required — the infringer must know |
| Claims substantially identical | Published claims vs granted claims |
| In practice | Rarely produce recovery |
The substantially-identical requirement is what defeats most claims. Applications are usually amended during prosecution, and once the granted claims differ meaningfully from the published ones, provisional rights fall away.
A published application confers no right to exclude. "Patent pending" is a signal with deterrent value and no legal force. See patent is pending.
Who reads published applications
| Party | Looking for |
|---|---|
| Competitors | Your direction, and art to submit |
| FTO analysts | Claims that might block their product |
| Examiners | Prior art against later applications |
| Potential buyers | Assets and families |
| Investors | Technical diligence |
Non-publication requests
| Available | Only if you will not file abroad |
| Certification | The invention will not be the subject of an application in a country requiring 18-month publication |
| Filed | With the application |
| If you later file abroad | Must rescind, and notify within 45 days |
| Failure to notify | Application regarded as abandoned |
This is a trap with a hard edge. Filing abroad after a non-publication request, without rescinding and notifying within the deadline, can abandon the application entirely.
Most applicants should not use it. It forecloses foreign filing, which is a large concession for eighteen months of additional secrecy.
What the delay means in practice
| Party | Consequence |
|---|---|
| You | Competitors cannot see your recent filings |
| Competitors | You cannot see theirs |
| FTO analysts | The gap cannot be closed |
| Examiners | Cite art they can see |
Everyone operates with the same blind spot, which is at least symmetrical.
Why the rule exists
| Purpose | Detail |
|---|---|
| Public notice | Others learn what is claimed |
| Prior art creation | Prevents duplicate effort |
| Harmonisation | Most countries publish at 18 months |
| The bargain | Disclosure for a limited monopoly |
Publication is the disclosure half of the patent bargain, delivered whether or not the monopoly half ever arrives.
The invisible window
| Applications filed in the last 18 months | Invisible to everyone |
| Effect on FTO searches | Cannot be complete |
| Effect on patentability searches | Same gap |
| Can more searching close it? | No |
| Who it affects | Everyone equally |
This is why no freedom-to-operate opinion can promise certainty. A competitor may have filed something blocking last month and nobody can see it.
It also means pending continuations are a live risk. A competitor whose parent published can still be pursuing new claims aimed at your product. See FTO meaning.
Foreign publication
| Jurisdiction | Publication |
|---|---|
| United States | 18 months from earliest priority |
| EPO | 18 months |
| PCT | 18 months |
| Most others | 18 months |
| Non-publication request | US only, and forecloses foreign filing |
Eighteen months is close to universal, which is why the US non-publication option requires certifying you will not file abroad.
A PCT application publishes at eighteen months regardless, so the PCT route and non-publication are mutually exclusive. See international patent law.
Reading other people's published applications
| Source | Shows | Cost |
|---|---|---|
| USPTO Patent Center | Application, full file history | Free |
| Google Patents | Full text, translations | Free |
| Espacenet | Worldwide families | Free |
| In the file history | Tells you |
|---|---|
| Office actions | What art the examiner cited |
| Responses and amendments | Where the claims are heading |
| Repeated rejections | The claims are under pressure |
| Notice of allowance | What will grant |
| Continuations pending | The family stays open |
Published claims are usually broader than what will grant. Treat them as a worst case rather than a forecast, and watch the amendments.
The file history is the more useful document. It shows the examiner's view of the prior art, free.
What to monitor after publication
| Watch | Why |
|---|---|
| Third-party submissions against your application | Art you must address |
| Competitor publications in your classes | New risks |
| Continuations in relevant families | Claims may be aimed at you |
| Foreign counterparts publishing | Different claims |
| Your own file history | What the examiner cited |
Published continuations are the highest-value thing to watch. They show a competitor pursuing further claims from a disclosure you already know about, frequently written with your product in view. See patent monitoring.
Defensive publication in detail
| Feature | Detail |
|---|---|
| Cost | Very low |
| Effect | Prior art against everyone, immediately |
| Right to exclude | None |
| Routes | Journals, technical disclosure services, own website with a date |
| Best for | Fear of being blocked, not blocking others |
It prevents anyone else patenting the invention, which is frequently the actual concern rather than wanting a monopoly yourself.
Publication as a strategy
| Use | Detail |
|---|---|
| Defensive publication | Stops others patenting it, cheaply |
| Signalling to competitors | Deterrent effect |
| Establishing prior art | Protects your freedom to operate |
| Attracting licensees or buyers | Visibility |
| Investor evidence | Something concrete to point at |
Defensive publication is underused where the concern is being blocked rather than blocking others. It costs very little, takes effect immediately, and prevents anyone else obtaining a patent on the invention.
