Listen to this post

In a recent precedential decision from the Federal Circuit, the appellate court explained that words of approximation like “about” are not inherently definite or indefinite — but when a patentee fails to provide sufficient guidance as to the meaning of such terms, it can be fatal to the claims. More specifically, in Enviro Tech Chemical Services, Inc. v. Safe Foods Corp., the court affirmed the lower court’s determination that all asserted claims of U.S. Patent No. 10,912,321 were invalid as indefinite under 35 U.S.C. § 112(b) because the term “about” failed to inform skilled artisans about the scope of the invention with reasonable certainty.

Background

Enviro Tech’s ’321 patent is directed to methods for treating poultry during processing to increase the weight of the poultry using peracetic acid. Representative claim 1 recites, among other things, a step of altering the pH of the peracetic acid-containing water “to a pH of about 7.6 to about 10 by adding an alkaline source.”

When Safe Foods was sued by Enviro Tech in U.S. District Court for the Eastern District of Arkansas for allegedly infringing the ’321 patent, it argued during claim construction that the term “about” was indefinite. The district court agreed, determining that both “about” and “an antimicrobial amount” were indefinite, and entered judgment that the asserted claims were invalid.

The Federal Circuit’s Analysis

In an opinion authored by Judge Alan Lourie, the Federal Circuit reiterated that terms of degree like “about” and “approximately” are not inherently definite or indefinite. However, when such words of approximation are used, the parameter’s range must be “reasonably certain based on the technological facts of the particular case.” The appellate panel systematically examined the intrinsic evidence — the claims, specification, and prosecution history — and found that none provided adequate guidance as to the scope of the term.

Starting with the claim language, the panel noted that, while the claims recite a “pH of about 7.6 to about 10,” they provide no guidance on how much below 7.6 or above 10 the pH could deviate while still meeting the claim limitation. The parties agreed that “about” means “approximately,” but the court observed that “approximately” provides no more clarity than “about” itself.

Turning to the specification, the court found it equally unhelpful. The specification recited numerous experiments where Enviro Tech would set a target pH, measure the actual pH, and proceed based on the difference. In a majority of experiments, Enviro Tech proceeded only when the difference was less than or equal to 0.3 of the target pH. However, there were notable exceptions. In one particularly informative experiment conducted at a major U.S. poultry processing plant with 5.8 million chickens, Enviro Tech proceeded with deviations between 0.35 and 0.5 of the target pH. The court concluded that this conflicting guidance “does not allow a skilled artisan to determine the scope of ‘about’ with reasonable certainty.”

The prosecution history also failed to clarify the scope of the term. In fact, the court found that Enviro Tech’s treatment of “about” during prosecution was inconsistent. In one office action response, Enviro Tech argued that “a peracetic acid solution at the lower end of the claimed range, pH 7.6” would not have been obvious over the prior art — notably omitting the term “about.” Yet, when discussing another claim in the same response, Enviro Tech included “about” in its arguments about pH values. Enviro Tech did not explain what “about” means at any point during prosecution.

Enviro Tech attempted to argue that its amendment of the lower boundary from “about 7.3” to “about 7.6” during prosecution demonstrated that “about” should be construed to mean less than or equal to 0.3 pH. The Federal Circuit was not persuaded and noted that Enviro Tech did not cite any remarks made to or by the examiner that would define the term and never offered any argument indicating what “about” means. The panel also observed that the prior art disclosed a pH as close as 7.0, which necessitated the amendment to the claims. The court noted that the prior art is “almost ‘about’ a pH of 7.6,” highlighting the imprecision of the term.

Takeaway

This decision serves as an important reminder that words of approximation in patent claims are not a free pass to avoid strict numerical boundaries. While the Federal Circuit has long recognized that terms like “about” and “approximately” may be appropriately used, patentees bear the burden of ensuring that the intrinsic evidence — particularly the specification and prosecution history — provides skilled artisans with reasonable certainty as to the scope of such terms. When the specification provides conflicting examples of permissible deviations, and the prosecution history is inconsistent in its treatment of the approximation term, the claims will likely fall as indefinite.

In short, if a patent relies on terms like “about” to define a claim boundary, the specification should provide clear and consistent guidance as to what deviations are acceptable — ideally with explicit statements defining the term.

Listen to this post

A recent precedential Federal Circuit decision further clarifies the limited scope of judicial review over PTAB institution-related rulings, holding that refusal by the Patent Trial and Appeal Board to resolve a disputed real-party-in-interest issue under 35 U.S.C. § 312(a)(2) is unreviewable on appeal. The April 29, 2026, decision in Federal Express Corporation v. Qualcomm Incorporated, No. 24-1236, provides important considerations for both petitioners and patent owners on institution-stage challenges and what remains open on appeal.   

Background

The case arises from inter partes review (IPR) petitions filed by Qualcomm Incorporated challenging four of Federal Express Corporation’s patents, including U.S. Patent No. 8,766,797 (the ’797 patent) related to systems and methods for accessing sensor-derived shipment information.

In February 2021, FedEx sued Roambee Corporation (since rebranded as Decklar) in the District of Delaware, asserting six FedEx patents, including the ’797 patent. Service of the complaint started Roambee’s one-year window under 35 U.S.C. § 315(b) to file its own IPR. On the very last day of that window, Qualcomm — a non-party to the Delaware litigation — filed IPR petitions targeting four of the six asserted patents. Qualcomm identified the Roambee litigation as a related matter but did not list Roambee as a real party in interest.

FedEx opposed institution under § 312(a)(2), which provides that an IPR petition may be considered only if it identifies all real parties in interest. The board declined to resolve the issue. Instead, the board held that it need not determine whether an unnamed party is a real party in interest where doing so is unnecessary to resolve the proceeding. The board instituted IPR, denied a later motion to terminate, and issued a final written decision finding all challenged claims unpatentable. FedEx appealed.

The Primary Issue on Appeal

Under 35 U.S.C. § 314(a), the director (or the board exercising delegated authority) decides whether to institute the requested inter partes review, and under 35 U.S.C. § 314(d), institution decisions are final and generally unreviewable. The Supreme Court held in Cuozzo Speed Techs., LLC v. Lee, 579 U.S. 261 (2016), that § 314(d) bars review of matters “closely tied to the application and interpretation of statutes related to” the institution decision. Challenges aimed at whether the agency properly instituted review fall squarely within that bar. The Federal Circuit extended the same framework to § 312(a)(2) real-party-in-interest determinations in ESIP Series 2, LLC v. Puzhen Life USA, LLC, 958 F.3d 1378 (Fed. Cir. 2020).

The Supreme Court has carved out a narrow exception where the agency acts outside its statutory authority after institution. In SAS Institute, Inc. v. Iancu, 584 U.S. 357 (2018), the petitioner did not challenge whether review should have been instituted, but rather how the board conducted the review after institution.

The question for the Federal Circuit was therefore whether FedEx’s challenge fit within the SAS exception or, instead, was an institution challenge.

The Federal Circuit’s Analysis

The Federal Circuit concluded that FedEx’s challenges “grounded in § 312(a)(2)” were not reviewable. The court viewed the real-party-in-interest requirement of § 312(a) as a threshold condition for institution and “integral to” that decision. Because the requirement has force only at the institution stage, any defect in the board’s handling of it necessarily targets the institution decision, regardless of how the challenge is framed. FedEx’s argument was an attack on whether institution should have occurred at all, and § 314(d) forecloses that attack on appeal.

