A trade-secret verdict can be perfectly proven on the merits and still collapse on the calendar. That is what happened to Insulet Corporation on May 28, 2026, when a divided Federal Circuit panel reversed its Defend Trade Secrets Act judgment against EOFlow and held that the three-year limitations period had run before Insulet ever filed suit. Judge Dyk wrote for the court, joined by Judge Reyna; Judge Prost dissented. The reversal wipes out a jury finding that EOFlow misappropriated four Insulet trade secrets, the joint-and-several liability imposed on EOFlow's entities and its CEO, and the permanent injunction the district court had entered.

The asserted secret was the technology behind Insulet's Omnipod, an adhesive, wearable insulin patch pump first marketed in 2005. EOFlow built a competing patch pump, the EOPatch, and the misappropriation theory centered on its second-generation EOPatch 2, developed after EOFlow hired several former Insulet employees who had signed confidentiality agreements. One of them, a former Insulet director of mechanical engineering, was alleged to have shared CAD files for the Omnipod Eros, soft-cannula design and manufacturing information, and occlusion-detection algorithm details with EOFlow between March and May 2018. The jury found those disclosures actionable and that the claim was not time-barred. The panel disagreed on the timing.

"we determine that EOFlow was entitled to JMOL on Insulet’s DTSA claim because the statute of limitations had expired before Insulet brought suit. We reverse."— Federal Circuit, source

The proof problem here is really a notice problem. Under 18 U.S.C. § 1836(d), a DTSA claim must be brought within three years after the misappropriation "is discovered or by the exercise of reasonable diligence should have been discovered." Insulet sued on August 3, 2023, so the critical question was whether Insulet knew, or should have known, of possible misappropriation before early August 2020. The panel did not have to choose between EOFlow's proposed inquiry-notice standard and Insulet's proposed Merck standard, because it concluded the claim was time-barred even under Insulet's more forgiving formulation.

Access plus similarity as the notice trigger

The analytical core of the opinion is the court's adoption — for limitations purposes — of an access-plus-similarity framework drawn from Uniform Trade Secrets Act case law. The court explained that a DTSA plaintiff "may rely on circumstantial evidence showing access to the trade secrets and similarities between those secrets and the accused product," citing the Third Circuit's Harbor Business Compliance and Oakwood Laboratories decisions. Because the DTSA was enacted in light of the UTSA and uses virtually identical limitations language, the court treated UTSA precedent — including the Fifth Circuit's Seatrax, the Third Circuit's Heraeus, and the Seventh Circuit's Sokol — as informing when a trade-secret owner is on notice. The same combination that lets a plaintiff infer misappropriation also tells a plaintiff when the clock starts.

Applied here, that framework was decisive. Insulet knew former employees with access to its Omnipod technology had joined EOFlow, and EOFlow's competing patch pump was visibly similar. Under Seatrax, comparable facts "put [the trade secret owner] on notice of possible misappropriation" and rendered the claim time-barred as a matter of law once suit came more than three years later. The court distinguished Sokol not as undermining the framework but as confirming it: the plaintiff there lacked knowledge of the defendant's actual use until a similar product was sold to a third party, which is exactly the kind of triggering event the framework looks for.

Reasonable measures, and the dissent

For trade-secret owners, the case is a sharp reminder that diligence runs in two directions. The same reasonable measures and monitoring that protect a secret also generate the knowledge that starts the limitations clock. Watching departing employees move to a competitor and then watching that competitor field a lookalike product is not a reason to wait for more certainty before suing — under this opinion, it can be the very notice that forecloses the claim if the owner waits too long. The prudent move is to treat access-plus-similarity signals as a litigation trigger, not merely an investigative one.

The damages dimension makes the limitations holding sting all the more. EOFlow's accused product, the EOPatch 2, was marketed in Europe and South Korea, and the litigation reached not only the corporate defendants but EOFlow's chief executive and major stockholder, who had been held jointly and severally liable. The permanent injunction the district court entered reached use or disclosure of the trade secrets "anywhere" — broad relief that the reversal now dissolves. For a medical-device company that built its case on the proposition that departing engineers carried Omnipod design, cannula, and occlusion-algorithm know-how into a competitor's next-generation pump, losing the entire judgment on timing is a hard result, and a cautionary one about how long a plaintiff can sit on access-plus-similarity signals before they harden into constructive notice.

This was not a unanimous court. Judge Prost dissented, and the majority devoted footnotes to defending the access-plus-similarity framework against her view, arguing that cases addressing the standard of proof for misappropriation are "directly relevant" to defining the elements of the claim and, by extension, what a plaintiff must have known. That methodological disagreement — whether proof-stage doctrine should drive notice-stage analysis — is the kind of fault line that can attract further attention. The panel also confirmed it had jurisdiction because the patent claims were "effectively dismissed with prejudice," and noted that its reversal mooted Insulet's parallel First Circuit appeal of the injunction. For now, the holding is straightforward and severe: a meritorious-looking trade-secret verdict over insulin-pump technology is gone, undone not by a failure of proof but by the running of time.