Estoppel Gutted: A Pelican’s Guide to Patent Litigation

by Dennis Crouch

In a significant decision narrowing the scope of Inter Partes Review (IPR) estoppel, the Federal Circuit has held that a patent challenger who previously pursued an IPR can still rely on system prior art in district court litigation, even when that system was fully documented in publications that could have been raised during the IPR.  A key implication from this case is that despite the estoppel provision, a printed publication from an IPR can be later used in the litigation as part of the same combination of references used in the IPR, so long as the publications are categorized by the patent challenger in the litigation as proving that the invention was known/used/on-sale. Ingenico Inc. v. IOENGINE, LLC, No. 23-1367 (Fed. Cir. May 7, 2025). Writing for a unanimous panel, Judge Hughes explained:

IPR estoppel does not preclude a petitioner from relying on the same patents and printed publications as evidence in asserting a ground that could not be raised during the IPR, such as that the claimed invention was known or used by others, on sale, or in public use.

The decision resolves a long-standing split among district courts regarding the proper interpretation of the term “ground” in 35 U.S.C. § 315(e)(2), which bars IPR petitioners from asserting in district court “any ground that the petitioner raised or reasonably could have raised during that inter partes review.” The Federal Circuit held that “ground” refers to the specific statutory basis for invalidity (e.g., anticipation or obviousness based on patents or printed publications), not the prior art references themselves.  The Federal Circuit clarified that ‘ground’ refers specifically to the statutory bases available in an IPR (anticipation or obviousness via patents/printed publications). As such, petitioners remain free to use identical prior art evidence to support district court invalidity grounds not available in IPR—such as prior public use or sale. The court explained:

[A] ground is not the prior art asserted during an IPR. In drafting 35 U.S.C. § 315(e)(2), Congress could have precluded petitioners from asserting in district court that the claim is invalid on any prior art that the petitioner raised or reasonably could have raised during that inter partes review, but Congress chose not to. Instead, Congress precluded petitioners from asserting grounds in district court. Thus, IPR estoppel does not preclude a petitioner from asserting the same prior art raised in an IPR in district court, but rather precludes a petitioner from asserting grounds that were raised or reasonably could have been raised during an IPR. The only anticipation and obviousness challenges that a petitioner can make during an IPR are that the claims were patented or described in a printed publication. IPR estoppel precludes these challenges in district court if they were raised or reasonably could have been raised during the IPR. But IPR estoppel does not preclude a petitioner from relying on the same patents and printed publications as evidence in asserting a ground that could not be raised during the IPR, such as that the claimed invention was known or used by others, on sale, or in public use.

Although the court’s statutory interpretation does not explicitly acknowledge broader policy implications, the practical effect significantly undermines the effectiveness of IPR estoppel, since patenting or publication typically provides virtually irrefutable evidence that an invention was “available to the public.”  With the one caveat here  being 102(a)(2) submarine prior art.

This interpretation creates a significant loophole in IPR estoppel. Consider a common scenario: a challenger files an IPR asserting that claims are obvious based on a combination of printed publications. After losing at the PTAB, under the Ingenico ruling, that same challenger can proceed to district court and argue the invention was obvious based based upon being “known or used by others” before the critical date—using those very same publications as evidence that the invention was publicly known. The challenger simply needs to recharacterize the invalidity theory from “described in printed publications” (a ground available in IPR) to “known to the public” (a ground unavailable in IPR), while relying on identical prior art evidence. This formalistic distinction between “grounds” and “evidence” effectively allows petitioners to relitigate the same substantive invalidity challenges that Congress intended to bar through § 315(e)(2). The court’s interpretation treats IPR estoppel as merely forcing challengers to change their legal theory while permitting them to recycle the same technical arguments and evidence.

[Note here that this particular case involved a pre-AIA patent and so the court was not faced with the of 102(a)(a)’s “available to the public” novelty.  This is a potential way to cabin-in the decision, but I expect it is an unlikely win.]

= = =

Ingenico originally challenged IOENGINE’s patents in IPR proceedings but had not raised any ground relating to a prior art system known as the DiskOnKey Firmware Upgrader. Although the petition was granted, some of IOENGINE’s patent claims survived. Under § 315(e)(2), once a final written decision issues in the IPR, the petitioner is then estopped from asserting “that the claim is invalid on any ground that the petitioner raised or reasonably could have raised during that inter partes review.”