The court rejected FedEx’s effort to invoke SAS. The challenge in SAS addressed a post-institution failure to follow a clear statutory guideline — the duty to decide all challenged claims — and did not seek to undo institution. By contrast, the relief FedEx ultimately sought — vacatur of the final written decision based on a § 312(a)(2) defect — would, in substance, mean the board should not have instituted at all. Reaffirming its prior holdings in ESIP and Ethanol Boosting Systems, LLC v. Ford Motor Co., 162 F.4th 1151 (Fed. Cir. 2025), the court emphasized that allegations of acting in excess of statutory jurisdiction “closely tied to the application and interpretation of statutes related to” the agency’s decision to initiate IPR, without more, do not overcome § 314(d)’s bar.

FedEx fared better on the merits, where the Federal Circuit vacated the board’s obviousness determination as to claims 6, 17, and 28 of the ’797 patent and remanded the case.

Key Takeaways

This decision reinforces several important principles for IPR practice:

  • § 312(a)(2) is not a procedural “gotcha.” Alleged defects in identifying real-parties-in-interest are unlikely to create appealable issues.
  • “Manner of proceeding” framing will not revive institution-tied challenges. The court rejected FedEx’s attempt to recast a § 312(a)(2) objection as a post-institution procedural error. Genuine SAS-style review remains limited to deviations from a clear statutory guideline after a valid institution.
  • Focus on the PTAB. The Federal Circuit’s ability to revisit PTAB decisions on appeal is significantly constrained, particularly for threshold and institution-stage determinations, so clients must focus on winning the fight at the PTAB level.

For practitioners, the message is consistent with the post-Cuozzo line of cases: Treat institution-stage arguments as decisive, and do not count on the Federal Circuit to revisit them later.

Listen to this post

In a significant decision for telecommunications patent law, a panel of the Federal Circuit issued a mixed ruling in Constellation Designs, LLC v. LG Electronics Inc, vacating a summary judgment of patent eligibility for certain claims while affirming eligibility for others, and affirming the jury’s findings of infringement and damages.  The April 28, 2026, opinion provides important guidance on how courts should apply the Alice/Mayo framework to distinguish between impermissibly abstract, result-oriented claims and patent-eligible claims directed to specific technological implementations.

The Claimed Technology

Constellation’s patents (U.S. Patent No. 8,842,761 (the ’761 patent); U.S. Patent No. 10,693,700 (the ’700 patent); U.S. Patent No. 11,019,509 (the ’509 patent); and U.S. Patent No. 11,018,922 (the ’922 patent)) relate to digital communication systems that use non-uniform “constellations” — visual representations of the relationship between digital bits and broadcast signals — to improve transmission capacity. Traditional/uniform constellations were designed with evenly spaced points to maximize the minimum distance between constellation points (dmin), on the theory that greater spacing would improve capacity at high SNRs.

The inventors of the asserted patents took a different approach. Rather than optimizing based on dmin, the inventors developed an iterative process to construct non-uniform constellations optimized for “parallel decode (PD) capacity” — a measure comparing the information entering the mapper on one side of the channel with what exits the demapper on the other.  According to the written description of the asserted patents, these capacity-optimized constellations “can achieve significantly higher performance gains” as compared to conventional constellations (achieving efficiency gains of more than 25%).

The Procedural Background

Constellation Designs sued LG Electronics for patent infringement in the Eastern District of Texas, alleging that LG willfully infringed various claims in the four asserted patents through sales of its televisions compatible with the Advanced Television Systems Committee (ATSC) 3.0 over-the-air broadcast television standard. The district court granted Constellation’s motion for summary judgment of patent eligibility, holding all asserted claims patent eligible under 35 U.S.C. § 101 as being directed to a technical solution to a technical problem — specifically, overcoming data loss in over-the-air transmission through PD-capacity-optimized constellations.

Following a jury trial, the jury found all asserted claims valid and that LG’s accused televisions infringed at least one claim of each asserted patent. The jury awarded Constellation $1,684,469 in past damages (based on a per-television royalty of $6.75) and found LG’s infringement willful. LG appealed the district court’s eligibility ruling, denial of judgment as a matter of law on non-infringement, and denial of judgment as a matter of law on no damages.

The Federal Circuit’s Analysis

The Federal Circuit, in an opinion by Judge Kara Stoll joined by Judge Alan Lourie and District Judge Paul Oetken sitting by designation, divided the asserted claims into two groups: (1) “optimization claims” from the ’761 and ’700 patents, which recite non-uniform constellations “optimized” for PD capacity; and (2) “constellation claims” from the ’509 and ’922 patents, which recite specific non-uniform constellations with defined characteristics. The panel reached meaningfully different conclusions for the two claim groups after applying the two-step Alice/Mayo framework.

The Optimization Claims: Ineligible as Result-Oriented

At Alice step one, the Federal Circuit panel found the optimization claims directed to the abstract idea of “optimizing” a constellation for PD capacity. The panel drew on the Supreme Court’s reasoning in O’Reilly v. Morse and its own precedent in ChargePoint, Inc. v. SemaConnect, Inc.  to conclude that claim 17 of the ’761 patent — representative of the optimization claims — was “an abstract, result-oriented claim directed to all ways of achieving a recited result.”

The court emphasized that the key distinguishing element of claim 17 — the “wherein” clause reciting a constellation “optimized for capacity using parallel decode capacity” — was drafted in a “result-oriented way” that covers every possible method of achieving a constellation “optimized for capacity using parallel decode capacity” without reciting how that optimization is to be achieved. While the specification described an iterative process for constructing a geometrically shaped constellation, including steps for selecting constellation size, checking feasibility, iterating to convergence, and outputting the optimized result, the § 101 inquiry focuses on the claim language itself, and the claims themselves did not incorporate those details. As the court noted, “the specification cannot be used to import details from the specification if those details are not claimed.”

At Alice step two, Constellation had argued that the claims recited an inventive concept based on the efficiency gains achieved and the inventors’ recognition that optimizing for PD capacity yields counterintuitive, improved results. The court rejected this argument, holding that Constellation’s alleged inventive concept “is the abstract idea itself,” which cannot transform claims into a patent-eligible application. The court found no additional elements — individually or as an ordered combination — that transformed the claims’ nature. As a result, the Federal Circuit vacated the lower court’s summary judgment of patent eligibility under 35 U.S.C. § 101 of the optimization claims.

The Constellation Claims: Eligible as Specific Implementations

The constellation claims fared differently. Unlike the optimization claims, the asserted claims of the ’509 and ’922 patents were directed to “specific constellations” that the inventors developed using the techniques described in the specification — including constellations with overlapping point locations and hundreds of specific non-uniform constellations identified by their precise coordinates.

Claim 21 of the ’509 patent was deemed representative of the constellation claims and “directed to solving a particular technological problem — overcoming capacity constraints to improve coding gains — using a particular technological solution — specific, non-uniform constellations with overlapping constellation point locations.” More specifically, claim 29 of the ’509 patent requires that a symbol constellation’s point locations be unequally spaced, that each constellation point carry a different label, and that at least two constellation points share the same location — a counterintuitive structural feature flowing directly from the PD-capacity optimization approach. The court held that these concrete structural limitations provide sufficient specificity to satisfy Alice step one without need to reach step two.

Other Holdings

The Federal Circuit also affirmed the district court’s denial of LG’s motions for judgment as a matter of law on non-infringement and no damages. On infringement, the court held that a patent owner may prove infringement of individual claim limitations using standards-related evidence for some claim elements while using product-specific evidence for other elements of the same claim — an approach the court found consistent with its earlier decision in Fujitsu Ltd. v. Netgear Inc. The court found substantial evidence supporting the jury’s infringement verdict, including expert testimony, source code analysis, internal LG testing documents, and the accused televisions’ compliance with the FCC-mandatory A/322 protocol of ATSC 3.0.

