The Federal Circuit recently affirmed a Board decision canceling all but five challenged claims, which survived not on their merits, but because the petitioner failed to adequately argue they were unpatentable. Keysight Techs., Inc. v. Centripetal Networks, LLC, Appeal 2025-1053 (Fed. Cir. July 21, 2026) (nonprecedential). Each surviving claim depends upon a base claim the Board concluded was obvious. Each surviving claim therefore might be obvious for the same reasons. The court acknowledged the concern: “The ghost in the room here whispers that we are allowing plainly unpatentable claims to continue to live on. … Be that as it may, we do not opine on the patentability of [those] claims.” Keysight, Slip Op. at 8. So, the claims remain presumptively valid and enforceable, leaving unresolved whether they are valid.

In its IPR petition, Keysight challenged all claims of Centripetal Network’s U.S. Patent No. 11,012,474 as obvious. But the petition offered only a header and a chart listing dependent claims 6, 14, 22, 30, and 33 as being obvious over the combination of two prior art publications. Missing from the petition was an element-by-element mapping of those claims to the prior art, expert testimony supporting that mapping, and substantive argument specifically explaining why these dependent claims recite obvious subject matter. The Board instituted anyway and then, in its final decision, held that Keysight had not carried its burden to prove those claims are unpatentable. Keysight, Slip Op. at 3–5; see SAS Inst. Inc. v. Iancu, 584 U.S. 357, 370 (2018) (prohibiting institution of only a subset of the challenged claims).

On appeal, the Federal Circuit affirmed under an abuse-of-discretion standard of review, applying Netflix, Inc. v. DivX, LLC, 84 F.4th 1371, 1376–77 (Fed. Cir. 2023), and Wasica Finance GmbH v. Continental Automotive Systems, Inc., 853 F.3d 1272, 1285 (Fed. Cir. 2017) (discussed here). Under Netflix, a petitioner who files a “vague, generic, and/or meandering” petition cannot “later fault the Board” for failing to divine unstated arguments; and under Wasica, a claim mentioned only in a header and not specifically argued is not adequately challenged. Applying both, the court concluded that Keysight’s header-and-chart treatment of those five dependent claims fell short, and that the Board did not abuse its discretion in so finding. The court also distinguished its earlier decision in In re NuVasive, Inc., 841 F.3d 966 (Fed. Cir. 2016) (discussed here), where the petition had at least given the patent owner “minimally sufficient” notice. Here, apparently, Keysight did not offer such notice because it “failed entirely even to address the unpatentability of claims 6, 14, 22, 30, and 33 in its petition.” Keysight, Slip Op. at 6.

The decision is straightforward. The Board and the Federal Circuit need not search for arguments the petitioner did not clearly make. As the court put it, quoting a colorful line from a Seventh Circuit panel: “Judges are not like pigs, hunting for truffles buried in briefs.” United States v. Dunkel, 927 F.2d 955, 956 (7th Cir. 1991); Keysight, Slip Op. at 6 (quoting later decisions for the same proposition). Here, the Board’s judges ostensibly found the arguments when they decided to institute, but lost them when they finally decided the case. Nevertheless, the practical lesson is simple: petitioners must prove that each challenged claim is unpatentable. If they fail to do so—whatever the reason—they should expect to lose as to that claim.

The affirmance is likely sound, but the result seems troubling—not because IPR estoppel now bars Keysight from asserting grounds it “raised or reasonably could have raised” against the five surviving claims. 35 U.S.C. § 315(e). The real problem is that those claims remain in force even though the court flagged them as potentially “plainly unpatentable.” The Patent Office has not evaluated their merits against the prior-art combination Keysight’s petition identified. The Office’s recent directives may complicate further requests to review these claims. See, e.g., Magnolia Med. Techs., Inc. v. Kurin, Inc., IPR2026-00097, Paper 17 (Director May 14, 2026) (precedential); see also, PTO Notice: “Pre-order Procedure Regarding Substantial New Question Determination in Ex parte Reexamination Proceedings”, 1545 O.G. 217 (April 1, 2026).

The Director can address this problem. As another recent post explained, 35 U.S.C. § 303(a) authorizes the Director to order reexamination “on his own initiative, and at any time,” when a substantial new question of patentability exists. The Supreme Court has recognized this authority as part of the Director’s inquisitorial role. SAS, 584 U.S. at 370–71. The Director does not need a third-party request, and the prior art need not be new to the prosecution record. Nor would IPR estoppel or Keysight prevent reexamination. Indeed, Keysight expressly declined to decide whether the surviving claims are patentable. Keysight, Slip Op. at 8.

The existing record may already be enough. The Patent Office has the prior-art combination, so examiners would not need to start from scratch. The Board already found that the same combination invalidated other challenged claims—including the base claims from which each surviving claim depends. Id. at 4. And at institution, the Board preliminarily found that Keysight had “establishe[d] sufficiently that the combined disclosures in [the prior art] teach the inventions covered by” the five surviving dependent claims. Id. at 3. This record may therefore present the substantial new question of patentability needed for reexamination.

A Director-ordered reexamination would put the prior-art combination before examiners, not the Board. That should avoid concerns about asking the Board to revisit the IPR while giving the public what the IPR did not: an examination-based decision on patentability. The Director has denied potentially meritorious IPR petitions on discretionary grounds and directed petitioners toward reexamination. See, e.g., In re Intel Corp., Case 2026-113 (Fed. Cir. Feb. 24, 2026) (mentioned here). Accordingly, § 303(a) reexamination should not be viewed as unduly burdensome—especially when the Office has already developed the record and the Federal Circuit highlighted the unresolved patentability question. Afterall, a guiding purpose behind post-issuance review is “to determine whether the patent at issue may have been improvidently granted and whether and to what extent that right should continue to be recognized at all.” Magnolia Med. Techs., Paper 17, at 7 (discussing IPRs).