USPTO Acting Director Coke Stewart’s introduction of settled expectations into the discretionary denial calculus for inter partes review (IPR) petitions has created an unexpected hurdle in the institution of IPRs. AD Stewart’s Proxense decision indicated that a petitioner could defend against settled expectations by proving a number of negatives, but it is still uncertain how those elements are weighed and what level of proof is needed for them. To date, the majority of petitions overcoming settled expectations involved claims of examiners making material errors during prosecution of the patent. Until the Proxense defense to discretionary denial is clarified or refined, petitioners should carefully scrutinize the prosecution history to look for material errors to support their IPR petition.
Creating Settled Expectations
Last March, AD Stewart took a direct role under 35 U.S.C. § 314 in first determining whether IPR petitions should be discretionarily denied. AD Stewart’s Interim Process for PTAB Workload Management Memorandum created a process for denying institution of IPR petitions for reasons unrelated to the merits of the petition itself. One of the many listed factors favoring “discretionary denial” of IPR petitions was the concept of “settled expectations of the parties, such as the length of time the claims have been in force.” However, most practitioners were surprised the following June at how this factor was implemented in iRhythm Tech., Inc. v. Welch Allyn, Inc. (IPR2025-00363 Paper 10). AD Stewart found the Petitioner’s failure to seek review earlier had created settled expectations that the patent would remain unchallenged. This factor outweighed previously determinative factors like a post-PTAB decision trial date, lack of previous investment in the litigation, and a high likelihood of the litigation being stayed.
AD Stewart’s initial application of settled expectations suggested the petitioner’s objective knowledge played a role, but that was quickly dismissed with Dabico Airport Sols., Inc. v. AXA Power ApS (IPR2025-00408). There, AD Stewart established that settled expectations apply even when the petitioner had not argued it and without petitioner’s prior knowledge of the patent or of possible infringement (Papers 10 and 21). The burden was also clearly placed on the petitioner to provide “persuasive reasons why the Office should review the challenged patent” to overcome the patent owner’s settled expectations.
AD Stewart also said in Dabico that there was “no bright-line rule on when expectations become settled.” However, a six-year limit seems to be taking shape as long as the continuity has not been broken by a reissuance (IPR2025-00608, Nvidia Corp. v. Neural AI, LLC) or a patent owner has shown a lack of settled expectations in the claims by failing to pay the maintenance fee or disclaiming similar claims in another proceeding (IPR2025-00225, Globus Med., Inc. v. Spinelogik, Inc.; IPR2025-00680, Nintendo Co., Ltd. et al. v. Resonant Sys., Inc.). AD Stewart has also referred older patents because the other patents in the same parallel litigation had already been referred (see IPR2025-00478, Advanced Micro Dev., Inc., et al. v. Concurrent Ventures, LLC, et al., Paper 10) or because there were a large number of technologically diverse patents in the related litigation (see IPR2025-00785, American Airlines, Inc. et al. v. Intellectual Ventures I LLC).
Otherwise, the settled expectation rule seems to create an entire generation of U.S. patents issued from approximately 2005 to 2019 that are presumptively immune to IPR challenges.
But the six-year cutoff is not based on when the patent is asserted, when the IPR petition is filed, or even when all pre-institution IPR briefing is completed. In Kahoot! AS v. Interstellar, Inc., Kahoot!’s opposition to discretionary denial pointed out that the IPR petition had been filed almost four months before the six-year cutoff for settled expectations, but AD Stewart still found that settled expectations necessitated discretionary denial (IPR2025-00696, Paper 9 at 5-6; Paper 12). In Yangtze Memory Tech Co., Ltd. v. Micron Tech., Inc., the petition had been filed and all the briefing on discretionary denial had been completed just before the six-year cutoff of U.S. Patent 10,475,737. Only the AD’s decision to discretionarily deny based on settled expectations fell after the six-year cutoff (IPR2025-00501, Papers 8, 10). It would initially appear that the cutoff is based on the date of the AD’s decision. But another patent included in the AD’s Yangtze (IPR2025-00500, U.S. Patent No. 10,475,737) decision of discretionary denial based on settled expectations had a November 2019 issuance date, putting its six-year cutoff over three months after AD Stewart’s July decision. So, the lack of a bright-line rule for settled expectations persists.
