Back in May 2023 I wrote a short post about the University of Texas study that used a GPT-style language model to reconstruct the gist of stories from fMRI scans (An API for the Human Mind, May 6, 2023). What I did not know at the time was that the UT System had already filed a provisional application on the method a full year earlier, fourteen months before the Nature Neuroscience paper. The resulting non-provisional application published as US 2025/0068841 A1, titled “Decoding Language from Non-Invasive Brain Recordings,” naming Alexander Huth and Jerry Tang as inventors. It allowed on the first action in June 2026, without a single rejection. Texas paid the issue fee on September 18, so a patent should issue later this fall. No continuation in the record yet, but I expect they’ll file one.

The basic idea here is that different words and ideas produce different patterns of neural activity across the brain, and fMRI can measure those patterns (via bloodflow) while a person listens to a story. Huth and Tang first trained an “encoding model” on many hours of scans.  Decoding then runs in the opposite direction.  A GPT-style language model proposes candidate word sequences; the encoding model predicts the brain activity each candidate should produce; and the system keeps the candidates whose predictions best match the actual scan. Repeat that word by word and the output is a running paraphrase of what the subject heard or imagined.  Ultimately, this results in the decoded output capturing the gist of the thought rather than exact wording.

Huth and Tang’s specification includes a section titled “Privacy Implications” explaining that their approach does not read unrestricted thoughts. But, there are an increasing number of researchers and companies staking out pieces of the path from brain signal to semantic representation to language model to text. In potentially revolutionary fields like this, there are often early precursors whose patent rights expire before commercialization. An example is Huth’s own doctoral advisor (Jack Gallant) whose patent recently expired for failure to pay a maintenance fee.

The published claim 1 is not simply “mind reading.” The claim recites a computer-implemented beam search:

• receive N word-sequence hypotheses;
• have a language model propose K continuation words for each;
• convert each of the N*K continuations, using an “encoding model,” into a predicted brain response across L measured brain regions or sensors;
• compare those predictions against the subject’s actual brain measurement for the current time period;
• keep the top N;
• repeat; and
• output a series of words.

Claim 14 separately covers building the encoding model itself, mapping neural-language-model word embeddings into voxel space. The dependent claims add fMRI (claim 2) and fNIRS (claim 3), and claim 5 covers a subject who “receives the stimulus by thinking, reading, or listening.”

As you might expect in the patent system, the claims go well beyond what they demonstrated in the academic paper. Tang et al., 26 Nature Neuroscience 858 (2023), trained on three subjects, each scanned for 16 hours while listening to stories, and worked only with fMRI. The claims cover fNIRS and a “different sensor.” The specification’s own privacy section reports that decoders trained on other people’s brains performed “barely above chance” and concludes that “subject cooperation remains necessary for decoder training.”

Gallant’s lab (mentioned above) filed a provisional in March 2009 that eventually issued as U.S. Patent No. 9,451,883, “Apparatus and Method for Decoding Sensory and Cognitive Information from Brain Activity.” Huth is not a named inventor, although his 2012 Neuron paper on semantic maps is cited as a reference. Claim 1 of that patent uses a similar encoding-model architecture: acquire training brain data; convert it into encoding-model parameters over “linearizing feature spaces”; build a “decoding database” of candidate items; generate a predicted brain activity signal for each item; compute the probability that a new scan corresponds to each prediction; select the best item; and reconstruct the stimulus.

Siemens has another similar patent that I was able to find. U.S. Patent No. 10,856,815, “Generating Natural Language Representations of Mental Content from Functional Brain Images,” associated with the “universal decoder” project discussed in Pereira et al., 9 Nat. Commun. 963 (2018). The government holds a confirmatory license through the Air Force, which funded the research. Claim 1 recites capturing fMRI data while a subject is exposed to training text that has been assigned semantic vectors; decomposing the images into a weighted set of “basis images”; mapping those weights to the dimensions of the semantic vector; decoding a semantic vector for a new scan; and “generating a text output based on the decoded semantic vector.” In 2008 Mitchell and Just were able to predict fMRI activation for concrete nouns from a text corpus (Mitchell et al., 320 Science 1191 (2008)), but I didn’t find a patent filing on point.

