2025-1807

Appeal from the United States District Court for the District of Massachusetts in No. 1:23-cv-11780-FDS, Judge F. Dennis Saylor, IV.

Decided: May 28, 2026

WILLIAM M. JAY, Goodwin Procter LLP, Washington, DC, argued for plaintiff-appellee. Also represented by MATTHEW GINTHER; ROBERT CARROLL, WILLIAM EVANS, ROBERT FREDERICKSON, III, ALEXANDRA LU, Boston, MA; ALEXANDRA D. VALENTI, New York, NY.

ELIZABETH PRELOGAR, Cooley LLP, Washington, DC, argued for defendants-appellants. Also represented by ELIZABETH M. FLANAGAN, Minneapolis, MN; LOWELL D. MEAD, Palo Alto, CA.

Page 2Before DYK, PROST, and REYNA, Circuit Judges .

Opinion for the Court filed by Circuit Judge DYK.

Dissenting opinion filed by Circuit Judge PROST.

DYK, Circuit Judge .

Insulet Corporation (“Insulet”) sued EOFlow, Co. Ltd., EOFlow, Inc., and its chief executive officer and major stockholder Jesse Kim (collectively, “EOFlow” unless otherwise noted) for trade secret misappropriation under the Defend Trade Secrets Act (“DTSA”) and for patent infringement. The district court bifurcated Insulet’s trade secret misappropriation claim from the patent claims.

EOFlow argued that the statute of limitations had expired for Insulet's trade secret misappropriation claim. The district court denied both parties’ summary judgment motions on the limitations issue. The case proceeded to a jury trial on the DTSA claim. The jury found that EOFlow misappropriated four Insulet trade secrets and that Insulet’s claim was not time-barred for any of the misappropriated trade secrets and awarded damages. The district court denied EOFlow’s posttrial judgment as a matter of law (“JMOL”) motion; entered judgment awarding Insulet damages; held both EOFlow entities and Mr. Kim jointly and severally liable for the damages; and issued a permanent injunction. After trial, the patent claims were dismissed without prejudice. EOFlow appeals.

We conclude that we have jurisdiction because the patent claims were effectively dismissed with prejudice. On the merits, 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.

Page 3The following summary is based on the undisputed testimony at trial, unless otherwise noted. Insulet is a medical-device manufacturer that makes an adhesive, wearable insulin patch pump called the Omnipod®, which was first marketed in 2005. After its initial launch, Insulet continued to work on next-generation versions of the Omnipod, including the Omnipod Eros. Insulet marketed the Omnipod Eros beginning in 2011.

EOFlow also developed an adhesive, wearable insulin patch pump called the EOPatch®. EOFlow never marketed the first-generation EOPatch, and this patch pump was not alleged to have been developed through trade secret misappropriation. However, Insulet alleged trade secret misappropriation, beginning in 2018, in connection with EOFlow’s efforts to develop a second-generation patch pump, known as the EOPatch 2. This pump was later marketed in Europe and South Korea.

EOFlow's misappropriation was alleged to have occurred when EOFlow hired several former Insulet employees to help develop and commercialize the EOPatch 2. These former employees had signed confidentiality and non-disclosure agreements at Insulet. One of the former Insulet employees, Steve DiIanni, the Director of Mechanical Engineering at Insulet, was allegedly “entrusted with details of every aspect of the research and development efforts to create the world's first insulin patch pump,' J.A. 321 ¶ 40, and possessed 'detailed technical information about Insulet’s Omnipod product, including but not limited to product specifications, material compositions, coating materials, . . . operational software requirements and details, dimensional tolerances, . . . and other product and manufacturing details.” J.A. 322 ¶ 41.

