The Meter Is the Problem: What AI Changed in Patent Litigation, and What We’re Doing About It
November 6, 2026 | Firm NewsWhat AI actually does in my practice
I want to be specific, because the claims made for AI in law are usually too vague to check. A competitor case generates volume on a scale that used to consume associate teams for months. Document productions run to hundreds of thousands of pages. Source code and technical documents have to be found, organized, and understood before anyone can take a position. The prior art has to be swept, and swept again, before claim construction. Expert reports have to be tested against the underlying record. In my practice today, AI tools do the first pass through that volume — sorting, clustering, and flagging the documents and code a lawyer then reads closely. They do the first sweep of prior-art and legal research, which a lawyer then verifies and builds on. They support the analysis behind claim charts and infringement and invalidity positions, so the lawyers’ time goes to the positions themselves rather than to assembling the raw material underneath them. None of this makes a patent case small. It makes the preparation proportionate: the weeks of first-pass billable work that used to be the unavoidable floor under every case are no longer unavoidable.What AI does not do
It does not exercise judgment, and in a competitor case judgment is nearly everything. It does not decide which claim construction position will persuade the judge at Markman — the hearing on which these cases so often turn. It does not judge whether an expert will hold up, or what the other side’s best argument really is, or what a case is worth, or when to settle and when to push toward trial. It does not negotiate. It does not stand up in court. I have argued before the U.S. Supreme Court and served as a special master, and I can tell you the part of this work that decides cases has not changed. Every filing that leaves this firm is reviewed and signed by a lawyer here. AI has made the preparation cheaper. The judgment is the same judgment it always was — it is just no longer buried under months of first-pass work.So where should the savings go?
If the first pass through a case now costs a fraction of what it did, hourly billing has no way to show the client the saving. The meter looks exactly the same whether the first pass took a week or an afternoon, and nothing on the invoice tells the client which one it was. The alternative is to hand the efficiency to the client in the only form that matters to a board: a price, stated in advance, that does not move. None of this requires hourly firms to be acting in bad faith. An honest firm’s hourly charge for the same matter should already be going down — and I would argue it has for my clients; I am happy to offer references who will say so. But the gap that remains is informational. Nobody can predict in advance by how much they will save. And the meter introduces disputes and questions of its own — line items challenged, judgment calls re-argued on an invoice, long after the work was done and for reasons that have nothing to do with its quality. We would rather eliminate those disputes than win them. A stated price does: the saving is agreed before the work begins, and there is nothing left to dispute. We are choosing the second one.The offer
On suitable cases — and only with the approval of the firm and the lawyer who originates the matter — we now offer phased fixed fees for IP litigation, including patent litigation. The case is divided into its natural phases: the response to the complaint, claim construction, discovery, expert work, dispositive motions, trial preparation, trial. Each phase carries one stated price, agreed before the phase begins. A client can budget phase by phase, and a board can approve a number instead of authorizing a process. Complexity is real, and pretending otherwise would just rebuild the hourly meter under a new name. So each case is scored against a rubric we will publish — a weighting from 1 to 5 that accounts for the number of rights in suit, the parties, the accused products, the technical complexity, parallel proceedings such as a PTAB case running beside the district court action, the pace of the schedule, and the exposure. The score is re-examined every six months, and it can move down as well as up as a case narrows. A fixed fee means the firm carries the variance inside a phase; that certainty is worth something, and it is priced in — openly, in the number itself. The selectivity is deliberate, and for cases of this size it is a feature. A firm that will fix a price on any case, sight unseen, is not pricing the case — it is pricing its own optimism. We offer this where the phases can be defined honestly, and we say so where they cannot. We are not inventing the idea that litigation can be bought this way. Johnson & Johnson’s legal department has for years priced litigation in component units, calibrated by the complexity of the matter. England’s courts now assign civil cases to complexity bands, each with stage-by-stage fixed costs. Sophisticated corporate buyers and at least one major court system have already concluded that a case’s price can be stated by phase and complexity. We think a company defending — or bringing — a competitor case should be able to buy its counsel the same way. I am a registered patent attorney, and this firm tries patent cases; Chambers USA ranks me Band 3 for intellectual property patent litigation in Washington. The point of this article is not the résumé. It is that the economics of the work have changed, and the pricing should say so. The full fee schedule follows this article and is available on request. If you have a case you would like us to score, we will tell you the weighting and the phase prices before you owe us anything beyond the conversation.Losing a § 101 Fight Is Not “Exceptional”: Federal Circuit Reverses $9.15 Million Fee Award in ChromaDex v. Elysium Health
November 3, 2026 | Firm NewsOn October 30, 2026, the Federal Circuit reversed a $9,151,133 attorneys' fee award under 35 U.S.C. § 285, holding that the district court abused its discretion when it treated the Supreme Court's Myriad decision as having foreclosed the patentee's eligibility position from the start. ChromaDex, Inc. v. Elysium Health, Inc., No. 2025-1227 (Fed. Cir. Oct. 30, 2026) (nonprecedential).
