Can you patent something and make it free? Yes, and the reasons for doing so are more practical than idealistic.
A patent is a right to exclude, not an obligation to. Nothing requires an owner to enforce, charge, or restrict. You can grant the world a royalty-free licence the day the patent issues.
The question worth asking first is what you are actually defending against. Stopping someone else patenting the invention and stopping someone suing you are different problems with different cheapest solutions.
The routes, compared
| Route | Stops others patenting it | Gives you a right to exclude | Cost | Counterclaim value |
|---|---|---|---|---|
| Defensive publication | Yes | No | Very low | None |
| Patent + royalty-free licence | Yes | Yes, unexercised | High | Yes |
| Patent + public pledge | Yes | Yes, limited by the pledge | High | Partial |
| Patent + defensive pool | Yes | Pooled | High | Pooled |
| Patent, then abandon | Until publication | No, after lapse | Moderate | None |
| Do nothing | No | No | None | None |
Doing nothing is the risky option. If you neither publish nor file, someone else may independently develop and patent the same invention, and then assert it against you. Prior use defences exist but are narrow and hard to prove.
Defensive publication is the cheapest adequate answer to that risk — once the invention is in the prior art, nobody can validly patent it.
But publication gives you nothing to fight back with, which is the whole argument for patenting instead.
Why "free" needs a mechanism
Intending an invention to be free does not make it free. Without a positive step, a third party can patent the same idea and assert it against everyone — including you.
| What you do | Can someone else patent it? |
|---|---|
| Nothing | Yes |
| Tell people informally | Possibly — if not provably public and dated |
| Publish, dated and accessible | No |
| Patent it yourself | No |
| Patent, then abandon | Not after publication |
"We told everyone at a conference" may or may not be enough. Prior art must be publicly accessible and datable, and an undocumented talk is harder to rely on than a published paper.
This is the whole argument for a deliberate mechanism. Good intentions leave the invention available to be captured; a dated publication or a filed application does not.
Defensive publication
Publishing an invention puts it into the prior art, and prior art defeats novelty for everyone who comes after.
| Method | Effect |
|---|---|
| Journal or conference paper | Prior art from publication |
| Technical disclosure service | Prior art, indexed and dated |
| Published patent application | Prior art, and preserves the option to prosecute |
| Company website or blog | Prior art if genuinely public and dated |
| Product release | Prior art through public use or sale |
Filing an application and letting it publish is the hybrid route. It becomes prior art at eighteen months whether or not you prosecute it, and you keep the option to pursue claims if circumstances change.
Dating and accessibility matter. Prior art must be publicly available; a disclosure nobody could find may be argued not to qualify. Indexed, dated services exist precisely to remove that argument.
Publication destroys your own foreign rights too. Most countries have no grace period, so publishing before filing forecloses patenting the same invention abroad — permanently and for you as well as everyone else.
Open patent licences
A royalty-free licence is a contract, not an abandonment. You keep ownership, and the terms do work.
| Term | Typical open licence position |
|---|---|
| Royalty | Zero |
| Scope | Sometimes limited to a field or a standard |
| Territory | Usually worldwide |
| Duration | Life of the patent |
| Defensive termination | Rights end if the licensee sues you |
| Sublicensing | Often permitted |
| Warranty | None |
Defensive termination is the clause that makes open licensing strategic rather than charitable. Anyone may use the technology freely, but asserting patents against you costs them their own permission to use it. That converts a giveaway into mutual deterrence.
The licence binds a later buyer of the patent. If you sell the patent, the buyer takes it subject to existing licences — which is what makes an open licence durable rather than reversible. See intellectual property licensing.
What free does not mean
A royalty-free licence is not the same as no restrictions, and the difference matters to anyone relying on it.
| Assumption | Reality |
|---|---|
| "Free means I can do anything" | Scope, field and territory limits usually apply |
| "The licence cannot be withdrawn" | Depends on the terms; pledges are usually irrevocable |
| "It covers all their patents" | Usually only specified ones |
| "It covers future patents" | Rarely, unless expressly stated |
| "No conditions attached" | Defensive termination is common |
| "It protects me from everyone" | Only from this owner |
The last row is the one that catches adopters. A royalty-free licence from one company says nothing about a third party's patent covering the same technology. Freedom to operate is a separate question. See freedom to operate.
Read what is actually covered. A pledge naming specific patents covers those patents; new filings on improvements are usually outside it unless stated.
Pledges and pools
| Mechanism | How it works |
|---|---|
| Public pledge | An irrevocable published commitment not to assert specified patents |
| Standards commitment | FRAND or royalty-free undertaking through a standards body |
| Defensive pool | Members contribute or the pool acquires patents; members get licences |
| Cross-licence network | Members agree not to assert against each other |
Pledges are generally treated as binding through promissory estoppel, because parties relied on them in building products. A pledge that could be withdrawn would give no one confidence to rely on it, and the enforceability follows from that reliance.
Scope precision matters more in a pledge than in a licence, because you cannot negotiate individually afterwards. A pledge covering "our patents relating to X" will be read against you at the edges.
Standards commitments travel with the patent. A FRAND undertaking made through standards body membership binds a later purchaser, which is why buyers check for them in diligence. See patent royalty rates.
Worked example: three answers to one situation
A company develops a data format it wants the industry to adopt. Adoption requires that nobody fears being sued.
Option A — defensive publication
| Cost | Very low |
| Time to effect | Immediate |
| Stops others patenting the format | Yes |
| Stops a third party asserting an earlier patent | No |
| Counterclaim material if sued | None |
| Foreign patent rights | Destroyed |
Cheapest and fastest. It guarantees the format stays unpatentable and gives adopters that assurance.
It does nothing about a patent that already exists, and it leaves the company with no defensive assets at all.
