Patent invalidation: the routes and what each costs

Patent invalidation is showing that claims should never have issued. It is distinct from non-infringement, which accepts the patent and denies that your product falls within it, and the two are usually pleaded together.

There are four routes, and choosing between them is the first substantive decision — because each permits different grounds, applies a different burden, and costs different amounts in both money and forfeited arguments.

The four routes compared

The burden difference is the headline. An issued patent is presumed valid under 35 U.S.C. 282, and a district court requires clear and convincing evidence to overcome that presumption. The PTAB applies a preponderance standard — more likely than not. The same prior art can fail in court and succeed at the Board.

Which grounds each route allows

This determines the forum more than anything else.

IPR and ex parte reexamination are limited to documentary prior art. Under 35 U.S.C. 311(b), an IPR may only be based on patents and printed publications. A devastating public-use case, or a strong section 101 argument, simply cannot be raised there.

Post-grant review is the broad option and the window is narrow. Available only within nine months of grant for AIA patents, but it permits almost any invalidity ground including eligibility and written description. Companies monitoring competitor grants use it precisely because of that breadth.

Inequitable conduct is court-only and renders the patent unenforceable rather than invalid — a different remedy with a much higher pleading standard.

Inter partes review

The most-used route, and the subject of a substantial shift in recent years.

Institution rates have fallen sharply, from roughly 65% in October 2024 to about 37% by February 2026, driven principally by changes to discretionary denial practice. A meritorious petition can now be denied because parallel district court litigation is likely to reach validity first.

The one-year bar is absolute. Under 35 U.S.C. 315(b), a petition must be filed within one year of service of an infringement complaint, and the bar extends to real parties in interest and privies.

Estoppel is the real cost. Under 315(e), a petitioner who receives a final written decision cannot later raise in court or at the ITC any ground raised or that reasonably could have been raised. A failed IPR removes documentary prior art defences from the parallel case.

See inter partes review for the full process, timeline and fee structure.

Ex parte reexamination

The underused route, and often the right one.

Anyone can request it, at any time during the patent's enforceable life, based on prior art patents and printed publications raising a substantial new question of patentability.

The Director decides within three months whether to order reexamination. If ordered, the proceeding runs before an examiner with special dispatch.

The requester's lack of participation is the central trade. After filing, the proceeding is between the examiner and the patent owner. You cannot respond to the owner's arguments, cannot appeal, and cannot prevent amendments that preserve the claims in narrower form.

No estoppel is the compensating advantage. A failed reexamination costs the fee and forecloses nothing, which makes it a substantially lower-risk option than an IPR.

Anonymity matters commercially. A request filed through counsel without naming the real party in interest lets a company challenge a patent without signalling interest — useful where the challenger is negotiating, or does not want to invite an assertion.

The risk of strengthening the patent is real. A patent that survives reexamination over the best art has been tested and is worth more. That cuts both ways depending on whether you expect to face it again.

District court

The route with everything available and the highest burden.

Invalidity is pleaded as a defence and usually as a counterclaim for a declaratory judgment, which keeps it alive even if the patentee dismisses.

Clear and convincing evidence is required, because of the statutory presumption of validity. This is a meaningfully higher bar than the PTAB's preponderance standard, and it is the reason challengers with documentary art prefer the Board.

All grounds are available, including the ones the PTAB cannot reach — public use, on-sale activity, section 101 and section 112. Where your best case is one of those, court is not a choice but a necessity.

Stays are common. District courts frequently stay a case pending IPR, particularly where the petition was filed early. That is often the actual purpose of petitioning.

A worked decision

A defendant served with a complaint. The prior art search produces:

Route analysis

The public demonstration is the best evidence and it cannot go to the Board. That single fact reshapes the decision.

Filing an IPR on the documentary art alone risks the worst outcome: institution, a final written decision against the petitioner, and estoppel foreclosing the German application in the district court case — where it would have supported the invalidity defence alongside the public use.

Ex parte reexamination is the interesting middle. It puts the German application in front of an examiner cheaply, carries no estoppel, and preserves everything for court. The downside is no participation and the possibility of narrowing amendments that preserve the claims.

A defensible strategy here: file the reexamination request on the German application, plead all grounds in court including the public use and section 101, and reserve the IPR decision until the reexamination outcome is visible — subject to the one-year bar, which must be diarised from service regardless.

What happens to invalidated claims

Cancelled claims cease to exist. The USPTO issues a certificate and the claim is gone.

Claims stand or fall independently. A patent with twenty claims may lose six and keep fourteen, and the survivors remain fully enforceable. This is why challengers target the asserted claims rather than the whole patent — the fee structure also escalates sharply beyond twenty claims.

Cancellation is effective against the world, and a patent with claims cancelled is worth correspondingly less on any patent valuation. A successful challenge benefits every party exposed to the patent, not just the challenger. That produces a collective action problem: everyone benefits, so nobody wants to pay. It is also why defensive aggregators and industry groups fund petitions.

Amendments during the proceeding. In reexamination the owner may amend claims, potentially preserving them in narrower form. Narrowed claims are subject to intervening rights, which protect anyone who was practising the original scope.

Invalidation versus design-around

Often the better question than which forum.

Design-around wins where the claim is narrow and the feature is peripheral. Changing an implementation detail to fall outside a claim is far cheaper than proving the claim should never have issued.

Invalidation wins where the claim is broad enough that any commercially sensible product infringes, where the patent will be asserted repeatedly, or where several parties share the exposure and can share the cost.

The analysis starts with a claim chart either way — you cannot assess whether a design-around is feasible without knowing precisely which limitation you need to avoid.

Before you commit to a route

1. Identify the asserted claims and target only those.
2. Establish whether your best grounds are documentary — that decides whether the Board is available at all.
3. Diarise the one-year IPR bar from service, and confirm no privy was served earlier.
4. Read the file history to see what the examiner considered, because references already of record are weaker.
5. Model the estoppel exposure before filing an IPR.
6. Check the current institution rate, not a figure from two years ago.
7. Cost the design-around alternative properly before spending on invalidation.
8. Consider ex parte reexamination where anonymity or estoppel-avoidance matters more than participation.