On damages, LG had challenged Constellation’s damages expert’s reliance on Zenith patent licenses as comparables for a built-in apportionment theory, arguing that the Zenith licenses covered different patents, technologies, and product types. The court rejected LG’s challenge and agreed with the district court that this was a Daubert admissibility challenge — not a sufficiency-of-evidence argument suitable for JMOL — and found sufficient evidentiary support for comparability based on technical and economic similarities established through multiple witnesses to support the jury’s damages verdict.

Key Takeaways

This decision reinforces several important principles for patent eligibility:

  • Result-Oriented Claiming Creates Eligibility Risk – Claims that define an invention solely by the result achieved — without reciting the specific method or structure through which that result is obtained — risk being characterized as directed to an abstract idea under Alice step one. Even where a specification describes the inventive process in concrete detail, those details must be reflected in the claims themselves to survive § 101 scrutiny.
  • Written Disclosure Cannot Save Broad Claims – Technical details in the specification will not cure eligibility defects if those details are not incorporated into the claim language itself.
  • Specificity Matters – Claims that recite specific implementations — such as particular constellation configurations — are more likely to survive eligibility challenges than claims that broadly cover all ways of achieving a result.
  • Standard-Based Infringement Proof Extends Limitation-by-Limitation – The court’s extension of Fujitsu confirms that patent owners may prove infringement of individual claim limitations through compliance with an industry standard — without being required to show that the entire claim is satisfied by the standard. This provides meaningful flexibility for patentees whose technology is incorporated into mandatory or widely adopted standards.

Moreover, novelty does not always equal eligibility. Indeed, the court rejected Constellation’s argument that evidence of novelty and non-obviousness — including unsuccessful IPR challenges — demonstrated an inventive concept, reaffirming that the eligibility inquiry is separate from the novelty and obviousness inquiries.

The case now returns to the Eastern District of Texas for further proceedings on the optimization claims.

Listen to this post

A pending Federal Circuit decision may force licensing counsel to rethink one of the most routine provisions in patent settlement agreements — the licensee’s denial of infringement.

Patent settlement agreements usually contain a denial of liability by the defendant/licensee, e.g., “[l]icensee denies any infringement of the asserted patents.”  For decades, this language has largely been viewed as harmless boilerplate — a standard concession to the settling defendant with no apparent downstream consequences. However, an appeal pending before the Federal Circuit suggests that this boilerplate sentence may have effects far beyond the agreement in which it appears.

The case, VDPP, LLC v. Volkswagen Group of America, Inc., No. 24‑2226, arises from VDPP’s August 2023 lawsuit in the Southern District of Texas asserting U.S. Patent No. 9,426,452 (the ’452 Patent), which is directed to a 3D display and glasses system. VDPP — a non-practicing entity (NPE) — accused Volkswagen of infringement for certain surround‑view camera systems.  Volkswagen moved to dismiss, arguing that VDPP failed to comply with the patent marking statute (35 U.S.C. § 287(a)) before the patent expired. The district court agreed with Volkswagen and dismissed the case with prejudice, denied leave to amend, and later granted Volkswagen’s motion for attorneys’ fees under 35 U.S.C. § 285 holding both VDPP and its counsel jointly and severally liable. VDPP appealed. Oral argument was held on April 9, 2026, before Chief Judge Kimberly Moore, Judge Alan Lourie, and Judge Tiffany Cunningham.

The issue on appeal, in plain terms

Under 35 U.S.C. § 287(a), a patent holder who licenses patented technology must ensure that products covered by the license are properly marked. If the patent holder — or its licensees — fail to mark, damages are limited to the period after the accused infringer receives actual notice of infringement. For an expired patent where only pre‑suit damages are available, failure to mark can eliminate any recovery entirely.

For NPEs that do not themselves manufacture products, § 287(a) often lies dormant. The obligation typically arises only once the patent holder licenses others to make products under the patent.  

The unresolved question now before the Federal Circuit is whether that marking obligation is triggered by a settlement agreement — even when the agreement contains an express denial of infringement by the licensee.

A skeptical reception at oral argument

VDPP’s position on appeal is that its prior settlement agreements with other alleged infringers were not the kind of licenses that trigger § 287(a) at all. According to VDPP, the agreements were simply litigation settlements in which the respective defendants paid to resolve disputes while expressly denying infringement. VDPP argued that the agreements did not permit the defendants “to produce a patented article for or under” the ’452 Patent and, therefore, should not be viewed as licenses in the traditional sense such that any imposed a marking obligation.

The panel appeared unpersuaded. For example, when VDPP claimed that the licensees’ denial of infringement should be dispositive, Chief Judge Moore pushed back:

Every defendant insists they don’t infringe.  And if I were to adopt a rule that says a patentee has no obligation to force marking upon a licensee if they claim they don’t infringe, then I fear we would pretty much never have marking by licensees.

What to watch for when the decision drops

The central question is whether denial-of-infringement language insulates a settlement from the marking trigger of § 287 — or whether the court will look past the label to the substance of what the agreement actually permits.  

If the court focuses on substance over form, a licensor with a portfolio of prior settlement agreements — each framed as a dispute resolution — may have unknowingly started a marking clock years ago. Subsequent defendants could point to those prior agreements and argue that the patentee and prior licensees were obligated to mark but failed to do so, cutting off all pre‑expiration damages. And if the court signals that substance controls, the choice of settlement instrument will likely matter far more than it has historically: A covenant not to sue or a bare release of past claims may be treated differently than a license, giving practitioners a structural tool to resolve disputes without inadvertently triggering obligations that could impair future enforcement.

Practical steps before the opinion issues

Counsel advising patent holders with enforcement portfolios should not wait for the decision to begin reviewing their settlement agreements. Three steps worth taking now:

  1. Audit prior settlement agreements. Review existing agreements to determine whether any were structured as outright licenses versus covenants or releases. If licenses exist, assess whether marking occurred — or realistically could have occurred — during the license period.
  2. Revisit current templates. For any active licensing negotiations, consider whether the settlement agreement being used is appropriate given the marking risk, particularly for patents approaching or past expiration.
  3. Address § 287 expressly. If a denial-of-infringement clause is non-negotiable, consider pairing it with explicit language requiring the licensee to mark any covered products — notwithstanding the denial.

The panel’s questioning at oral argument suggests an uphill battle for VDPP. But the § 287 issue presented here — whether a denial-of-infringement in a settlement agreement defeats a marking obligation that would otherwise attach — is genuinely novel. A clear ruling in either direction will be a meaningful addition to patent licensing practice. 

Stay tuned.

Listen to this post

Last week, the Supreme Court vacated a copyright infringement judgment against an ISP provider and remanded the case to the Fifth Circuit for reconsideration in light of the Cox Communications decision (Grande Comm’s Networks v. UMG Recordings, Inc., 2026 WL 922501, at *1 (U.S. April 6, 2026)). Specifically, the Court directed the Fifth Circuit to reconsider its prior decision upholding a contributory infringement liability verdict against ISP Grande Communications Networks.

The Fifth Circuit concluded that a Texas federal district court had correctly upheld a jury verdict finding Grande Communications Networks liable for the willful contributory copyright infringement of over 1,400 songs after being made aware that its service subscribers were peer-sharing copyrighted songs. At the trial court level, the record label plaintiffs claimed Grande Communications Networks was liable for secondary copyright infringement for failing to ban its subscribers who the labels had accused of illegally sharing copyrighted material through the software BitTorrent. On the issue of contributory copyright infringement, the three-judge panel of the Fifth Circuit determined that the district court applied the correct legal standard by determining that Grande Communications Networks could be secondarily liable for copyright infringement committed by its subscribers “if it induced, caused, or materially contributed to its subscribers’ infringing activity.”