Option B — patent plus open licence
| Cost | $8,000–$20,000 plus $14,470 in fees |
| Time to effect | 2–4 years to grant |
| Stops others patenting the format | Yes |
| Adopter assurance | Strong — a published licence they can read |
| Counterclaim material if sued | Yes |
| Defensive termination available | Yes |
Most expensive and most protective. Adopters get a licence with terms rather than a promise, and the company gets an asset it can counterclaim with.
The maintenance fees are a permanent cost on a patent generating no revenue, which is a real consideration over twenty years.
Option C — publish now, file too
| Cost | Filing cost, decision deferred |
| Effect at 18 months | Publication makes it prior art |
| Option preserved | Prosecute if the landscape changes |
| Adopter assurance | Moderate until the position is announced |
The pragmatic middle. Filing then publishing achieves the prior art effect automatically at eighteen months, while keeping the option to obtain claims if a competitor turns hostile.
Announce the intention clearly if adoption is the goal, because a pending application with no stated position looks like a threat rather than a gift.
Prior user rights, and why they are not enough
If you neither file nor publish, you rely on a narrow statutory defence.
| Prior user rights under 35 U.S.C. 273 | |
|---|---|
| Protects | Commercial use begun at least 1 year before the patent's filing or disclosure |
| Scope | Your use only — not a general licence |
| Transferable | Only with the business, in narrow circumstances |
| Burden | Clear and convincing evidence, on you |
| Covers expansion | Limited |
It is a defence, not a right. You can keep doing what you were already doing, and proving it requires contemporaneous documented evidence of commercial use from more than a year before the other party filed.
It does not stop the patent being asserted against your customers or suppliers, and it gives you nothing to counterclaim with.
Which is the practical argument for publishing at minimum. A defensive publication costs very little and prevents the patent existing at all, rather than requiring you to defend against it with evidence you may not have kept.
The cost of keeping a free patent alive
| Fee | Due | Large | Small (40%) | Micro (20%) |
|---|---|---|---|---|
| First | 3.5 yrs after grant | $2,150 | $860 | $430 |
| Second | 7.5 yrs | $4,040 | $1,616 | $808 |
| Third | 11.5 yrs | $8,280 | $3,312 | $1,656 |
| Total | $14,470 | $5,788 | $2,894 |
A freely licensed patent generates no revenue and costs the same as any other. Which means someone has to keep deciding it is worth paying for.
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
Defensive patents are among the first to lapse when budgets tighten, because they have no revenue to defend them at the third fee. If the commitment matters, budget for all three fees at the outset rather than deciding each time.
Lapse is not the same as a pledge. Once the patent expires it can no longer be used defensively, and the counterclaim value disappears with it.
Worked numbers: publication versus patent
A company deciding how to protect a format it wants adopted.
| Defensive publication | Patent + open licence | |
|---|---|---|
| Preparation | $500–$2,000 | $8,000–$20,000 |
| USPTO fees | — | ~$2,000 large entity |
| Time to effect | Immediate | 2–4 years |
| Maintenance fees | None | $14,470 over the life |
| 20-year total | ~$2,000 | ~$35,000 |
| Blocks others patenting it | Yes | Yes |
| Counterclaim asset | None | Yes |
| Foreign rights | Destroyed | Preserved if filed first |
Seventeen times the cost, for a counterclaim asset and preserved foreign options. Whether that is worth it depends entirely on litigation risk in the field.
In a heavily litigated area the patent is cheap insurance. In a quiet one the publication does the whole job for two thousand dollars.
The middle route — file, publish, decide later — costs the filing but defers the maintenance commitment, since a patent can be allowed to lapse at the first fee once the prior art effect has been achieved.
Choosing between them
| Your goal | Best route |
|---|---|
| Stop anyone patenting this | Defensive publication — cheapest and immediate |
| Give adopters confidence | Patent + published open licence |
| Deter suits against you | Patent + defensive termination |
| Support a standard | Standards commitment, royalty-free |
| Cheapest possible | Defensive publication |
| Keep foreign options open | File first, publish after |
The order matters if foreign rights are in play. Publishing before filing destroys them everywhere without a grace period. Filing first, then publishing, keeps every option open at modest extra cost. See prior art.
Announcing the position
An open commitment only works if people can find it and rely on it.
| Element | Why |
|---|---|
| A stable public page | Adopters need somewhere to point |
| Patent numbers listed | Scope must be identifiable |
| The licence text itself | Not just a summary |
| Whether future patents are included | Usually they are not |
| Defensive termination terms | Adopters need to know the condition |
| A contact for questions | Signals it is maintained |
Publish the licence text, not an intention. "We do not intend to enforce" is weaker than a granted licence, because reliance on a stated intention is harder to establish than reliance on a contract.
Keep the page maintained. A pledge page last updated years ago, listing patents that have since lapsed, undermines confidence in the commitment rather than supporting it.
Patenting something and making it free: the checklist
- Decide which risk you are managing — others patenting it, or others suing you. They have different cheapest answers.
- If the goal is only to block patenting, publish. It is immediate and costs very little.
- File before publishing if foreign rights might matter. Publication destroys them where there is no grace period.
- Use a dated, indexed publication route so the disclosure's availability cannot be disputed.
- If you want counterclaim material, file. A published disclosure gives you nothing to assert.
- Publish the licence terms, not just an intention. Adopters need something they can read.
- Include defensive termination so the licence ends for anyone who sues you.
- Define pledge scope precisely. You cannot renegotiate a pledge afterwards.
- Budget all three maintenance fees at the outset — a free patent is a twenty-year cost with no revenue.
- Remember the licence binds a later buyer, which is what makes the commitment durable if the patent is ever sold.