by Dennis Crouch
The Solicitor General has filed a brief urging the Supreme Court to deny certiorari in Thaler v. Perlmutter, the case testing whether artificial intelligence can be recognized as an author under copyright law. The government's opposition takes a notably restrained approach, relying entirely on statutory interpretation while avoiding the policy arguments about innovation incentives and economic impact that have dominated public discourse. The brief includes a strategic reframing of the question presented -- attempting to narrow the case its unusual facts while leaving open the genuinely difficult questions about human-AI collaboration that will define copyright's future in the near term. Brief for the Respondents in Opposition, Thaler v. Perlmutter, No. 25-449 (U.S. Jan. 2026).
The case arises from computer scientist Stephen Thaler's application to register copyright in "A Recent Entrance to Paradise," a visual artwork generated by his AI system called the "Creativity Machine." Thaler's application listed the Creativity Machine as the sole author and stated the work was "created autonomously by machine" without any human creative contribution.
The Trump Administration brief is notable for its restrained approach, relying entirely on statutory text and structure without invoking policy arguments about innovation or economic impact. At the same time, the government's framing of the human authorship requirement appears more permissive than the Copyright Office's recent registration decisions have suggested. The brief quotes the district court opinion for the proposition that that copyright protection remains available for AI-assisted works in very broad terms:
The rule requires only that the author of that work be a human being—the person who created, operated, or used artificial intelligence—and not the machine itself.
This language could be read to encompass a broader range of AI users than the Office has previously accepted.
In these situations, the US Gov't brief will be signed by lawyers from both the DOJ and Copyright Office. In this case, we only see DOJ attorneys, which suggests some disagreement in approach. Of course, the Trump Administration (in control of the DOJ) is currently in litigation to remove the Copyright Registrar Shira Perlmutter, and so a lack of cooperation is not surprising.
The USPTO classifies patent applicants into three entity categories: large entities (regular undiscounted), small entities (businesses with fewer than 500 employees, independent inventors, nonprofits), and micro entities (those meeting additional income and filing history limits). These classifications determine fee structures, with small entities receiving a 61% discount and micro entities an 80% discount on most USPTO fees. These fee categories create an convenient mechanism for categorizing patent applicants and their patenting activity.
I analyzed USPTO disposition data for published patent applications from January 1-27, 2026, examining applications that either issued as patents or were abandoned during this period. Frankly, I was surprised by the disparity. Overall, 74% of disposed applications resulted in patents. But the allowance rate varies dramatically by entity size: large entities achieved an 80% allowance rate, small entities 61%, and micro entities just 40%.
Comcast is pushing its proper venue argument. After being denied by the district court and a Federal Circuit panel, it has now petitioned for en banc rehearing. In re Comcast Cable Communications, LLC, No. 26-104 (Fed. Cir. Dec. 9, 2025).
The underlying dispute presents a question the appellate court seemingly has not squarely addressed: whether the “acts of infringement” prong of 28 U.S.C. § 1400(b) requires a plaintiff to allege that the defendant performed all steps of a claimed method within the judicial district, or whether alleging the performance a subset suffices.
Patent Venue Under § 1400(b): The patent venue statute provides that a civil action for patent infringement "may be brought in the judicial district where the defendant resides, or where the defendant has committed acts of infringement and has a regular and established place of business." 28 U.S.C. § 1400(b). After TC Heartland LLC v. Kraft Foods Group Brands LLC, 581 U.S. 258 (2017), a domestic corporation "resides" only in its state of incorporation. For defendants sued outside their home state, plaintiffs must establish both "acts of infringement" and a "regular and established place of business" in the chosen forum.
Post TC Heartland (2017), I think of Section 1400(b) as being quite narrow. But its 1897 origin was not as a restriction on patent plaintiffs but as an expansion of their options. At the time, the general venue provision limited civil suits to the district where the defendant was an “inhabitant.” Patent holders complained that this forced them to travel across the continent to sue infringers operating businesses in their own backyards. The 1897 statute responded by adding a second venue option: districts where the defendant had committed acts of infringement and maintained a regular and established place of business. This gave patentees meaningful access to forums where infringement was actually occurring.
Since then, however, the general venue statute has expanded greatly so that it now (basically) encompasses any district where the defendant is subject to personal jurisdiction. What began as a plaintiff-friendly expansion became, through statutory drift, more of a defendant’s shelter.
Docs:
Comcast began when Sandpiper sued Comcast in the Eastern District of Texas for infringement of five method patents related to content delivery network (CDN) technology. For Sandpiper, EDTex venue was proper because Comcast had committed acts of infringement in the district and also had offices located there (regular and established place of business).