Page 4Mr. DiIanni provided assistance to EOFlow in developing a second-generation patch pump. Between March and May 2018, Mr. DiIanni shared computer-aided design (CAD) files for the Omnipod Eros, information about the design and manufacture of the Omnipod's soft cannula, and information regarding the Omnipod’s occlusion-detection algorithm with EOFlow. Insulet alleges that these disclosures, among others, constituted trade secret misappropriation.¹

On August 3, 2023, Insulet brought suit against EOFlow in the District of Massachusetts asserting trade secret misappropriation under the DTSA and patent infringement. Insulet moved for a preliminary injunction against EOFlow on the DTSA claim. The district court granted the preliminary injunction and bifurcated the DTSA claim and patent claims. We reversed the district court's grant of preliminary injunction. Insulet Corp. v. EOFlow, Co. (“ Insulet I ”), 104 F.4th 873 (Fed. Cir. June 17, 2024). We concluded that the district court “fail[ed] to address the statute of limitations,” failed to conduct “a tailored analysis as to what specific information actually constituted a trade secret,' and 'abused its discretion in reaching its findings as to irreparable harm and the public interest.” Insulet I , 104 F.4th at 883.

On remand, EOFlow and Insulet both moved for summary judgment on the statute of limitations issue. Under the DTSA, the statute of limitations expires “3 years after Page 5the date on which the misappropriation with respect to which the action would relate is discovered or by the exercise of reasonable diligence should have been discovered. For purposes of this subsection, a continuing misappropriation constitutes a single claim of misappropriation.' 18 U.S.C. § 1836(d). EOFlow contended that the DTSA should be interpreted to incorporate an inquiry-notice standard and that Insulet's claim was time-barred as a matter of law under that standard. Insulet, in turn, argued that the Supreme Court’s decision in Merck & Co. v. Reynolds , 559 U.S. 633 (2010), had rejected an inquiry-notice standard for a similar statute; that the Merck discovery standard applied; and that its claim was not time-barred as a matter of law under that standard. The district court agreed with Insulet that the Merck standard applied, but concluded there were “material factual disputes concerning the time at which a reasonably diligent company should have discovered the alleged misappropriation” under the Merck standard and denied both summary judgment motions. J.A. 23. At trial, both parties presented evidence on the statute of limitations issue.

The jury found EOFlow misappropriated four of Insulet's trade secrets: (1) the CAD files for the Omnipod; (2) the design and manufacturing process for the Omnipod soft cannula; (3) the design history file for the Omnipod Eros; and (4) the occlusion-detection algorithm for the Omnipod. The jury also found that, for each asserted trade secret, Insulet’s claim was not barred by the statute of limitations; and that EOFlow, Co. Ltd. and Mr. Kim willfully and maliciously misappropriated three of the trade secrets (the CAD files, occlusion-detection algorithm, and design history file). The jury awarded Insulet $170,000,015 in compensatory damages and $282,000,005 in exemplary damages. After the verdict, on January 3, 2025, the district court dismissed Insulet's patent claims 'without prejudice.” J.A. 21412.

Page 6EOFlow moved for JMOL on the DTSA claim on multiple grounds, including that Insulet’s DTSA claim was timebarred under both the inquiry notice and Merck standards. The district court denied EOFlow's motion.

Insulet requested a permanent injunction. EOFlow argued that the permanent injunction was entirely duplicative of the jury’s damages award because Insulet’s damages model was based on enjoined future sales of EOPatch 2. The district court agreed that the injunction provided relief that overlapped significantly with the jury's damages award and reduced the damages award to $25.8 million in compensatory damages and $33.6 million in exemplary damages to prevent double recovery. On April 24, 2025, the district court entered judgment and a permanent injunction enjoining EOFlow’s disclosure and use of the trade secrets, enjoining the commercialization of 'any product that was designed, developed, or manufactured, in whole or in part, using or relying on the Trade Secrets, anywhere in the world,” J.A. 111, and ordered EOFlow to assign certain patent applications to Insulet.

EOFlow timely appealed to our court. Insulet then appealed the judgment and injunction to the First Circuit and moved to transfer this appeal to the First Circuit.² On July 7, 2025, we denied Insulet's motion to transfer and requested that the parties address our jurisdiction in their merits briefs.

Page 7We first consider whether we have jurisdiction. 28 U.S.C. § 1295(a)(1) grants this court exclusive jurisdiction “in any civil action arising under . . . any act of Congress relating to patents.” Our jurisdiction thus “depends on whether the plaintiff’s complaint as amended raises patent law issues.” Chamberlain Grp., Inc. v. Skylink Techs. , Inc. , 381 F.3d 1178, 1189 (Fed. Cir. 2004). A dismissal with prejudice of patent claims does not deprive our court of jurisdiction, as dismissals with prejudice “are adjudications on the merits, and not constructive amendments to the complaint.' Id. at 1189-90. Whether a dismissal is with prejudice 'is functional, rather than semantic.' Id. at 1190.