Background
ChromaDex and the Trustees of Dartmouth College own U.S. Patent Nos. 8,197,807 and 8,383,086, directed to dietary supplement compositions containing isolated nicotinamide riboside (NR), a form of vitamin B3. NR occurs naturally in cow's milk, but only in trace amounts, bound to whey protein and not bioavailable. In 2023, the Federal Circuit affirmed summary judgment that the asserted claims are patent-ineligible under § 101 as directed to a product of nature. ChromaDex, Inc. v. Elysium Health, Inc., 59 F.4th 1280 (Fed. Cir. 2023).
Elysium then moved for fees under § 285. The district court (Chief Judge Connolly, D. Del.) found the case exceptional: in its view, Myriad "addressed the specific issue presented," ChromaDex's summary judgment briefing "mentioned Myriad just once" and made no attempt to distinguish it, and ChromaDex's position "stood out" from the dozens of other § 101 challenges the court had seen. It awarded Elysium $9,151,133, including roughly $3.9 million in fees incurred before the Markman hearing.
The Reversal
The Federal Circuit (Prost, Chen, and Stoll, Circuit Judges; opinion by Judge Chen) held that it was "legal error to treat Myriad as the beginning and the end of the patent-eligibility analysis." Myriad answered one question — that merely isolating a naturally occurring molecule, without more, does not confer eligibility. But ChromaDex's principal argument went further: under Diamond v. Chakrabarty and Natural Alternatives International, Inc. v. Creative Compounds, LLC, 918 F.3d 1338 (Fed. Cir. 2019), it contended that its compositions had markedly different characteristics and significant utility over milk, because isolated NR can increase NAD+ biosynthesis in a way naturally occurring NR cannot — an argument backed by expert testimony. Evaluating that position required several considerations beyond Myriad, and the Federal Circuit's own 2023 decision had worked through them at length before rejecting the argument.
Under Octane Fitness, a case is exceptional when it "stands out from others with respect to the substantive strength of a party's litigating position," and a position need not be correct to be reasonable — it must be "exceptionally meritless" to support a fee award. Because ChromaDex's position rested on a genuine, precedent-based argument, it was not exceptionally meritless, and the district court's contrary legal premise was an abuse of discretion. The court also refused Elysium's request for a remand: neither the district court's analysis nor the record could support an exceptionality finding, and fee disputes should not become "a second major litigation."
What It Means
- Losing is not standing out. A patentee who loses a hard-fought § 101 motion does not face fee exposure merely because a prominent precedent pointed the other way. The court must assess the position actually advanced.
- Anchoring matters. ChromaDex's argument survived fee review because it was built on on-point Federal Circuit precedent (Natural Alternatives) and expert support — not on ignoring Myriad.
- Fee motions get one fair shot. By reversing outright rather than remanding, the court signaled that exceptionality cannot be rebuilt on a second record — a meaningful check on fee-driven satellite litigation.
The full opinion is available from the Federal Circuit: https://www.cafc.uscourts.gov/opinions-orders/25-1227.OPINION.10-30-2026_2779593.pdf
Court Dismisses DentalEZ Claims Against Professional Sales Associates
October 27, 2026 | Firm NewsThe U.S. District Court for the Western District of Washington has dismissed Professional Sales Associates, Inc. ("PSA") from DentalEZ, Inc. v. Firstar Dental Company, No. C25-1368JLR. In an order dated October 20, 2026, Judge James L. Robart granted PSA's motion to dismiss and dismissed DentalEZ's two Lanham Act claims against PSA — false advertising and false designation of origin — with prejudice and without leave to amend. The order also dismissed DentalEZ's Washington Consumer Protection Act claim and its request for punitive damages, which were asserted against Firstar.
DentalEZ, a manufacturer and wholesaler of dental equipment, alleged that PSA and Firstar used a presentation to pitch Firstar's competing products to national dental dealers. The court held the presentation was not commercial advertising to the relevant purchasing public and that DentalEZ had not identified an actionable false statement or source identifier under the Lanham Act.
Lowe Graham Jones represented PSA. Only DentalEZ's tortious-interference claim against Firstar survives.
Read the order (PDF).