I also did some digging and found a 1967 Air Force research publication in which volunteers were trained to switch their brain alpha rhythms on and off (using eye movements) and then used that switch to send Morse code through an EEG. E.M. Dewan, Occipital Alpha Rhythm Eye Position and Lens Accommodation, 214 Nature 975 (1967). I’m not sure why they waited so long, but the Air Force roughly patented the concept in 2003 (U.S. Patent No. 6,529,773). The ‘773 patent claims a “mentally controlled optical communication system” in which the user’s altered alpha waves spell out Morse code and switch on a light transmitter. The specification incorporates the 1967 Nature paper by reference and quotes Martin Caidin’s 1968 novel The God Machine on Dewan’s “alpha adepts.”

Eligibility and the mental process irony. Section 101 excludes “mental processes,” meaning claims that could be performed in the human mind. Although these are claims about reading thoughts, none of them can be performed in the human mind.

The privacy question. Patent claims say what a decoder does with a brain signal. They say nothing about whose signal it is or who may use the output, and legislators have started to fill that gap. Colorado and California amended their consumer privacy statutes in 2024 to treat “neural data” as sensitive information. In September 2025, Senators Schumer, Cantwell, and Markey introduced the MIND Act, S. 2925, which would direct the FTC to study the governance of “information obtained by measuring the activity of an individual’s central or peripheral nervous system through the use of neurotechnology” and report to Congress within a year. It would also bar federal agencies from buying or operating neurotechnology inconsistent with forthcoming OMB guidance. But the MIND Act is not becoming law this session.

Requests for continued examination are climbing again. Applicants filed nearly 150k RCEs in published utility applications over the 12 months ending August 2026, up 13% from the prior 12 months. That reverses a five-year slide as seen in the chart below.

But, as I explain below, most of the increase is not a change in after-final applicant behavior. The choice to file an RCE arises at a final rejection, and the key change is that the USPTO has been mailing more final rejections – that accounts for most of the increase.  But, we also have the end of the free After Final Consideration Pilot (AFCP 2.0) in December 2024. At least some of the Applicants who once tried an AFCP amendment now go straight to an RCE.

It does not appear that the 2025 fee rule slowed things down significantly.  The fee change raised the first RCE fee 10% to $1,500 and second and later RCEs 43% to $2,860.  But the RCE filing rate is up for both first RCEs and subsequent RCEs.

That leaves the question of why the Office is mailing more final rejections: 55 per 100 disposals over the past year, up from 50.  Some of that rise is due to the shift from examining continuations to examining original applications. As I explained in my September 7 post, the Office pulled continuations forward in the examination queue beginning in 2022 and pushed them back in 2025, and original applications are more likely to receive final rejections compared with continuations. But this only explains part of the recent rise (and I don’t have a further explanation).

Guest post by Professor Paul R. Gugliuzza

Since Congress created the Federal Circuit in 1982, the court has repeatedly expanded the class of cases that, in its view, “arise under” patent law and so fall within its exclusive jurisdiction. The court’s recent decision in Insulet Corp. v. EOFlow Co. is another example. Insulet is a trade secret case. Yet the Federal Circuit held that the case arose under patent law based on its prediction about the outcome of a hypothetical future patent infringement dispute.

The court’s decision overlooks a fundamental principle of the law of federal jurisdiction: namely, that the “arising under” analysis looks solely at the plaintiff’s well-pleaded complaint, not to issues that may (or may not) arise in future litigation.

Some background. In Insulet, the district court entered a $60 million judgment for Insulet on a jury verdict of trade secret misappropriation. Insulet’s original complaint had asserted patent claims, too, but those claims were bifurcated from the trade secret claims, never adjudicated, and eventually dismissed without prejudice by an amendment to Insulet’s complaint.

EOFlow appealed its trial loss to the Federal Circuit. Insulet contested jurisdiction because its amended complaint contained no patent claims, citing a line of circuit precedent holding that, when all patent claims were dismissed without prejudice, the Federal Circuit lacks jurisdiction.

The court nevertheless heard EOFlow’s appeal on the merits and reversed the trade secret verdict in Insulet’s favor.