Comcast admits that it does operate in EDTex and may perform some of the claimed steps, but argues none of the asserted claimed methods are fully performed within EDTex. Since plaintiffs have the burden of showing proper venue, Comcast’s legal argument is that Sandpiper had not alleged that all steps of any claimed method was performed within the district.
The district court sided with the patentee — holding that venue is proper so long as the plaintiff alleges that the defendant performed at least one step of the claimed method in the forum district. Comcast then petitioned for mandamus but that was denied based upon adequacy of a post-judgment appeal. Comcast has now asked the full court to reconsider its case.
The substantive question is interesting and important. Comcast’s argument rests primarily on the Federal Circuit’s 2005 decision in NTP, Inc. v. Research In Motion, Ltd., 418 F.3d 1282 (Fed. Cir. 2005). In that case, the focus was on infringement – asking when method claims were infringed “within the United States” under 35 U.S.C. § 271(a) when some steps occurred abroad. The court held that infringement of a US process patent requires each and every step be performed within the US. This is bolstered by Limelight Networks v. Akamai, 572 U.S. 915, 921 (2014), where the Supreme Court explained that “[a] method patent… is not infringed unless all the steps are carried out.”
Comcast argues that NTP‘s logic regarding § 271(a) applies with equal force to § 1400(b). The venue statute asks where the defendant “has committed acts of infringement.” If an “act of infringement” of a method claim requires performance of all steps (as NTP holds for § 271(a)), then venue should likewise require all steps to occur in the district. The textual argument has considerable force. That said, when I first read the statute I thought of “acts” as in “acts of a play,” where each act is a component part of the larger dramatic work rather than the complete performance itself. Under this alternative reading, the statutory phrase “acts of infringement” would refer to the constituent steps or components that together make up the complete infringement, not to fully consummated infringing acts. In their opposition brief, Sandpiper argues Congress deliberately chose “acts of infringement” rather than just “infringement”; intending to differentiate between “constituent conduct” and “completed violation.” The brief cites 18 U.S.C. § 2331(4)’s definition of “act of war” as meaning “any act occurring in the course of . . . declared war” as a parallel statutory usage. Sandpiper also tries to distinguish NTP’s holding as being driven by extraterritorial concerns rather than domestic venue allocation. Different policy considerations allow for an arguably different interpretive approach.
The district courts appear divided. Compare AML IP, LLC v. Bath & Body Works Direct, Inc., 2024 WL 3825242 (E.D. Tex. Aug. 13, 2024) (NTP controls; plaintiff must allege performance of all method steps within the district); SEVEN Networks, LLC v. Google LLC, 315 F. Supp. 3d 933 (E.D. Tex. 2018) (one step rule; venue is proper so long as the plaintiff alleges that the defendant performed at least one step of the claimed method in the forum district). That said, Sanpiper notes that the circuit split is quite one-sided with AML appearing as the only EDTex case following an all-the-steps rule.
I find myself skeptical of Comcast’s position when taken to its logical extreme. The patent venue statute is a creature of the 19th century, enacted in an era when the geographic distribution of infringing activity was far simpler to conceptualize. A factory making infringing widgets either operated in a particular district or it did not. Method patents claiming distributed computing processes present a fundamentally different situation. When a CDN operates by distributing content from servers across dozens of districts to users in all fifty states, no single district may contain all the claimed steps. Under Comcast’s “all steps” rule, the “acts of infringement” prong of § 1400(b) would become a nullity for such technologies. The only available venue would be the defendant’s state of incorporation. Congress provided two bases for venue, and an interpretation that collapses them into one for an entire category of patents seems to go too far.
Note: Sandpiper’s opposition brief cites Valeant Pharms. N. Am. LLC v. Mylan Pharms. Inc., 978 F.3d 1374 (Fed. Cir. 2020) and Celgene Corp. v. Mylan Pharms. Inc., 17 F.4th 1111 (Fed. Cir. 2021) for the proposition that an “act of infringement” may encompass activities that are only “part of” the completed infringement. Those are ANDA cases where § 271(e) creates “paper infringement” (filing of the ANDA with the FDA constitutes the infringing act). In Valeant, the court reserved the question of what preparatory activities might establish venue, and Celgene suggested that acts “fairly . . . part of the submission” could suffice even if they don’t constitute the complete infringement. But the ANDA context is arguably distinguishable. There, the infringement is a single unified act (submitting the ANDA) and the question is whether preparatory work contributing to that act occurred in the district. Comcast’s situation is structurally different: the question isn’t whether preparatory activities for a single act occurred locally, but whether performing one discrete step of a multi-step method claim is sufficient when other steps occur elsewhere.
