The Docket — August 2, 2026

I · Trade-Secret Law

The Instruction That Pointed the Wrong Way

Ninth Circuit · trade secrets

Two appellate decisions from the same court this summer share a hidden hinge, the humble question of who must prove a fact and how far a trial judge’s choices survive review. One lives inside the engineered world of radio-frequency power electronics; the other lives far outside it, among endangered fish and irrigation water on the Klamath River.

The first begins with a rich verdict that came apart over a single jury instruction. A company had pleaded two overlapping trade-secret claims, dropped one in the middle of trial, and left the jury reading a rule written for the claim that no longer existed.

$40Mverdict vacated

$17Mfee award vacated

2–1panel split

The setup

Comet makes radio-frequency power delivery systems that feed the plasma used to etch and coat semiconductor wafers, and XP Power is a competitor in power hardware.[1] A federal jury in California found that XP had misappropriated Comet’s trade secrets, meaning it had taken and used confidential information that draws value from being secret, and it awarded $40 million, split evenly between compensatory and punitive damages.[3] The trial court added a permanent injunction and roughly $17 million in attorney fees, and XP appealed.[3]

The question

Both of Comet’s claims lived under trade-secret statutes, one federal and one state, and the two allocate a key burden in opposite directions.[1] The federal Defend Trade Secrets Act, the DTSA, makes a plaintiff prove that its information was not “readily ascertainable through proper means” by others, so secrecy is part of the plaintiff’s own case.[1] California’s trade-secret act, the CUTSA, instead treats ready ascertainability as an affirmative defense, a point the defendant must raise and prove.[2] The narrow question on appeal was which allocation the jury should have been told to apply once only the federal claim remained.[4]

The reasoning, walked

The two claims were tried together, so the court gave one instruction placing the burden of ready ascertainability on XP, which is correct under California law.[1] Partway through trial Comet voluntarily dismissed its state claim and pressed only the federal one, and XP objected that the burden should now flip to Comet.[2] The judge kept the defendant-burden instruction, and the panel held that this “erroneously flipped the burden of proof on an essential element of Comet’s claim,” because the DTSA puts that burden on the plaintiff.[1]

Comet’s first answer was that XP had proposed the very instruction it now attacked, which would make this invited error, a doctrine barring a party from appealing a mistake it caused.[1] The panel rejected that reading of the record. XP had offered the instruction while both claims were live, then objected in time once Comet dropped the state claim, and the combination of that dismissal and the court’s ruling left XP a real dilemma rather than a free choice.[1] Arguing in the alternative, the court held, did not waive the objection.[2]

The harder move was harmlessness, since even a wrong instruction can support a verdict if it is more probable than not that a properly charged jury would have reached the same result.[1] The majority found the error was not harmless for three linked reasons. Reading the charge as a whole did not cure it, because independent economic value and ready ascertainability measure different things, one the information’s usefulness and the other its secrecy.[1] The evidence on whether rivals could have reverse-engineered the technology was sharply in conflict, with XP’s expert pointing to obtainable physical dimensions and Comet’s pointing to schematics and test data that were not public.[3] And the jury had awarded different sums for different secrets, so the court could not tell how a correctly instructed jury would have sorted them.[1]

On that basis the panel reversed and remanded for a new trial, undoing the $40 million award, the injunction, and the fees.[1] The panel’s author, Judge Hamilton, added a separate concurrence on a distinct remedies worry, whether granting both unjust-enrichment damages measured by the defendant’s avoided development costs and an injunction is a forbidden double recovery.[1] He concluded it is not, reasoning that avoided-cost damages redress a past benefit while an injunction prevents future harm, and he set apart a Second Circuit decision that had rejected an avoided-cost award where the defendant built no competing product.[2] Judge Bumatay dissented, agreeing the instruction was wrong but calling the error harmless, since in his read the record “easily” showed the jury would have found the information not readily ascertainable, and the large compensatory award served as a reliable proxy for that finding.[1]

As a matter of argument structure, the majority reasons more cleanly, and this is an assessment of reasoning, not legal advice.[1] Its harmlessness holding rests on a concrete feature of the verdict, the different dollar figures assigned to different secrets, which the record cannot confidently reconcile.[1] The dissent reasons from the verdict’s bottom line backward, treating a large damages number as proof of a finding the jury was never correctly asked to make.[1]

A verdict can be undone not by the facts but by who the jury was told had to prove them.

