Patent trolls

Patent trolls in 2025: the year in numbers

2025 was a banner year for patent trolls — and not by accident. Filings surged while the defendant's cheapest weapon, the PTAB, was quietly taken off the table. Here's what the reporting shows, and why the two are connected.

10 min read · Updated Jun 8, 2026

If you defend against patent trolls, 2025 was a bad year — and the data explains exactly why. Patent litigation surged, non-practicing entities drove almost all of the increase, and the single cheapest tool a defendant has for killing a bad patent — challenging it at the USPTO's Patent Trial and Appeal Board — became dramatically harder to use. Those two facts are not a coincidence. This is the year in numbers, drawn from public reporting and cited throughout.

The headline: litigation surged, NPEs drove it

Patent litigation rose sharply in 2025. Filings ran well ahead of the prior year's pace, with the increase driven almost entirely by non-practicing entities — analysts pegged the NPE-driven rise at roughly 15–20% over 2024 (Unified Patents data, as reported by Maynard Nexsen).

By the numbers, through the first nine months of 2025 (Unified Patents, Q3 2025 in Review):

  • NPEs filed 56.4% of all district-court patent cases — a clear majority — versus 39.7% for operating companies.
  • In the high-tech sector, the NPE share was 91.4%. Nearly every high-tech patent suit was brought by an entity that makes nothing.
  • Eight of the top ten district-court patent plaintiffs were NPEs, including entities filing under names like Torus, HyperQuery LLC, WirelessWerx IP, and Linfo IP — the kind of shell-style naming that is itself a tell.

This is the volume-based model operating at full throttle: a majority of the federal patent docket is now non-practicing entities asserting patents they bought rather than built.

The Eastern District of Texas is back

For a few years after TC Heartland tightened venue rules, troll filings spread out — heavily to the Western District of Texas (Waco) and Delaware. In 2025 the Eastern District of Texas reclaimed the throne (Unified Patents, Q3 2025):

Venue Share of all patent cases Share of NPE filings
Eastern District of Texas 28.3% 45.4%
Western District of Texas 11.3% 17.4%
District of Delaware 11.0% 8.0%

Together, the two Texas districts accounted for 40% of all patent cases and 62.8% of all NPE cases in the first nine months of 2025. The pull factors are familiar: speed, predictable procedure, and plaintiff-friendly juries. When Waco changed its case-assignment rules to break up a single judge's docket, plaintiffs simply drifted back east. The forum-shopping that TC Heartland was supposed to curb never went away — it just re-concentrated.

The real story of 2025: the PTAB clampdown

Here is what made 2025 different, and why it matters more than the filing numbers. The defendant's most powerful low-cost weapon is inter partes review — a USPTO proceeding that can invalidate a bad patent for a fraction of the cost of litigating it in court. In 2025, the USPTO made that weapon far harder to fire.

The mechanism is "discretionary denial" — the Board declining to even consider a petition's merits, usually because parallel district-court litigation is moving quickly. The timeline:

  • Late February 2025: Acting Director Coke Morgan Stewart rescinded the 2022 Vidal guidance that had limited discretionary denials (the guidance that protected petitions with compelling merits, Sotera stipulations, and parallel ITC cases). (USPTO; IPWatchdog)
  • March 2025: Stewart issued a new interim process that bifurcated institution — the Director, with PTAB judges, would separately decide whether to discretionarily deny — and expanded the grounds, including a patent's "settled expectations" based on how long it had been in force. (WilmerHale)

The effect on institution rates was immediate and severe. Within calendar year 2025 (RPX):

Calendar quarter PTAB institution rate
Q1 2025 68%
Q2 2025 42%
Q3 2025 23%
Q4 2025 41% overall — but just 14% under the Director's new process

By October 2025, RPX reports, Director John Squires had taken personal control of the institution decision and instituted only 27 of 196 petitions in Q4 — a 14% rate, often with no written explanation. For the full fiscal year (FY2025), the USPTO's official institution rate landed around 50%, a five-year low, even though petition filings actually ticked up (1,433 petitions, 145 more than FY2024). (USPTO PTAB Trial Statistics) The petitions kept coming; the Board just stopped instituting them.