With the recent whiplash of policy changes from director to director, one might assume that settled expectations may disappear if John Squires is confirmed as Director. However, Squires has echoed AD Stewart’s justification for the settled expectation rule, expressing his belief that “it is to the benefit of all stakeholders if prior art is identified and applied at the earliest stage of examination or post issuance” (Questions for the Record, May 28, 2025 at 23). Unless someone effectively opposes the establishment of the settled expectations rule as, for example, a substantive rule in violation of the Administrative Procedure Act’s requirements or “shenanigans” outside of the Director’s statutory limits (see Cuozzo Speed Techs. v. Com. for Intell. Prop., 579 U.S. 261, 275 (2016)), we should assume that settled expectations are here for the term.
AD Stewart has partially justified the six-year cutoff for settled expectations using the six-year limit on infringement damages in 35 U.S.C. § 286. But while Section 286 aims to deter patent owners from sitting on their rights or risk losing out on damages, the settled expectations rule rewards patent owners for waiting until settled expectations have set in before initiating patent litigation (which could then trigger an IPR petition). AD Stewart’s policy in conjunction with Section 286 creates a sweet spot of 5½ to 6 years from patent issuance to initiate litigation in which the patent owner surrenders no damages, but AD Stewart will be unlikely to issue her discretionary denial decision before the patent owner’s expectations have settled.
The Difficulties of the Proxense Defense
In Intel Corp. v. Proxense LLC (IPR2025-00327, Paper 12), AD Stewart offered guidance for overcoming settled expectations, but they have proven to be largely ineffective in practice. Although AD Stewart stated that a patent that issued some time ago but was not “commercialized, asserted, marked, licensed, or otherwise applied in a petitioner’s particular technology space, if at all” may not have settled expectations, such arguments have not fared well. For example, in TSMC Ltd. v. AICP LLC (IPR2025-00682), the Petitioner argued that the patent had been dormant and unasserted until it was purchased to use against TSMC, but AD Stewart found that this “does not defeat Patent Owner’s settled expectations” (Paper 10 at 40-42; Paper 17). In Kahoot!, AD Stewart found that the assertion that the patent had not previously been asserted “does not defeat settled expectations.” In Google LLC v. SoundClear Tech. LLC, Petitioner asserted that SoundClear was a non-practicing entity that had purchased the dormant patent and had never “commercialized, asserted, marked, licensed, or otherwise applied [it] in a petitioner’s particular technology space” (IPR2025-00345, Paper 12 at 2-3). But this did not defeat settled expectations “without more” (Paper 15).
In Murata Manf., Co. Ltd. v. Georgia Tech. Res. Corp., AD Stewart acknowledged that the patent had not been commercialized or any product marked for it, but the mere fact that the Patent Owner was active in the Petitioner’s technology space was enough to overcome the lack of commercialization (IPR2025-00383, Paper 13). AD Stewart also pointed to a discussion of the technology between the parties two years before the patent’s issuance. The implication is that the Proxense defenses are mooted if a petitioner had actual knowledge of the patent or should have known about the patent. In iRhythm, the Petitioner asked AD Stewart to reconsider her first settled expectations decision, explaining that the Patent Owner had allowed the patent to lie dormant and not asserted it until the lawsuit (IPR2025-00363, Paper 11). But AD Stewart denied reconsideration without comment (Id., Paper 13).