In deciding it had jurisdiction, the Federal Circuit’s reasoning went like this:

• One of Insulet’s dismissed patent claims had alleged an act of infringement in 2018.
• Section 286 of the Patent Act prohibits recovery for infringement that occurred more than six years before suit.
• So, if Insulet were to refile its patent infringement claims today, the court would deny recovery for the 2018 act of infringement.
• Consequently, the dismissal of Insulet’s patent claims, though explicitly “without prejudice,” was, in the Federal Circuit’s view, “function[ally]” a dismissal with prejudice, which establishes Federal Circuit jurisdiction over any subsequent appeal.

That sort of multistep analysis resting on conjecture about what might happen in a yet-to-be-filed case is, under both binding Supreme Court precedent and sound principles of judicial administration, the wrong way to decide jurisdiction.

As I argue in an amicus brief supporting Insulet’s pending petition for rehearing en banc, the correct question is, simply: what claims appear in the operative, well-pleaded complaint? In Insulet, the amended complaint contains no patent claims. So, the Federal Circuit does not have jurisdiction. End of story.

Instead, however, the Federal Circuit panel imagined a new patent infringement complaint that Insulet might file in the future, identified a defense EOFlow might raise, and hypothesized how a court would resolve that defense to conclude that the case currently before it did, in fact, arise under patent law.

The 1L civil procedure chestnut, Louisville & Nashville Railroad v. Mottley, fully resolves the jurisdictional dispute in the Insulet case. Under Mottley‘s “well-pleaded complaint rule,” arising under jurisdiction depends on the claims in the plaintiff’s complaint, not on defenses and not on issues that might arise later.

If there were any doubt, the Supreme Court reiterated just last year, in Royal Canin U.S.A., Inc. v. Wullschleger, that, under the well-pleaded complaint rule, an amendment eliminating claims (such as Insulet’s amendment to remove its patent claims) can “wipe the jurisdictional slate clean.”

Insulet isn’t the first time the Federal Circuit has relied on a hypothetical patent dispute to expand its jurisdiction. For instance, the circuit had long held that it had exclusive jurisdiction over state law malpractice cases against patent lawyers because the cases raised questions about what would have happened in a prior patent case had the lawyer not erred.

In Gunn v. Minton, the Supreme Court rejected that line of Federal Circuit precedent. Indeed, the Federal Circuit has a poor record before the Supreme Court on these sorts of jurisdictional or procedural questions.

In a recent study, Mark Lemley and I reviewed every patent-related case the Supreme Court has decided since the Federal Circuit began operating four decades ago. We found that the conventional wisdom about the Federal Circuit (that the Supreme Court is constantly correcting it on matters of substantive patent law) is overstated. Instead, where the Federal Circuit performs poorly is in patent-related cases raising questions of jurisdiction or procedure. In those cases (sixteen in total), the Supreme Court disagreed with the Federal Circuit fifteen times. Including in Gunn.

Those numbers reflect a recurring problem with the Federal Circuit. Congress created the court to provide uniformity and expertise in patent law, but the court has relied on those goals to expand its own authority (vis-à-vis state courts, district courts, and the regional circuits) while ignoring transsubstantive principles of federal law.

Insulet is another example. Instead of applying the well-pleaded complaint rule and sending that trade secret case to the First Circuit, the panel imagined a hypothetical future patent dispute (that might never actually arise) to keep the case in the Federal Circuit.

The en banc court should correct the panel’s error. As Ryan Vacca has shown, though, the Federal Circuit has used en banc review far less often in recent years than it did earlier in its history.

That development removes one of the few ways to correct mistakes of patent law without Supreme Court intervention. When a regional circuit errs, another circuit can disagree, prompting reconsideration or Supreme Court review. On patent law, the Federal Circuit speaks for the whole country.

Supreme Court review supplies a check, but the Court is hearing fewer patent cases than it did a decade ago.

Insulet gives the Federal Circuit a straightforward opportunity to correct itself, for the correct rule is simple: the court should (like all other federal courts) decide jurisdiction by looking at the operative complaint alone.

Paul R. Gugliuzza is the Thomas W. Gregory Professor of Law at University of Texas at Austin School of Law.