Grande Communications Networks argued before the district court that material contribution is not a valid basis for contributory copyright liability because “the Supreme Court has recognized two — and only two — types of contributory copyright infringement: (1) where the defendant distribute[s] a product or service without any commercially significant, non-infringing use, and (2) where there is clear expression or other affirmative steps taken to foster infringement by the defendant.” The district court instructed the jury that Grande Communications Networks was contributorily liable if it induced, caused, or materially contributed to the infringing activity, and that this standard was met if Grande Communications Networks “could have taken basic measures to prevent further damages to copyrighted works, yet intentionally continued to provide access to infringing sound recordings.” In its opinion, the Fifth Circuit panel — relying in significant part on the Fourt Circuit’s decision that was reversed in the Cox Communications case — found that there was “no basis to reverse the jury’s verdict that Grande is liable for contributory infringement” because “(1) intentionally providing material contribution to infringement is a valid basis for contributory liability; (2) an ISP’s continued provision of internet services to known infringing subscribers, without taking simple measures to prevent infringement, constitutes material contribution; and (3) the evidence at trial was sufficient to show that Grande engaged in precisely that conduct[.]”

Recall that in the Supreme Court’s March 25 decision Cox Communications v. Sony Music Entertainment, which was the subject of a previous blog post, the Court held that Cox Communications, also an internet service provider, could only be found to be contributorily liable for copyright infringement if it affirmatively induced infringement by a subscriber or offered a service designed primarily for infringing uses. In Cox Communications, the Court held that an ISP cannot be held contributorily liable for copyright infringement merely because it continued providing internet service to its subscribers who had been the subject of infringement complaints, and that contributory liability requires proof that the provider actually intended its service to be used for infringement — shown either through active inducement or by offering a service with no substantial non-infringing uses. The holding in Cox Communications appears to vindicate Grande Communications Networks’ position before the trial court, as well as setting a significantly higher bar for contributory copyright infringement than the “material contribution” standard applied by the district court and Fifth Circuit.

Listen to this post

Cox Communications v. Sony Music Entertainment | Decided March 25, 2026

On March 25, 2026, the U.S. Supreme Court delivered a landmark decision that will reshape not only how copyright law applies to the internet for years to come, but could impact other areas of intellectual property law as well. In Cox Communications, Inc. v. Sony Music Entertainment, the Court held that internet service providers cannot be held indirectly liable for their customers’ copyright infringement simply because the ISPs knew the infringement was happening but failed to prevent it. The decision reversed and remanded a billion-dollar judgment against ISP Cox Communications and drew a more clearly defined line around secondary copyright liability.

How We Got Here

In 2018, Sony Music Entertainment and dozens of other major record labels sued Cox Communications — with six million subscribers, one of the largest broadband providers in the United States — in federal district court in Virginia. The theory of liability? Cox Communications had received thousands of infringement notices between 2013 and 2014 identifying specific subscriber accounts that had been repeatedly caught pirating the plaintiffs’ copyrighted music. Cox Communications terminated only 32 customers for copyright infringement during that period — while simultaneously terminating hundreds of thousands of subscribers for nonpayment. The labels argued that Cox Communications’ decision to keep known infringers on its network was motivated by a desire to protect subscription revenue, a combination that made Cox Communications legally responsible for their customers’ copyright-infringing misconduct.

A jury agreed with the plaintiffs, finding Cox Communications liable under two separate legal theories — contributory infringement and vicarious infringement —awarding the music labels $1 billion in statutory damages for willful copyright infringement.

On appeal, the Fourth Circuit split the baby: It threw out the vicarious liability finding but upheld the contributory infringement verdict, reasoning that knowingly supplying a service to someone you believe will use it to infringe copyrights is “exactly the sort of culpable conduct sufficient for contributory infringement.” The Fourth Circuit ordered a new trial on damages, and Cox Communications appealed. The Supreme Court granted cert in June of 2025 on the contributory infringement question.

What the Court Decided

Writing for seven justices, Justice Clarence Thomas reversed the Fourth Circuit in Cox Communications’ favor. The opinion’s core holding is seemingly and certainly consequential: “Under our precedents, a company is not liable as a copyright infringer for merely providing a service to the general public with knowledge that it will be used by some to infringe copyrights.”

Instead, to prove contributory infringement, a copyright owner must show that the provider actually intended for its service to be used for infringement — and the required intent can be established in only two ways. Either the provider affirmatively induced the infringement, or it offered a service specifically designed or tailored to facilitate the infringing activity.

The Court held that Cox Communications did neither. The Court acknowledged that Cox Communications knew infringement was occurring on its network. But general internet access, the Court noted, has numerous non-infringing, lawful uses — it is not a service designed for piracy the way that, say, certain peer-to-peer file-sharing platforms had been (more about Grokster later). And Cox Communications did not encourage its customers to download music illegally; its terms of service explicitly prohibited it. Failing to terminate known infringing accounts, the Court concluded, is not the same as intending infringement.

The opinion puts it plainly: The Fourt Circuit’s reasoning conflicted with the “Court’s repeated admonition that contributory liability cannot rest only on a provider’s knowledge of infringement and insufficient action to prevent it.”

The Court also addressed the labels’ argument that the Digital Millennium Copyright Act (DMCA) — which creates a “safe harbor” for ISPs that take steps to address repeat infringers — implies that providers who fail to meet those safe harbor requirements must therefore be liable. The Court rejected this reading, noting that the DMCA expressly states that failing to qualify for a safe harbor “shall not bear adversely upon” an infringement defense. In other words, missing the safe harbor does not automatically equal liability.

Why This Decision Is Bigger Than Music

The music industry is continuing the fight; the Recording Industry Association of America called the ruling a disappointment and urged policymakers to examine its impact. The labels argued throughout the case that Cox Communications knowingly facilitated “theft on a massive scale,” and there was trial evidence suggesting that some Cox Communications employees were reluctant to terminate infringing accounts precisely because they did not want to lose subscription revenue. However, none of that, in the Supreme Court’s view, is enough.

But the decision’s reach extends well beyond streaming services and record labels. The case attracted amicus briefs from a wide range of civil liberties and First Amendment organizations. The Fourth Circuit’s reasoning, if left standing, would have threatened to hold any general-purpose internet platform liable when users commit infringement, so long as the platform knew about it and did not act decisively enough. That standard would have hung over every social media platform, cloud storage provider, and AI company that hosts or processes user-generated content.

X’s amicus brief to the Court put it thusly: The lower court’s rule could “wreak havoc” on the technology industry and specifically on AI, noting that if content creators can sue AI platforms whenever users employ their tools to violate copyright, tech companies would have no choice but to drastically curtail their services. Given the wave of copyright litigation currently targeting AI companies, the Court’s decision to cabin contributory liability to intentional actors arrives at a particularly consequential moment.

Cox’s Potential Impact Beyond Copyright Law – Patent Law Practitioners, Pay Attention

Central to Justice Thomas’ opinion is something that patent litigators cannot/should not ignore: a deliberate, explicit alignment of contributory copyright liability with its counterparts in patent law. The Court did not stumble into patent territory by accident. It went there on purpose, and in doing so it may have set the stage for a tightening of indirect patent infringement doctrine.