Three outliers in the application of the Proxense defenses are Shenzen Tuozhu Tech. Co., Ltd. v. Stratasys, Inc. (IPR2025-00531) and the recent Home Depot U.S.A., Inc. v. H2 Intellect LLC (IPR2025-00480) and Apple Inc. v. Ferid Allani (IPR2025-00856) decisions. In Shenzen, the Petitioner supported its Proxense defense against a ten-year-old patent with discovery showing that the Patent Owner admitted that it did not rely on any practicing products and could not explain why its marking obligations had been met (Paper 10). AD Stewart also noted that the numerous and varied patent technologies in Shenzen’s parallel litigation also indicated that “the Board is better suited to review a large number of patents involving diverse subject matter.” In Home Depot, AD Stewart found that the twelve-year-old patent’s prior assertion against only smartphones, tablets, and watches technology was far outside of Home Depot’s hardware “technology space” (Paper 11). In Apple, the Patent Owner’s admission that it waited to assert the patent until after the patent’s expiration and eleven years after Petitioner denied the need for a license, overcame settled expectations. The early denial of a need for a license in Apple contrasts with the lack of assertion in Murata or the late assertion in DataDome S.A. et al. v. Arkose Labs Holdings, Inc. (IPR2025-00693, Paper 14 at 2-3).
These cases indicate that the Proxense defenses are most effective when supported by patent owner admissions, when the patent technology is far outside of the petitioner’s technology, or where settled expectations were disrupted by early denial of the patent’s value. In Apple and Home Depot, these arguments even overcame the fact that the PTAB’s decisions would issue well after each of the district court’s.
The AD’s Proxense decision also suggests that a petitioner could overcome settled expectations by explaining how “a significant change in law may have occurred” since the patent’s issuance that “directly bears on the patentability of the challenged claims.” This argument also offers a difficult road since it would require a clear and significant change in the law which occurred after the patent issuance and can be shown to have affected the challenged patent’s prosecution. This change would likely have to affect validity under 35 U.S.C. §§ 102 or 103 since those are the only grounds for an IPR petition. There have certainly been some significant changes in novelty/obviousness law between 2005 and 2019. But it is a daunting legal and technical challenge to reverse engineer them into a defense against settled expectations. To date, no petitioner has successfully applied this argument to overcome settled expectations.
Material Error as the Most Successful Strategy
Surprisingly, the most successful method for overcoming settled expectations has been to attack the USPTO by arguing that the patent examiner made a material error in examination of the underlying patent. Findings of material error have overcome settled expectations and led to PTAB referral at least eight times.
Material error by the examiner can be shown where the petition presents the same or substantially the same art or argument previously presented to the Office, but which the Office erred in evaluating in a manner material to the patentability of the challenged claims (IPR2019-01469, Advanced Bionics, LLC v. Med-EL Elektromedizinische Geräte GmbH, Paper 6 at 7-8). Material errors allowing for referral of the petition have taken several forms with the reapplication of previously reviewed prior art being the most interesting. In Anthony Inc. v. ControlTec LLC, AD Stewart referred the IPRs to the PTAB because the patent examiner had overlooked Petitioner’s prior art which had been included in an International Search Report but not included in the U.S. examination (IPR2025-00636, Paper 8 at 16-19, Paper 9). AD Stewart agreed that the examiner’s material error of ignoring that prior art, even though it had not been presented in an IDS, overcame 17-18 years of settled expectations. In Eunsung Global Corp. v. Hydrafacial LLC (IPR2025-00445), the Petitioner successfully argued that the elements alleged to be absent from the prior art by the examiner’s Reasons for Allowance were actually found in the prior art listed in the patent’s References Cited, which the examiner had acknowledged but not fully considered (Paper 14 citing Paper 12 at 15-21). In TSMC Ltd.et al. v. Marlin Semiconductor Ltd. (IPR2025-00847), AD Stewart referred a petition on a fifteen-year-old patent for material error where the examiner had applied prior art to object to patent claims but then allowed amended claims even though the amended elements that were the basis for allowance were found in that prior art (Paper 11). These were all material errors that overcame settled expectations.