It gives no right to exclude, which is the trade. See can you patent something and make it free.
Redacted and early publication
| Option | Detail |
|---|---|
| Request early publication | Available, with a fee |
| Why | Starts provisional rights sooner |
| Redacted publication | Narrow circumstances |
| Effect on secrecy | Same — publication is publication |
Early publication is occasionally useful where you want the disclosure on the record as prior art against competitors sooner, or to start provisional rights.
It does nothing about the eighteen-month blind spot for anyone else.
Worked example: the secrecy decision
A company has a manufacturing process competitors cannot detect from the product.
| Option | Effect |
|---|---|
| File a patent | Publishes at 18 months — secrecy gone forever |
| Term | 20 years from filing, then public anyway |
| Keep as a trade secret | No expiry, while it stays secret |
| Risk | Independent discovery, reverse engineering |
| File and abandon | Worst outcome — published, no patent |
| Consideration | Weight |
|---|---|
| Can competitors detect it from the product? | No — favours trade secret |
| Could they independently develop it? | Possible |
| Would we enforce a patent on it? | Hard to detect infringement |
| Value of 20 years vs indefinite | Indefinite wins if secrecy holds |
They kept it secret. A process that is undetectable from outside is difficult to enforce as a patent and valuable to hold as a secret.
Filing and then abandoning is the worst outcome available, and it happens when applicants file to feel protected without deciding what they actually want.
Watching a competitor's application
| What to check | Tells you |
|---|---|
| Published claims | The worst case, not the forecast |
| Office actions | What art the examiner found |
| Responses and amendments | Where the claims are heading |
| Repeated rejections | Claims under pressure |
| Continuity data | Whether continuations are pending |
| Foreign counterparts | Different claims elsewhere |
Amendments are the most informative part. A claim narrowed twice is heading somewhere specific, and the direction is visible before grant.
All of it is free in USPTO Patent Center.
What publication does not mean
| Not | Reality |
|---|---|
| You have a patent | An application under examination |
| The claims are final | Usually narrowed before grant |
| You can enforce it | No right to exclude yet |
| Examination has happened | Publication is independent of it |
| Others cannot file | They can; your disclosure is prior art against them |
Publication makes it prior art
| Against | Effect |
|---|---|
| Later applications by others | Your disclosure blocks them |
| Their claims | Cited by examiners |
| Your own later applications | Also cited against you |
Your own publication can be cited against your later filings, which is a reason to keep families open through continuations rather than filing afresh.
The trade in one table
| You give | You get |
|---|---|
| Permanent public disclosure | The possibility of a patent |
| Trade secret protection, forever | 20 years from filing |
| Visibility to competitors | Prior art position against them |
| Openness to third-party submissions | Provisional rights, narrowly |
This is the patent bargain in its plainest form. Disclosure now in exchange for a limited monopoly, and the disclosure happens whether or not the monopoly ever materialises.
Timeline in context
| Event | Timing |
|---|---|
| Priority filing | Day 0 |
| Publication | 18 months |
| First office action | 19.9 months average from filing, FY2024 |
| Disposal | 26.3 months average, FY2024 |
| Grant | Variable |
Publication frequently precedes the first office action. Your application becomes public before an examiner has said anything about it.
Which means competitors may read it before you know whether it will grant.
After publication
| Task | Detail |
|---|---|
| Expect competitor attention | Your direction is visible |
| Watch for third-party submissions | Art cited against you |
| Consider provisional rights | Notice required to use them |
| Track your own file history | Publicly visible |
| Decide on continuations | Before the parent issues |
Checking a published application
| Step | Source |
|---|---|
| Read the claims | Worst case, not forecast |
| Read the file history | Where claims are heading |
| Check continuity data | Anything pending? |
| Check foreign counterparts | Different claims |
All free in USPTO Patent Center.
Patent application publication: the checklist
- Publication happens at eighteen months from earliest priority, automatically.
- It ends trade secret protection permanently. That decision is irreversible.
- Keep anything you want secret out of the application entirely.
- Abandoning does not restore secrecy. The publication stands.
- Provisional rights are narrow. Notice required, claims must be substantially identical.
- A published application confers no right to exclude.
- Use non-publication requests rarely, and never if you might file abroad.
- Read competitors' file histories. Free, and more informative than the claims.
- Treat published claims as a worst case, not a forecast.
- Accept the eighteen-month blind spot. No amount of searching closes it.