To understand why Cox matters for patent law, you need to understand where copyright’s secondary liability doctrine came from in the first place: patent law. Copyright’s contributory infringement framework has no explicit statutory basis. Unlike the Patent Act — which codifies both induced infringement under 35 U.S.C. § 271(b) and contributory patent infringement under 35 U.S.C. § 271(c) — the Copyright Act says nothing about liability for facilitating another person’s copyright infringement. Courts (including the Supreme Court) have filled that gap by borrowing from patent law, citing what the Court in Cox called the “historic kinship” between the two fields of IP law.

The Court acknowledged in Cox that this kinship runs through a trio of the Court’s landmark secondary liability decisions: Kalem Co. v. Harper Brothers (1911), Sony Corp. of America v. Universal City Studios (1984), and Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd. (2005). In each of those cases, the Court looked to patent law’s standards for indirect infringement — particularly the staple article of commerce doctrine and the inducement rule — when deciding how far copyright’s secondary liability should reach. Cox is the latest branch in the copyright-patent indirect liability family tree.

1. What Cox Actually Holds — and Why It Tracks Patent Law So Closely

Justice Thomas’ majority opinion makes the patent law connection explicit and structural. After announcing the two-pathway framework for contributory copyright liability — inducement or a service tailored to infringement — the opinion states flatly: “These two forms of contributory infringement track patent law.”

The majority then walks through the patent parallels in detail. Under § 271(b), a party is liable for actively inducing patent infringement only when it expresses “an affirmative intent that the product be used to infringe.” Under § 271(c), contributory patent infringement requires that the defendant sell a product “knowing the same to be especially made or especially adapted for use in an infringement of such patent” — but the Court has long held that mere knowledge is not enough; instead, the product must lack substantial non-infringing uses. Both of those standards map directly onto the two pathways for indirect copyright infringement the Court adopted in Cox.

The opinion then drives the point home with a passage that will surely become a staple in indirect patent infringement defense briefing: “This Court has repeatedly made clear that mere knowledge that a service will be used to infringe is insufficient to establish the required intent to infringe.” While the statement was made in a copyright infringement case, because it was made by deliberate reference to patent law’s borrowed structure, how can it not also be read as an endorsement of the same principle across both fields?

2. Broader Implications for Patent Indirect Infringement Doctrine

The timing of the Cox decision is significant. Just five weeks after Cox was decided, the Supreme Court is scheduled to hear oral arguments on April 29 in Hikma Pharmaceuticals USA, Inc. v. Amarin Pharma, Inc., a pharmaceutical patent case squarely about the standard for induced infringement under § 271(b).

But the effects of Cox likely will not stop at the pharmaceutical industry.

The staple article doctrine found in § 271(c) already protects sellers of components that have substantial non-infringing uses from contributory patent infringement. Cox reinforces the same logic in the copyright context and — given the explicit cross-referencing — could be cited to cabin future expansions of either doctrine. Defendants in technology patent cases who supply general-purpose components will seemingly have a cleaner road to arguing that their products are not “especially adapted” for infringing use.

Just as the Court held in Cox that an ISP providing general internet access does not contribute to infringement simply because some users infringe, software platform companies facing patent claims over end-user conduct will argue that general-purpose platforms — those capable of substantial non-infringing uses — should be evaluated by the same demanding intent standard. The logic is seemingly the same whether the underlying IP right is copyright or patent.

The Federal Circuit Court of Appeals (which has exclusive jurisdiction over cases arising under patent law) has at times allowed induced patent infringement cases to proceed on less robust allegations of specific intent, particularly in the Hatch-Waxman context. Cox at least would seem to foreclose the proposition that knowledge of infringement plus inaction equals inducement liability, requiring instead positive conduct that promotes infringement. For instance, the Court identifies in Cox the kind of conduct that would seemingly be necessary to support a finding of inducement — “evidence of express promotion, marketing, and intent to promote” infringement — which might mitigate against findings of inducement in the patent context on a more passive standard.

Cox Communications, Inc. v. Sony Music Entertainment, No. 24-171, 607 U.S. ___ (2026). Decided March 25, 2026. Opinion by Justice Clarence Thomas, joined by Chief Justice John Roberts and Justices Samuel Alito, Elena Kagan, Neil Gorsuch, Brett Kavanaugh, and Amy Coney Barrett. Justice Sonia Sotomayor concurred in the judgment, joined by Justice Ketanji Brown Jackso

Listen to this post

Elon Musk’s year of controversy continues as startup Operation Bluebird attempts to take flight with its rival social media platform — “twitter.new” — by asking the U.S. Patent and Trademark Office (USPTO) to find that Musk’s X Corp. has abandoned its “Twitter” and “tweet” trademarks. X Corp. responded to this petition on December 16 by asking the District Court of Delaware to find that Operation Bluebird is actively infringing on X Corp.’s trademarks by publicly positioning itself to take over the world-famous branding. While Musk publicly remarked in 2023 that X would “bid adieu to the Twitter brand and, gradually, all the birds,” the USPTO and the U.S. District Court for the District of Delaware will soon be tasked with determining whether the marks are deemed “abandoned.”

Key Elements of Trademark Abandonment: Non-Use and Intent Not to Resume Use

The Lanham Act provides that a mark is deemed abandoned “[w]hen its use has been discontinued with intent not to resume use.” In practical terms, a party claiming abandonment must establish two elements: (1) the mark’s non-use and (2) intent not to resume use.

Courts may infer intent to use from the circumstances. Importantly, any “use” must constitute a bona fide use in commerce — that is, use made to identify the source of goods or services and not merely to reserve rights in the mark. The Lanham Act further provides that three consecutive years of non-use creates a rebuttable presumption of abandonment, shifting the burden to the trademark owner to produce evidence of continued use or intent to resume use.

Beyond statutory abandonment, trademarks may also be abandoned in other ways. For example, owners risk losing marks when they become a generic identifier of goods or services, when the associated goods or services are discontinued, or when the owner neglects to police or protect its rights.

An Owner’s Rebuttal: Proving Use or Intent to Resume Use

A trademark owner’s strongest chance of surviving an abandonment claim is evidence of the mark’s current use in commerce, which may include advertising or marketing materials, website screenshots, product packaging or labels, social media promotion tied to sales, or shipping and transaction records. Even where a mark has fallen out of active commercial use, an owner may rebut abandonment by demonstrating a documented and objective intent to resume use. Such evidence can include business plans referencing the mark, product development timelines, licensing negotiations, draft marketing materials, or internal communications discussing relaunch plans.

What Happens After Abandonment?

Once a trademark is deemed abandoned, for all practical purposes, the owner loses all rights associated with the mark. Any protection tied to prior use is extinguished, and the mark generally returns to the public domain — available for adoption and registration by others. While a former owner may attempt to revive the mark, prior registrations typically provide little advantage. Abandonment is, in most cases, permanent.

Potential Abandonment of the Iconic Blue Bird

Despite X Corp.’s successful renewal of its “Twitter” trademark in 2023, Musk and the X Corp. formally dropped the Twitter branding in July of that same year. Operation Bluebird is thus poised to argue that, as of July 2026, X Corp. has presumptively abandoned the mark due to three years of non-use.

Notably, X Corp.’s renewal filing included a screenshot of a “Twitter Ads” webpage featuring both the “Twitter” word mark and the iconic blue bird logo.

Today, however, that same webpage — ads.twitter.com — appears without either mark. While a Google search still displays the hyperlink title “Advertise on X (Twitter)” and X Corp.’s new suit indicating that users still access the platform through the domain “twitter.com,” such residual references may not qualify as bona fide trademark use. X Corp. also proffered within its new suit that they still actively defend and maintain the Twitter trademarks, but those efforts still may fall short.