This approach turns typical IPR strategy on its head. Conventional wisdom has been to avoid relying on prior art included in the prosecution history because 35 U.S.C. § 325 allows denial of an IPR petition for asserting the same or substantially the same prior art. In fact, in much of the briefing in the above cases the Patent Owners argued that the prior art was the same or cumulative while the Petitioners argued it was not, but offered material error as an alternative if the prior art was found to be cumulative. Ironically, AD Stewart agreed with the Patent Owners and denied their motions for discretionary denial because cumulative or identical prior art with overlooked elements indicated a material error by the examiner. It is unlikely that the cited prior art would have overcome settled expectations had it not previously been cited and resulted in a material error on the part of the examiner.
AD Stewart has not limited material errors to overlooked elements in prior art before the examiner at the time of the challenged patent’s prosecution. In Xencor, Inc. v. Merus N.V., AD Stewart referred an IPR petition to the PTAB because the examiner made the material error of allowing a patent based on the absence of teachings in the prior art although a patent cited in a later IDS for a continuation of the challenged patent included those teachings (IPR2025-00604, Paper 12). In Western Digital Tech., Inc. v. Godo Kaisha IP Bridge 1, AD Stewart considered it a material error for the examiner to allow a patent based on the absence of elements in the prior art only to find them in a newly cited reference when examining the patent’s child applications (IPR2025-00701, Paper 9).
Another material error overcoming settled expectations is overgeneralization in patent examiner allowances. For example, in Skullcandy, Inc. v. Earin AB, the examiner had cited to elements in the Reasons for Allowance that were in the first independent claim, but not in the other independent claims, bringing the latter claims’ validity into question (IPR2025-00690, Paper 9). In Activision Blizzard, Inc. v. Milestone Entertainment, LLC (IPR2025-00708), the patentee had overcome examiner-cited prior art by narrowing its claim terms, but later broadened them again in subsequent continuations (Paper 13 citing Paper 12 at 13-15). The Petitioner argued that this broadening put the claims back within the teachings of the prior art, but the examiner had failed to reconsider the prior art. Similarly, in USAA Federal Savings Bank v. Pacid Tech. LLC, the Petitioner noted that the patents had been allowed without substantive office action, leading AD Stewart to note that they relied upon an ancestor application to a different category of technology (IPR2025-00697, Paper 9, Paper 8 at 24). AD Stewart found that although the two technologies differed, the examiner made the material error of continuing to apply the same prior art for analysis. Based on this, AD Stewart referred all the IPRs to the PTAB even though the patents’ issuances ranged back to 2017.
Other material errors recognized by AD Stewart that were not used to overcome settled expectations, but likely could be successful, are a failure to consider proceedings related to the patent (IPR2025-00557, Røde Microphones, LLC, et al. v. Zaxcom, Inc., Paper 11) and a failure to consider the previously presented prior art in light of newly presented prior art (IPR2025-00152, Tesla, Inc. v. Charge Fusion Tech., LLC, Paper 11).
Conclusion
AD Stewart’s decisions suggest that Petitioners facing patents that issued five or more years earlier should carefully review the Reasons for Allowance, if any are given, and determine if references in the prosecution history, including in parent and child applications, actually disclose elements identified as the alleged point of novelty distinguishing the patented claims. Petitioners should also be sensitive to prior art that may have been reviewed in related forums, such as International Search Reports or other proceedings, which were ignored, overlooked, or simply misapprehended during examination of the challenged patent. Petitioners should further confirm that a challenged patent that was a continuation received the full review rather than being sped along because its claims were similar to its parents’ issued claims. The examiner may have overlooked subsequent modifications in claim language resulting in invalidity. Until the Proxense defenses are further clarified, capitalizing on examiner material error may be the best course to overcome settled expectations.
[View source.]