Here, petitioner could have but did not asserted grounds related to the DiskOnKey Firmware Upgrader — in particular, there were several associated printed publications, including a Readme user guide that included substantial functionality detail, documentation related to the DiskOnKey Upgrade software, and Screenshots showing the Firmware Upgrader’s operation.  In this case, the Firmware Upgrader software itself was not seen as a printed publication, but rather “system art.”

After failing in the IPR, Ingenico argued in the district court litigation that the claims are invalid based upon the DiskOnKey System, including its Firmware Upgrader feature.  As part of its argument, the patent challenger relied heavily on the documentation cited above as well as the system itself — arguing that those showed that the invention was on sale prior to IOENGINE’s invention.   The district court (CAFC Judge Bryson sitting by designation) allowed the evidence and the jury found the claims invalid. On appeal, the patentee unsuccessfully argued this evidence should have been barred as estopped under 315(e).  The patentee argued that the system itself should also be barred because its disclosure was functionally equivalent to the barred printed publications.  As discussed above, the appellate panel rejected these arguments.

The court arguably went further than necessary by broadly interpreting ‘ground,’ potentially limiting the practical reach of § 315(e)(2) in many scenarios.  Rather, the documents could have been allowed in because as part of a combination of references (including the actual product) that collectively showed that the invention was on sale. That ground as a whole could not have been brought because some of the prior art provided was not available in an IPR. However, the court’s decision went much further, seeming to so narrowly interpret the estoppel provision as to eliminate its purpose for most situations. In a later holding, a court could pull back from the more expansive statements by noting their dicta status.

A second caveat to my analysis of the case requires is the untested (but I believe very reasonable) conclusion that a “patent” or “printed publication” will almost certainly fit in the statutory category of being “otherwise available to the public.” But, that quoted provision was added to Section 102 as part of the America Invents Act of 2011 and has not been fully interpreted by the courts.  Still, I generally think of the new provision as broader than the “known or used by others” provision of pre-AIA 102.

A final point worth noting is the broader policy implication of the Federal Circuit’s interpretation, which potentially invites strategic gamesmanship by patent challengers. Although the court’s textual analysis here is grammatically persuasive, it does not appear to fully consider congressional intent behind the estoppel provision—to prevent duplicative litigation and encourage efficient resolution of patent disputes at the PTAB. Moving forward, one immediate solution could be for the USPTO to require Sotero+ style stipulations before granting IPR review.  I also expect that we will quickly see additional litigation on this front as failed IPR petitioners seek to present their “new” grounds (on old prior art) to the courts.

The panel here included Judges Hughes, Prost, and Dyk. Kerry Timbers (Sunstein LLP) argued for the patent challenger and was joined on the brief by Kevin Mosier.  Noah Liebowitz (Dechert LLP) argued for the patentee, and was joined on the brief by Gregory Chuebon, Michael Fisher, and Michael Joshi.

One more final paragraph – this one about the Judges.  The panel of judges on this case were particularly bad for the patentee. Each of the three judges has a history of IPR decisions that tend to disadvantage patent owners. Judge Dyk has developed a reputation for reversing pro-patentee PTAB determinations.  Recently in Valve Corp. v. Ironburg Inventions Ltd., 2025 WL 1177824 (Fed. Cir. Apr. 23, 2025), he vacated the Board’s non-obviousness finding for applying the wrong legal standard regarding motivation to combine. Similarly, in Honeywell Int’l Inc. v. 3G Licensing, S.A., 124 F.4th 1345 (Fed. Cir. Jan. 2, 2025), Judge Dyk reversed the Board’s patentability determination, criticizing it for failing to recognize that “the motivation to modify a prior art reference to arrive at the claimed invention need not be the same motivation that the patentee had.” Judge Prost has likewise demonstrated a consistent anti-patentee tendency.  One recent case on point is Cytiva BioProcess R&D AB v. JSR Corp., 122 F.4th 876 (Fed. Cir. Dec. 4, 2024). In that case, Judge Prost’s opinion reversed the Board’s determination that certain claims were not unpatentable by applying broad inherency principles, concluding that “recitation of a naturally flowing property of [the otherwise obvious] claims ‘cannot overcome a prima facie case of obviousness.'” She also expanded prior art eligibility in Lynk Labs, Inc. v. Samsung Elecs. Co., 125 F.4th 1120 (Fed. Cir. Jan. 14, 2025), holding that published patent applications could serve as prior art as of their filing date rather than publication date. Judge Hughes, who authored the present opinion, previously invalidated patent claims through expansive claim construction in Samsung Elecs. Co. v. Power2B, Inc., 2025 WL 957287 (Fed. Cir. Mar. 31, 2025), where he found the PTAB had erroneously construed a key claim limitation.