Given the marks’ apparent absence from active branding, the USPTO and Delaware District Court will need to determine whether X Corp. can demonstrate a concrete and ongoing intent to resume use of the Twitter marks. Perhaps revival plans exist behind the scenes or perhaps the blue bird has truly flown the coop.

If X Corp. fails to muster up the evidence and the marks are deemed abandoned, maybe a new social media tycoon may well be born from Twitter’s discarded feathers. Until then, we can perch by our windowsills and watch how it all unfolds.

Listen to this post

If you’ve spent time in the trenches of a tech deal (as a buyer, a seller, or an advisor), you know that intellectual property (IP) representations and warranties can be a minefield. They look simple on the surface: The seller represents that they own the IP, have the right to use it, and that no one else is infringing or claiming rights. Easy enough, right?

But in the fast-moving world of technology, those IP clauses can carry big risks. A single poorly defined representation can turn into a post-closing headache or even a full-blown dispute. Let’s break down what IP representations and warranties really mean in tech deals, the common pitfalls to watch for, and how to keep your deal from turning into a last-minute scramble.

The Risk with “Standard” IP Representations

Most deal templates start with a “standard” set of IP representations. They usually include statements like:

  • The company owns or has sufficient rights to all IP used in its business.
  • No third party is infringing the company’s IP.
  • The company’s products don’t infringe anyone else’s IP.

These sound harmless (maybe even necessary), but the problem is that in tech deals, standard often means overbroad. A typical rep might say something like, “The company owns or has valid rights to all IP necessary to operate the business.” That might be fine if you’re selling a single product that was built entirely in house. But if your business touches multiple product lines, uses open-source components, relies on third-party APIs, or licenses content from vendors, that clause can sweep in assets outside the scope of the deal.

For example: Imagine a company selling its e-commerce platform division. The IP representations might inadvertently cover the company’s internal HR software, its data analytics engine, or even shared brand assets that aren’t part of the sale. Suddenly, a representation intended to cover “deal assets” has ballooned to include unrelated tech.

When reviewing IP representations, pay attention to scope.

  • Define what exactly is included in “intellectual property” or “IP” and include a schedule listing all intellectual property included in the deal.
  • Limit representations to “IP owned or used in connection with the business” being sold — not the entire company.
  • Be clear about what “the business” actually means in the definitions section.
  • Confirm that shared assets, like code libraries or trademarks used across divisions, are carved out or properly licensed post-close.

If you’re the seller, you want to make sure you’re not inadvertently representing ownership of IP that isn’t being sold. If you’re the buyer, you want comfort that the assets you’re acquiring are actually covered by the representations and warranties, but not so broad that you inherit unrelated risk.

Carveouts and Limitations: “Material,” “Knowledge” and the Art of the Qualifier

Once the basic scope is in place, the next question is how robust should the representations be? That’s where carveouts and qualifiers come in. Without qualifiers, IP representations can be dangerously absolute.

For example: “No third party has claimed that the company’s products infringe any IP rights.” That sounds good, but what if some random patent troll sent a demand letter two years ago that everyone ignored? Without a qualifier, that’s technically a breach, even if the claim was baseless.

However, the scope of the rep can be reined in by adding: “To the company’s knowledge, no third party has claimed in any material respect that the company’s products infringe any IP rights.” Now you’ve narrowed it to actual or constructive knowledge of key personnel and filtered out minor or immaterial issues.

Here are a few qualifiers that you’ll see in almost every tech deal:

  • “Material” or “material adverse effect”: limits representations to issues that would actually matter to the business.
  • “Knowledge”: restricts the rep to what management actually knows (or should reasonably know).
  • “Except as set forth on Schedule [X]”: lets the seller disclose known exceptions, like ongoing IP disputes or open-source dependencies.

Buyers generally want representations to be as tight as possible, while sellers push for these qualifiers to reduce risk. Sometimes, the appropriate qualifiers can depend on the size of the deal. The sweet spot is somewhere in between: meaningful protection for the buyer without setting traps for the seller.

Pro tip: If you’re the seller, resist the temptation to gloss over open-source use, pending IP claims, or third-party licenses. Disclose them clearly on the schedules. Buyers appreciate transparency, and properly disclosed issues usually won’t count as a breach later.

Post-Closing IP Issues: The Work Doesn’t Stop at Signing

Even with the cleanest IP representations, your work isn’t done when the ink dries. In tech deals, post-closing IP housekeeping is critical and often where problems surface.

1. Updating and Assigning IP Agreements

    After closing, someone has to make sure all the underlying IP agreements are properly updated or assigned. That includes:

    • Updating license agreements to reflect the new ownership structure.
    • Reassigning employee inventions and contractor IP to the right entity.
    • Recording patent, trademark, and copyright assignments with the U.S. Patent & Trademark Office or the U.S. Copyright Office.
    • Updating domain registrations, source code repositories, and cloud accounts with the new ownership and contact information.

    If these details slip through the cracks, the buyer could end up using assets they technically don’t own, which can create big problems down the line, especially if you need to enforce IP rights or bring in investors.

    2. Revising Licenses and Commercial Agreements

    Tech businesses rely heavily on licensed-in technology — APIs, SDKs, databases, or software libraries. Post-closing, those licenses often need to be revised or re-papered to reflect the new structure. You might find that a license was “non-assignable,” meaning it doesn’t automatically transfer with the sale. In that case, you’ll need a new agreement or consent. Skipping this step could mean you’re technically in breach of a key vendor contract.

    3. Monitoring for Indemnity Triggers

    Most purchase agreements include IP indemnities — promises that if someone sues claiming infringement, the seller will step in and cover it. Post-closing, buyers should keep an eye out for any IP claims or demand letters that might trigger these provisions. The key is to act quickly: Most indemnity clauses require prompt notice to the seller, and missing that deadline can waive your rights. Sellers, on the other hand, should keep tabs on what’s happening with the IP they sold. If the buyer starts using the technology in new ways, say integrating it into a new platform or launching in new markets, those uses might not be covered by the original indemnity.

    Wrapping It Up: Practical Tips for Tech Deal IP Representations

    When it comes to IP representations and warranties, clarity beats coverage every time. Overbroad, absolute language might feel protective, but in practice it just creates uncertainty and post-closing risk. Here are a few practical takeaways:

    1. Define the “business” precisely. Make sure IP representations only apply to assets used in the business being sold.
    2. Use qualifiers wisely. “Material” and “knowledge” can save both sides headaches, but overuse can make representations meaningless.
    3. Schedule everything. If there’s an open-source dependency, a pending dispute, or a third-party license, put it on the disclosure schedule.
    4. Follow through post-close. Update assignments, revise licenses, and monitor for indemnity issues. Deals don’t end at signing; they just move into maintenance mode.
    5. Collaborate between deal teams and engineers. Legal teams often draft representations in isolation, but engineers know where the IP skeletons are buried. A quick internal audit before signing can prevent surprises later.

    At the end of the day, IP representations and warranties aren’t just legal boilerplate — they’re the connective tissue of a tech deal. They ensure the buyer actually gets the technology they’re paying for, and that the seller isn’t on the hook for risks they didn’t intend to assume. Handled thoughtfully, these clauses protect both sides and keep the focus where it belongs: growing the technology, rather than litigating over it.

    Listen to this post

    The Federal Circuit’s recent decision in Canatex Completion Solutions, Inc. v. Wellmatics, LLC is a good example of the courts’ ability to correct obvious errors in patent claims through claim construction. Indeed, the panel’s reversal of the lower court’s ruling demonstrates that judicial correction of evident errors in patent claims — even when the error appears multiple times throughout the patent document — is available to patentees when only one reasonable correction exists.