Normally, an amendment to a complaint that dismisses patent claims without prejudice will divest this court of its jurisdiction over the appeal. Krauser v. BioHorizons, Inc. , 753 F.3d 1263, 1269 (Fed. Cir. 2014); see also Royal Canin U. S. A., Inc. v. Wullschleger , 604 U.S. 22, 34 (2025) (describing how “an amendment can wipe the jurisdictional slate clean'). However, a dismissal nominally without prejudice may effectively be a dismissal with prejudice if the statute of limitations has run. See Hilton Int’l Co. v. Union de Trabajadores de la Industria Gastronomica , 833 F.2d 10, 11 (1st Cir. 1987) (“The weight of authority . . . is that dismissal . . . even though labelled ‘without prejudice,’ is, in fact, with prejudice if the statute of limitations has run.”); see also Chamberlain , 381 F.3d at 1190.

EOFlow argues that Insulet’s voluntary dismissal of its patent infringement claims (and the amendment to its complaint to delete the patent claims) functioned as a dismissal with prejudice because refiling of the claims would be barred by the six-year statute of limitations, 35 U.S.C. § 286. In particular, EOFlow points out that the complaint alleged infringement occurring in June 2018, whenPage 8 'EOFlow imported at least one device for display at the 2018 ADA meeting,” J.A. 1132 ¶ 134, and that this claim was barred by the six-year limitations period at the time of the dismissal, see 35 U.S.C. § 286. We agree with EOFlow. Insulet is barred from “refiling of the same claim” of patent infringement because the statute of limitations expired in June 2024 with respect to acts of infringement in June 2018, and the dismissals occurred on January 3, 2025, more than six years later. See Semtek Int’l Inc. v. Lockheed Martin Corp. , 531 U.S. 497, 506 (2001) (equating an “adjudication upon the merits” to a dismissal where refiling is barred).

Insulet nonetheless argues that the dismissal instead functions as a dismissal without prejudice. First, it argues that section 286 does not bar Insulet from bringing suit related to acts of infringement committed more than six years prior to filing suit; it merely prevents recovery for such acts. For present purposes, this is a distinction without a difference. A dismissal is with prejudice if the “dismissal[ ] altered the legal positions of [the parties] vis-à-vis the dismissed claim.” Chamberlain , 381 F.3d at 1190. Under either view of the statute, “the legal positions” of the parties were altered by the dismissal.

Second, Insulet argues that in addition to the alleged act of importation from 2018, it 'alleged multiple additional acts of importation of the pirated device . . . occurring in 2022 at the earliest.” Appellee’s Br. 29. But, as both concurring opinions for the en banc court in Arellano v. McDonough recognized, ongoing patent infringement ''comprises a 'series of discrete infringing acts,' each of which is a distinct harm giving rise to an independent claim for relief that starts a new limitations period.''¹ F.4th 1059, 1091 (Fed. Cir. 2021) (en banc) (Dyk, J., concurring) (quoting Arellano , 1 F.4th at 1070 (Chen, J., concurring) (quoting Petrella v. Metro-Goldwyn-Mayer, Inc. , 571 U.S. 663, 671–72 (2014)) (emphasis in original)). Regardless of whether Insulet may still file patent Page 9infringement claims for some alleged acts of infringement, the statute of limitations had expired with respect to EOFlow’s alleged June 2018 infringement.

Finally, Insulet argues that the cases establishing that a dismissal is with prejudice when the statute of limitations has run only apply to involuntary dismissals, not voluntary dismissals. But it is well-established that this is also “clearly the rule if the dismissal is voluntary” and “any other result . . . would be most unfair.' Hilton , 833 F.2d at 11. Because Insulet could not refile suit predicated on this alleged act of infringement, the dismissal of its patent claims, at least in part, functioned as a dismissal with prejudice. We therefore have jurisdiction.