What turns on it

For anyone litigating trade secrets in the Ninth Circuit, the lesson is mechanical and expensive, that pleading parallel federal and state claims and then dropping one can silently misalign the jury charge, and the surviving statute’s burden must control.[4] The ready-ascertainability element lands close to home for engineers, because it asks whether a competitor could have reached the same design by fair means such as reverse engineering, which is often the whole fight.[3] Concretely, a $40 million verdict and a $17 million fee award are now unwound, and the dispute returns to the district court for a second trial under a corrected instruction.[1] The concurrence’s conclusion that avoided-cost damages and an injunction can coexist also gives future plaintiffs a clearer route to stacking those remedies.[2]

Sources

II · Water Rights & Endangered Species

Discretion on the Klamath

Ninth Circuit · water rights

Water in the Klamath Basin runs chronically short, and each drought forces the same hard triage between farms that need irrigation and fish that need flow.[1] This case turns on a narrow, load-bearing question asked first, whether a federal wildlife law reaches the agency that operates the water at all.[2]

240kacres served

2–1panel on the ESA question

Affirmeddisposition

The setup

The Klamath Project, authorized in 1905, is a federal irrigation system straddling the California–Oregon line, watering roughly 240,000 acres and several wildlife refuges from Upper Klamath Lake.[1] Two sucker species and coho salmon in the system are listed as threatened or endangered, and after severe drought the Bureau of Reclamation adopted operating rules that held water back for minimum lake levels and downstream flows.[3] Irrigation users challenged those rules, and when Oregon’s water agency ordered the species-protection releases stopped in April 2021, the United States and the Yurok and Hoopa Valley Tribes sued to keep the water moving for the fish.[1]

The question

Section 7(a)(2) of the Endangered Species Act, the ESA, requires every federal agency to ensure its actions do not jeopardize a listed species.[2] The Supreme Court has read that command to reach only discretionary agency action, not steps an agency is legally compelled to take, in a 2007 decision, National Association of Home Builders v. Defenders of Wildlife.[2] So the real question became whether the Bureau has discretion over Klamath deliveries, given the water-supply contracts it signed with irrigators, because without discretion the ESA duty would not attach.[1]

The reasoning, walked

A wrinkle framed everything, because a 1999 Ninth Circuit decision, Klamath Water Users Protective Ass’n v. Patterson, had already held that Section 7 governs the Bureau’s operation of this project.[2] A three-judge panel must follow its own circuit’s prior rulings unless they are “clearly irreconcilable” with a later Supreme Court or circuit decision, the standard from Miller v. Gammie.[1] The whole dispute reduced to whether the intervening discretion cases had knocked Patterson out from under that rule.[2]

The majority held they had not.[1] It read Home Builders narrowly, as reaching only the case where a second federal statute forces an agency into a specific action that collides with ESA compliance, which the Reclamation Act’s broad direction to manage the project is not.[1] It then turned to the contracts, noting that most contain shortage clauses letting the Bureau curtail deliveries when water is “unavailable due to drought or other reasons.”[1] Those clauses, the court reasoned, supply a legal basis to withhold water for species protection, so the Bureau keeps the discretion that triggers Section 7.[2]

The panel disposed of two further arguments quickly. It found no judicial taking of the irrigators’ water rights, a claim that a court order can itself amount to a seizure of property, because deciding that the ESA applies is not the same as adjudicating who owns the water.[1] And it held the federal court had jurisdiction, since doctrines that would send the matter elsewhere, including Colorado River abstention, which lets a federal court defer to parallel state litigation, address the state adjudication of water rights, not the federal statutory question here.[1] Reviewing the summary judgment de novo, meaning with fresh eyes and no deference to the court below, the panel affirmed.[2]

Judge R. Nelson dissented on the ESA point while joining the rest.[1] In his view Patterson is clearly irreconcilable with Home Builders and with a more recent circuit case, NRDC v. Haaland, that asks contract by contract whether an agency actually kept discretion.[2] The right question, he wrote, is not whether the ESA applies to the project in the abstract but whether these specific contracts leave the Bureau any room to choose.[1] The shortage clauses do not create discretion, he argued, they merely excuse the Bureau from liability when some other legal duty compels it to cut deliveries, and “the duty to comply with mandatory legal obligations is not a source of discretion.”[1] He would have largely reversed and sent the case back for the contract-by-contract analysis the majority skipped.[2]

The whole water fight narrows to a single word: discretion.

Judged purely as argument structure, the dissent isolates the operative question more precisely, since after the discretion cases it is the contract terms, not a 1999 precedent’s broad phrasing, that decide the outcome.[2] The majority’s stronger move is textual, reading the “or other reasons” clause as an affirmative grant of authority rather than a mere liability shield, which if correct answers the dissent on its own ground.[1] This is an assessment of reasoning, not legal advice.[1]

What turns on it

The discretion gateway is the whole game for federally run water across the arid West, because if contract duties can strip an agency of discretion, they can also switch off its ESA obligations.[2] A forceful dissent that reads two higher decisions the other way is the kind of split that draws petitions for rehearing en banc, before all the circuit’s active judges, or review by the Supreme Court.[1] For now the outcome is concrete, as Klamath irrigators stay subject to species-driven curtailments and the basin’s salmon and suckers keep their court-protected flows.[3] Beneath the water fight sits a quieter structural question that recurs across the federal courts, when a panel may treat its own circuit’s precedent as overtaken by later law.[2]

Sources