Why the two trends are one trend

Put the pieces together and the surge in troll litigation isn't a mystery — it's a predictable response to the PTAB clampdown:

flowchart TD
  A["USPTO expands<br/>discretionary denials<br/>(Feb–Mar 2025)"] --> B["IPR institution rate<br/>collapses<br/>68% → ~14–23%"]
  B --> C["Fewer IPRs instituted<br/>= fewer district-court<br/>stays granted"]
  C --> D["Defendants lose the<br/>cheap, fast off-ramp"]
  D --> E["NPEs file more suits<br/>NPEs = 56% of the docket"]
  E --> F["Settlement leverage<br/>shifts back to the troll"]

  style A fill:#fff7ed,stroke:#c2410c,color:#1c1917
  style B fill:#fee2e2,stroke:#be123c,color:#1c1917
  style E fill:#fff7ed,stroke:#c2410c,color:#1c1917
  style F fill:#fee2e2,stroke:#be123c,color:#1c1917

An IPR is most valuable not just because it can kill a patent, but because filing one often wins a stay of the parallel district-court case — pausing the expensive litigation machine while the USPTO decides. When the PTAB stops instituting, district courts stop staying, and the troll's cost-pressure playbook runs unopposed. Commentators reported PTAB institution rates as low as 10.3% over a recent nine-month window by late 2025 (IPLaw360, via Maynard Nexsen). NPEs noticed, and filed accordingly.

The 2024 record, for context

The escalation didn't start in 2025. The Lex Machina 2025 Patent Litigation Report — which covers calendar year 2024 — documented the run-up (LexisNexis / Lex Machina):

  • A 22% surge in U.S. patent case filings over 2023.
  • A record $4.3 billion in patent damages awarded.
  • The Eastern District of Texas alone saw over 1,000 new patent lawsuits — more than twice the next-busiest district.
  • A 35% jump in design-patent suits and a 20%+ rise in ANDA (pharma) litigation.

So 2025's NPE surge built on a 2024 that was already a multi-year high. The trend line points up.

Reform is pending — and stuck

Three bipartisan bills aimed at the patent system were live in the 119th Congress in 2025, all of which would, on balance, strengthen patent-holder rights:

  • PERA (Patent Eligibility Restoration Act of 2025, S.1546 / H.R.3152, introduced May 1, 2025) would abolish the judicial exceptions to patent eligibility — effectively overruling Alice and Mayo, the §101 doctrines defendants rely on to knock out abstract software and diagnostic patents.
  • PREVAIL Act (S.480 / H.R.1574, reintroduced February 2025) would tighten who can file an IPR and add standing and estoppel restrictions — making the PTAB harder for defendants to use, on top of the administrative clampdown already underway.
  • RESTORE Patent Rights Act (reintroduced February 2025) would restore a presumption of injunctive relief, partially undoing eBay v. MercExchange — handing NPEs back the injunction threat that made them so dangerous before 2006.

As of mid-2026, none had advanced beyond committee. But the direction of all three, combined with the USPTO's 2025 administrative moves, points the same way: a harder environment for defendants.

One more 2025 marker for the record: AIPLA released its 2025 Report of the Economic Survey in February 2026, the latest installment of the biennial cost data this site relies on for the litigation and prosecution cost articles.

What it means if you're a defendant

The 2025 data changes the defensive calculus in concrete ways:

  • Don't assume an IPR stay. The reflexive 2018-era move — file an IPR, win a stay, pressure a cheap settlement — is much less reliable now. Model the case on the assumption that the PTAB may discretionarily deny, and have a district-court plan that stands on its own.
  • Petition quality and timing matter more than ever. With the Director personally gatekeeping institution, a petition has to be filed early and present compelling merits to have a shot. Marginal petitions that might have squeaked through in 2021 now get denied.
  • Non-PTAB defenses get more weight. Early non-infringement reads, § 101 eligibility motions in district court (still good law, for now), and § 285 fee-shifting threats carry more of the load when the IPR off-ramp narrows.
  • Pre-built invalidity is worth more. When the cheap administrative path to invalidation is throttled, the value of having §102 / §103 analysis ready before you're sued goes up, not down. That's the entire premise of this database.

Bottom line

2025 was, by the numbers, a banner year for patent trolls. NPEs filed a majority of all district-court patent cases (56.4%) and the overwhelming share of high-tech suits (91.4%); the Eastern District of Texas reabsorbed the docket; and the USPTO's expansion of discretionary denials drove PTAB institution rates from 68% to as low as ~14% within the year, pulling the defendant's cheapest weapon off the table. Filings rose because the cost of not fighting back is, once again, the only number the troll needs.

The reforms that might change this — at the USPTO or in Congress — currently tilt the other way. For now, the asymmetry this site exists to fight got a little wider in 2025. Understanding the data is the first step to closing it.

Sources

This article reflects public reporting available as of mid-2026 and is for general education; it is not legal advice. Litigation statistics and USPTO policy are moving quickly — verify the current figures against the linked primary sources before relying on them.