    The Technology and the Typographical Trap

    The technology at issue in U.S. Patent No. 10,794,122 centers around a releasable connection device used in oil and gas well operations. More specifically, the two-part downhole tool allows operators to disconnect components when the lower part of the tool becomes stuck in the wellbore, thus enabling retrieval of the upper part while leaving the stuck component for later recovery. The claimed connection device includes a “first part” (further downhole in the wellbore) with an external connection profile and a “second part” (closer to the surface) containing a releasable engagement profile that internally engages the first part’s connection profile. Claim 1 of the ʼ122 patent reads as follows:

    1. A releasable connection for a downhole tool string, comprising

    a first part comprising an external connection profile; and

    a second part comprising:

    an outer housing;

    a releasable engagement profile which internally engages the connection profile of the first part and which is configured to expand radially to release the connection profile of the first part;

    a locking piston positioned within an internal cavity of the second part, the locking piston configured to move axially along the second part between a locking position that directly constrains the releasable engagement profile into engagement with the connection profile of the first part and a release position that permits the releasable engagement profile to expand radially to release the connection profile of the second part;

    an expansion chamber in fluid communication with the locking piston; and

    a source of fluid pressure in communication with the expansion chamber, wherein, upon activation, the source of fluid pressure is configured to apply fluid pressure to move the locking piston from the locking position toward the release position.

    The remaining two independent claims also contain the italicized language in claim 1 above that dictates how a locking piston moves between positions to control the connection between the first and second part. In the release position, the piston “permits the releasable engagement profile to expand radially to release the connection profile of the second part.” The problem is that the claims never previously mentioned any “connection profile of the second part” — only a “connection profile of the first part.”

    The District Court’s Indefiniteness Determination

    Boring down on this antecedent basis problem during claim construction, defendants Wellmatics and GR Energy Services argued the claims were invalid for indefiniteness due to “the connection profile of the second part” lacking any prior reference in the claim. Canatex acknowledged the error but contended it had a simple, obvious correction: changing “second” to “first” so the phrase would refer to “the connection profile of the first part” — the only connection profile actually described in the claims.

    The District Court for the Southern District of Texas disagreed and held all asserted claims of the ʼ122 patent invalid for indefiniteness. The court reasoned that the error was not evident from the patent’s face because it appeared pervasively throughout the ʼ122 patent — in all three independent claims, in the abstract, and twice in the written description. This pervasiveness suggested “an intentional drafting choice and not an error at all.” In fact, the court speculated that the drafter may have intended to provide an antecedent basis for a connection profile of the second part, making the error the use of “the” instead of “a.” The court also took issue with Canatex’s failure to seek correction from the USPTO under 35 U.S.C. § 255, which suggested the error was “neither minor nor evident.”

    The Federal Circuit’s Reversal: A Rigorous Standard Met

    Applying the demanding standard for judicial correction of claim errors, the appellate panel reversed the district court. The opinion articulates several necessary requirements for judicial correction via claim construction:

    1. The error must be evident from the patent’s face — obvious from the perspective of one skilled in the art;
    2. The correction cannot be subject to reasonable debate based on the claim language, specification, and prosecution history; and
    3. The correction must involve only obvious minor typographical or clerical errors.

    The demanding standard for judicial correction serves multiple purposes. First, it respects the public notice function of patents by ensuring corrections are limited to situations where relevant readers would constructively understand the intended meaning. Second, it distinguishes judicial correction from USPTO correction under Section 255, which covers broader classes of errors and operates only prospectively because it can result in claim scope different from what readers would have understood. The panel’s opinion emphasized that while the standard is exacting, it was clearly satisfied here.

    Why the Error Was Evident

    The Federal Circuit found the error obvious on multiple grounds. Starting with the claim language itself, the panel explained that a skilled artisan would immediately recognize that the phrase “the connection profile of the second part” requires but lacks an antecedent. No connection profile of the second part appears anywhere earlier in the claim. In addition, the reference to the second part in this context was not deemed to make any functional sense. The claim structure requires the second part’s “releasable engagement profile” to engage with the first part’s connection profile by embracing it internally. Upon activation, the locking piston moves to a release position that permits the releasable engagement profile to expand radially. According to the appellate panel, the idea that this expansion would release an unidentified component of the second part itself — rather than the first part’s connection profile that it embraces — is “jarringly anomalous.”

    Turning to the specification of the ʼ122 patent, the appellate panel found that it confirmed the error’s obviousness. In particular, the figures lack a “connection profile” in the second part. In addition, a portion of the written description states that the locking piston “locks releasable engagement profile 26 into engagement with connection profile 16 of the first part, and a release position… that permits releasable engagement profile 26 to release connection profile 16 of second part 14.” This portion both correctly identifies connection profile 16 as belonging to the first part and incorrectly attributes it to the second part — making the error undeniable to a relevant reader.

    Why Only One Reasonable Correction Exists

    The panel found that changing “second” to “first” in the phrase at issue was the only reasonable correction because the claim language and specification both show that what gets released through radial expansion is the connection profile of the first part — the very component the releasable engagement profile embraces in the connected state. In this regard, the Federal Circuit rejected the proposed alternative corrections from defendants as unreasonable. For example, one proposed correction from the defendants echoed the district court’s view that the device contains an unnamed “connection profile” on the second part, so the claim could be corrected by changing “the” to “a” or adding antecedent language. In making this proposal, defendants pointed to “tubing connector 22” as a possible connection profile of the second part. The panel found this proposal to be illogical given that tubing connector 22 sits at the far proximal end of the second part and serves to connect the device to the wireline extending to the surface, and not to connect or disconnect the device’s two parts. The panel likewise dismissed defendants’ expert’s opinion that the varying language across instances (sometimes adding figure reference numbers) suggested deliberate drafting rather than copy-paste error, noting the variations merely added figure labels without changing component understanding. Defendants remaining proposed alternatives were deemed by the panel to not be reasonably debatable alternatives to the simple, logical change of one word because each would substantially alter claim scope and meaning beyond what the claims and specification clearly describe.

    Key Takeaways

    The case now returns to the district court for further proceedings with the corrected claim construction. With “second” properly read as “first,” the parties can now address infringement and any remaining validity issues on the merits. Regardless of how it turns out, this decision reinforces some key takeaways for navigating claim construction and adjacent validity challenges:

    Correction Must Be Textually Simple

    The correction here changed a single word without altering claim structure or adding new limitations. More complex corrections that significantly revise claim language or add substantial new text will not qualify as “minor” corrections even if otherwise appropriate.

    Pervasive Errors Do Not Preclude Correction

    The district court’s conclusion that repeated errors suggest intentional drafting was rejected by the Federal Circuit. While obviously not optimal, consistent copy-paste errors throughout a patent document do not transform an obvious mistake into deliberate language choice. A court should focus on whether the error and its correction are evident to skilled artisans, not on how many times the error appears.

    Functional Context Reveals Errors

    The Federal Circuit emphasized how the claim’s functional description made the error obvious. When claim language describing mechanical interactions makes no physical or logical sense as written, a simple correction to restore coherent meaning consistent with the specification might be the answer instead of invalidation under 35 U.S.C. § 112.

    Specification Passages Can Confirm Errors

    Self-contradicting specification language — such as the passage correctly identifying a component as belonging to the first part while later attributing it to the second part — provides strong evidence that similar claim language contains the same error.

    Failure to Seek USPTO Correction Is Not Dispositive

    The district court gave weight to Canatex’s failure to pursue Section 255 correction before litigation. The Federal Circuit’s reversal indicates this factor should not preclude judicial correction when the demanding standard is otherwise met. Indeed, Canatex did seek USPTO correction during the appeal, which the PTO denied — but this did not prevent the Federal Circuit from making the correction judicially.