= = =
Odds are against the court granting en banc hearing of the mandamus action. However, the Federal Circuit will eventually have to resolve this question, whether now or in a future appeal from final judgment. The underlying tension reflects one that has long been part of patent law: how to apply statutory language drafted for a different technological era to innovations that its drafters could not have imagined.
An amicus brief filed by the US Chamber of Commerce argues that adds some data that forum shopping is a huge concern. And, in particular that NPEs are filing 60%+ of patent suits, with heavy EDTex concentration.
]]>by Dennis Crouch
The Federal Circuit's recent nonprecedential decision in Guardant Health v. University of Washington focuses on single-reference obviousness findings.
The vast majority of obviousness cases are based upon two or more references that, when combined, teach each limitation of the claimed invention. The key question in those cases boils down to whether a person of ordinary skill in the art would have had a sufficient motivation to combine those references to form the invention as well as a reasonable expectation of success in that endeavor. In the single reference situation, the Federal Circuit in Guardant Health found that neither of those test make sense - particularly when elements appear together in a single embodiment in the prior art.
by Dennis Crouch
IPR institution rates under Director John Squires have begun to climb. My tracking data shows the six-week moving average rising from lows around 15% in August and September 2025 to somewhere in the range of 35-55% by late December 2025 and into January 2026. The rebound remains far below the historic average of roughly 67% that prevailed before Acting Director Coke Morgan Stewart began centralizing discretionary denial authority followed by Director Squires assumption of personal control of all institution decisions in October 2025.
Still, the uptick signals something meaningful: the IPR system is not dead, and a discernible framework for institution has begun to emerge from what initially appeared to be an era of blanket denials. The patterns in recent grants reveal that Director Squires is operating with something resembling an equitable framework that requires petitioners to go beyond the statutory "reasonable likelihood" threshold to encompass broader considerations of fairness and systemic efficiency. The USPTO's recent designation of numerous institution decisions as precedential or informative in early January 2026 provides additional clarity about the principles the Office considers settled enough to guide future practice.
by Dennis Crouch
The Federal Circuit has a new expert witness case -- this time siding with the patentee and finding that two experts were improperly excluded mid-trial. But, the case includes a sharp dissent over where to draw the line between a court's Daubert gatekeeping function and the jury's role as factfinder. Barry v. DePuy Synthes Companies, No. 2023-2226 (Fed. Cir. Jan. 20, 2026).
In the case, the experts had presented testimony that strayed somewhat from the court's claim construction and also included survey results using questionable methodology. Writing for the majority, Judge Stark concluded that this testimony was (1) permissible applications of the court's claim construction and (2) that methodological criticisms went to evidentiary weight rather than admissibility. Judge Prost dissented, warning that the majority's approach "contravenes the principles embraced in EcoFactor and the 2023 amendments" to Federal Rule of Evidence 702 and "will undermine district courts' abilities to exercise their important gatekeeping function."
The dispute in this case highlights a conceptual tension in how we think about expert testimony standards. All testimony must clear a basic relevance threshold to be admissible. At the other end of the spectrum, we have a higher standard for actually carrying the burden of proof and sustaining a verdict. Expert testimony occupies middle ground. Rule 702 and Daubert impose reliability requirements that exceed ordinary admissibility standards, recognizing that experts wield particular influence over juries. But, where exactly expert testimony sits on this spectrum (and who decides) is the central tension between the majority and dissent.
by Dennis Crouch
I only recently started using mobile check deposits. I don't get very many checks these days, but when I do it is a real pain to drag myself to the the bank to deposit -- particularly because I don't carry my ATM card with me. Great convenience -- and also a great patent story.
USAA is a financial services association cooperatively owned by U.S. military members, veterans, and their families. Because USAA serves military families whether stationed at home or deployed overseas to places like Iraq and Afghanistan, it does not operate a traditional branch network. This created a persistent problem: how could USAA members deposit checks when stationed far from any bank? Starting around 2005, a team at USAA began researching technology that would allow members to deposit checks from anywhere in the world using devices they already owned. The challenge was formidable: as of the 2006 priority date, the banking industry conventional wisdom was that specialized check-scanning equipment was the only way to reliably capture check images meeting the technical requirements for remote deposit under the Check 21 Act. USAA's engineers developed a system using a downloaded mobile app to assist customers in capturing compliant check images without specialized equipment, launching Deposit@Mobile in 2009 as the first mobile check deposit service offered by any U.S. bank. Within a month, it became the number one financial app in the Apple app store, with members depositing over 1.5 million checks in the first year.