Inter partes review petitions fell to an all-time quarterly low of 57 in the second quarter of 2026. The story is no longer just decline; it is substitution. Ex parte reexamination requests hit an all-time quarterly high of 336 in the same quarter, and reexams now account for roughly three-quarters of all post-grant validity filings. The IPR era’s center of gravity has moved.
The numbers
IPR petitions hit an all-time low. 57 IPR petitions were filed in Q2 2026, down 51.3% from Q1 2026 and 83.8% from Q2 2025. Monthly filings bottomed at 22 in June, versus 131 in January 2025.
Reexam, in sharp contrast, set an all-time high. 336 ex parte reexamination requests were filed in Q2 2026, up 30.2% from Q1 and 148.9% year-over-year, edging past the previous record (330 in Q3 2012, set by a one-time fee rush). Reexams made up 74.7% of post-grant filings in the first half of 2026; in the first half of 2025 that share belonged to IPRs (72.9%). Samsung (61 requests), Apple (28), and Unified Patents (25) led the surge.
The institution rate is recovering, on a thin base. The IPR institution grant rate reached 47.4% in June 2026, up from 21.5% in January 2026 and an all-time monthly low of 19.4% in August 2025. The trailing 12-month rates remain low: 27.2% for petitions against NPE-owned patents, 41.4% against operating companies.
A notable reversal within the quarter: the institution rate against NPEs (47.0%) rose above the rate against operating companies (35.6%) in Q2, the first flip since the discretionary-denial surge began.
Procedural denials fell with volume. PTAB procedural (discretionary) denials dropped from 156 in Q1 to 50 in Q2, a 67.9% decline, consistent with far fewer petitions reaching an institution decision.
The backlog kept shrinking. Pending PTAB matters fell to 696 at the end of June 2026, just 36.2% of the November 2018 peak of 1,923 (and down from roughly 815 in March).
District court held steady. 1,007 new patent cases were filed in Q2 2026, in line with recent quarterly volumes. NPEs filed 48.6% of Q2 district court cases, the first quarter below 50% since Q1 2018.
What’s driving the numbers
The collapse in IPR filings is the demand response to the discretionary-denial regime built over the past year: Director-level control of institution decisions, the October 2025 proposed rules codifying discretionary-denial practice, and the Federal Circuit’s decision in In re Motorola Solutions, Inc., which insulated the Director’s discretionary institution authority from judicial oversight. When institution odds fell into the 20s%, petitioners stopped filing and took their invalidity arguments to the Central Reexamination Unit instead. The same operating companies that historically led PTAB filings (Samsung, Apple, Cisco, Intel, Google) now top the reexamination rankings.
Two second-quarter developments cut the other way and are worth noting. First, the USPTO’s Tesla, Inc. v. Bulletproof Property Management, LLC decision (designated informative June 15, 2026) declined discretionary denial across seven IPR petitions with actual reasoning: no trial date set in the parallel case, a Sotera stipulation, prior art the examiner effectively overlooked, patents too young for “settled expectations,” and evidence of U.S.-based manufacturing. It is the clearest institution roadmap petitioners have been given since the surge began. Second, the USPTO’s April 2026 “pre-order” procedure now gives patent owners a fee-free, 30-page opportunity to oppose a reexam request before a substantial-new-question determination; patent owners used it in roughly 42% of eligible requests, and monthly reexam filings retreated from 136 in May to 79 in June. Whether that dampens the reexam boom is the trend to watch in Q3.
What it means for companies
Validity fights have moved, not disappeared. District court filing volume was unaffected by the PTAB’s contraction: 1,007 new cases in Q2. Invalidity is increasingly resolved in front of a district judge and jury, which is slower and more expensive, and puts fee-shifting under 35 U.S.C. § 285 (the Octane Fitness standard) squarely back in the cost-benefit analysis for both sides.
Reexamination is the new default challenge, with real limits. It is ex parte (the requester exits after filing), it is slower, and it lacks IPR’s estoppel-driven finality. The pre-order procedure now gives patent owners a first shot at stopping it. Companies substituting reexam for IPR should price those differences in.
NPE dynamics are shifting at the margins. NPE district court share slipped below 50% for the first time since early 2018, and Q2 institution rates ran higher against NPE-owned patents than operating-company patents. The discretionary-denial regime that emboldened NPE filing through 2025 may be recalibrating; it is too early to call it a turn.
Timing and forum strategy carry more weight than ever. An accused infringer now has three paths (IPR, reexam, district court invalidity) with sharply different odds, speeds, and estoppel consequences. That choice has to be modeled at the demand-letter stage, not after service.
How to read it (the part the headlines miss)
“IPR is dead” remains an overstatement, and Q2 gave the counterpoint: a 47.4% June institution rate and an informative Director decision explaining, factor by factor, how a petition survives discretionary review. What died is the reflexive IPR. The tool still works when a petition clears institution; the Tesla v. Bulletproof roadmap (early filing, a broad stipulation, overlooked art, young patents) shows what clearing it now requires. The practical question for a defendant is no longer “should we file an IPR?” but “which validity forum, on which schedule, gives this case its best odds?” For the first time since the AIA, the honest answer is often not the
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
[View Source]