    Prosecution History Matters But Was Absent Here

    The court confirmed that prosecution history can preclude correction if it suggests a different claim interpretation. However, neither party identified anything in the prosecution suggesting the disputed language should be read other than with the obvious correction. This distinguishes the case from Chef America, Inc. v. Lamb-Weston, Inc. where prosecution history confirmed the patentee’s claim language was problematic.

    In short, Canatex demonstrates that the claim construction process remains flexible enough to correct obvious mistakes when doing so gives effect to the meaning skilled artisans would clearly understand from the patent as a whole. Accused infringers considering indefiniteness arguments should carefully evaluate whether claimed errors are genuinely ambiguous or merely typographical before getting too comfortable relying on this type of invalidity challenge. More promising indefiniteness challenges will involve claim language where multiple reasonable interpretations exist or where the specification provides no clear resolution.

    Listen to this post

    In a pair of recently issued companion decisions, the Federal Circuit awarded Causam Enterprises, Inc. a victory on patent ownership (Causam v. ITC (ITC appeal)) only to pull the rug out from under the company by declaring the very patent claims at issue unpatentable (Causam v. ecobee (IPR appeal)). Unfortunately for Causam, this is a shining example of winning the battle but losing the war.

    The saga centers on U.S. Patent No. 10,394,268, which relates to demand response functionality for electric power grids designed to help utilities balance supply and demand. Causam filed a complaint with the International Trade Commission alleging that respondents Resideo Smart Homes Technology, its domestic affiliate Ademco, Inc., and other respondents imported smart thermostats that infringed the ʼ268 patent. At the same time, ecobee Technologies, which also had been named as a respondent in the ITC complaint, petitioned for inter partes review of the patent claims before the Patent Trial and Appeal Board (PTAB) (IPR2022-01339). The two proceedings wound their way through the system in parallel, ultimately landing before the Federal Circuit on the same day for oral argument.

    Causam has asserted the patent against a range of smart thermostat and energy management companies, including ecobee Technologies ULC and Resideo Smart Homes Technology (Tianjin).

    Round One: The ʼ268 Patent Ownership Drama

    In the ITC proceeding, the respondents challenged whether Causam even owned the ʼ268 patent — a threshold question intended to defeat Causam’s standing to even bring the ITC case. The ownership dispute turned on a series of agreements involving inventor Joseph Forbes. In 2007, Forbes assigned the ancestor application (the ʼ909 application) to America Connect, which later became Consert. That assignment covered “all divisions, reissues, continuations and extensions thereof” but conspicuously omitted continuation-in-part applications.  

    So, here is where it gets interesting. The ʼ268 patent claimed priority to the ʼ909 application through a continuation-in-part — the ʼ761 application, which was filed in 2012. A 2013 settlement between Forbes and Consert listed the ʼ761 application as an “Excluded Patent.” Forbes then assigned the ʼ268 patent to Causam in 2014.

    Despite the lack of any reference to continuation-in-part applications in the 2007 assignment to Consert, the administrative law judge (ALJ) ruled that the assignment to Consert from Forbes covered “all progeny” and, thus, Causam could not and did not own the ʼ268 patent.  The ALJ also found that accused smart thermostats did not infringe the asserted claims of the ʼ268 patent. The full commission adopted the ALJ’s infringement finding, but punted on the ownership issue.

    On appeal, the Federal Circuit addressed whether Causam had Article III standing to challenge the commission’s decision — which required determining if Causam actually owned the patent. The court held that at this stage (after a full evidentiary record), Causam needed to provide evidence of ownership, not just assert it. Citing its tandem decision in the IPR appeal, the appellate panel explained that the burden is “the same as that of a plaintiff moving for summary judgment in the district court.”

    The panel then dove right in to contract interpretation. Despite intervenors arguing that “continuations” in the 2007 assignment should be read to include continuations-in-part, the Federal Circuit explained that continuations and continuations-in-part are “widely understood to be different” and have been treated separately in the Manual of Patent Examining Procedure since 1948. Critically, a continuation may not add new matter to the parent application, while a continuation-in-part does include new matter. Indeed, an inventor “might well wish to assign away continuations of an invention and keep the rights to continuations-in-part (and the included new matter).”

    The panel further opined that reading “continuations” to include continuations-in-part would amount to “insert[ing] words into the contract that the parties never agreed to.” On this basis, the court concluded that the 2007 assignment unambiguously excluded the ʼ761 application, so Forbes had good title when he assigned the ʼ268 patent to Causam in 2014. Victory for Causam, right? Well, not so fast.

    Round Two: The IPR Appeal Delivered the Knock-Out Punch

    While Causam won on the ownership issue, the holding in the companion IPRappeal mooted the noninfringement issue in the ITC appeal. In this aspect, the panel declined to address Causam’s challenge to the commission’s noninfringement determination in the ITC appeal based on its affirmation of the PTAB’s holding that all challenged claims of the ʼ268 patent — including claim 1 (the only claim at issue in the ITC appeal) — were unpatentable as obvious in the IPR appeal. Claim 1 reads as follows:

    1. A method for managing an electric power flow within an electric power grid, comprising:

    a client device receiving a power control message from a load management server, the power control message indicating at least one of an amount of electric power to be reduced and an identification of at least one controllable device to be instructed to disable the electric power flow to at least one associated power consuming device;

    the client device issuing a power management command to the at least one controllable device, the power management command causing the at least one controllable device to disable the electric power flow to the at least one associated power consuming device to provide a reduction in consumed power; and

    generating measurement and verification data corresponding to the reduction in consumed power.  

    The italicized claim limitation was front and center in the IPR appeal. During the IPR, Causam argued this required generating data during the demand response event. But, the PTAB disagreed, construing the limitation to encompass both actual measurements during the event and estimates based on device power usage measured before the event. 

    The Federal Circuit sided with the PTAB explaining that the claim language itself imposed no temporal restriction and that adopting Causam’s narrow construction would exclude embodiments disclosed in the specification. For example, one embodiment described a system that determines “the amount of steady-state power each device consumes when turned on” and uses that known consumption data when deciding which loads to turn off during high-demand periods.  Those “known” power-consumption loads are estimates obtained before any demand response event occurs — directly contradicting Causam’s proposed limitation. Another embodiment described accessing a database after a power savings event ends to find “the actual amount (or average amount) of power that would have been used” during the event. With the panel’s finding that claim 1 of the ʼ268 patent had been properly construed by the PTAB, Causam’s obviousness challenge collapsed and the PTAB’s unpatentability determination stood.

    Because Section 337 — the basis for the ITC action — requires infringement of a “valid and enforceable” patent, Causam had no right to an exclusion order. The panel thus proceeded to dismiss the ITC appeal as moot, leaving Causam with nothing but an empty victory on the ownership question.

    The Takeaways

    Parallel proceedings can create strategic nightmares. Causam fought hard to establish ownership in the ITC case, only to watch the patent claims evaporate in the IPR. The company ended up with clear title to an unpatentable patent — a trophy nobody wants. However, these companion appeals at least offer several drafting lessons:

    • Be careful what you disclose in your specification. Causam’s broad embodiments — using estimates and measurements taken before or after demand response events — undermined its attempt to narrow the claims during litigation. The Federal Circuit will not let you exclude what you have already clearly described as within the scope of the invention without something more in the prosecution history.
    • When preparing assignment agreements, be specific. The distinction between continuations and continuations-in-part matters, and courts will not read one term to include the other. If you want comprehensive coverage, say so explicitly.