We next address the statute of limitations issue. We review the district court’s denial of EOFlow’s JMOL motion on statute-of-limitations grounds de novo. Proveris Sci. Corp. v. Innovasystems, Inc. , 536 F.3d 1256, 1266-67 (Fed. Cir. 2008) (citing Hochen v. Bobst Grp., Inc. , 290 F.3d 446, 453 (1st Cir. 2002)).

The statute of limitations for federal trade secret misappropriation claims runs “3 years after the date on which the misappropriation with respect to which the action would relate is discovered or by the exercise of reasonable diligence should have been discovered. For purposes of this subsection, a continuing misappropriation constitutes a single claim of misappropriation.” 18 U.S.C. § 1836(d).

EOFlow argues that the so-called inquiry-notice standard applies and that inquiry-notice standard begins the limitations period “‘when the first event occurs that would prompt a reasonable person to inquire into a possible injury at the hands of the defendant.'' Appellants' Br. 28 (quoting Epstein v. C.R. Bard, Inc. , 460 F.3d 183, 187-88 (1st Cir. 2006) (applying Massachusetts law)). Insulet argues that (1) inquiry notice is not the proper standard Page 10because, in Merck , the Supreme Court rejected inquiry notice in construing the discovery standard under a similar statute; and (2) the same discovery standard discussed in Merck should apply to the DTSA. Under the Merck standard, the statute of limitations begins whenever a plaintiff did discover or should have discovered the facts underlying its cause of action, whereas under the inquiry-notice standard, the statute of limitations begins to run when the plaintiff or a reasonably diligent plaintiff would have begun an investigation that would allow it to discover these facts. Merck , 559 U.S. at 651–53.

As the district court concluded, the factual circumstances present here demonstrated that Insulet 'plainly had at least a cause for concern that defendants had engaged in prohibited conduct by early 2019,” before the critical date of August 3, 2020 (three years before Insulet filed its complaint) (“critical date”), J.A. 21, and that if the inquiry-notice standard applied, “then defendants would almost certainly be entitled to summary judgment on the trade secret claim.” J.A. 46. Insulet disputes the district court’s apparent conclusion that it was on inquiry notice before the critical date. The parties also disagree as to whether, applying the Merck standard, substantial evidence supported the jury’s verdict.

We need not decide whether EOFlow's proposed inquiry-notice standard or Insulet’s proposed Merck standard applies. We conclude that even under the more demanding Merck standard, the statute of limitations expired before Insulet filed its complaint because the undisputed evidence shows that Insulet knew or should have known before the critical date the facts it needed to sufficiently plead a trade secret misappropriation claim against EOFlow. In doing so, we do not, as the dissent suggests, “conflate [our] application of the discovery rule with the inquiry-notice standard,” Dissent 1, but apply the more rigorous Merck standard.

Page 11Before turning to the facts of the present case, we must consider the applicable standard for pleading a claim for trade secret misappropriation under the DTSA. EOFlow argues that a claim for trade secret misappropriation under the DTSA may be properly pled and proven through factual allegations demonstrating the combination of the defendants’ access to the trade secret and the similarity of the trade secret to the defendant's design.³ We agree.

Cases addressing the standards for pleading and proof of misappropriation under the DTSA establish that, to prove a trade secret misappropriation claim, “[a] plaintiff may rely on circumstantial evidence showing access to the trade secrets and similarities between those secrets and the accused product.'⁴ Harbor Bus. Compliance Corp. v. Firstbase.io, Inc. , 152 F.4th 516, 530 (3d Cir. 2025) (applying the DTSA) (citing Oakwood Laboratories, Inc. v. Thanoo , 999 F.3d 892, 912 n.19 (3d Cir. 2021) (applying the DTSA)). Interpretations of the Uniform Trade Secret Act (“UTSA”) also inform our interpretation of the DTSA. When it enacted the DTSA, “Congress was aware of the role and limitations of the UTSA as model legislation for the states, and it recognized the DTSA and the UTSA as Page 12similar.” Attia v. Google LLC , 983 F.3d 420, 425 (9th Cir. 2020). One House report noted that the limitations period of the DTSA is 'identical to the limitations period of the UTSA.” H.R. Rep. No. 114-529, at 13 (2016). Given that the DTSA was enacted in light of the UTSA and, in relevant respect, uses virtually identical language, 5 it is appropriate to consider cases interpreting the UTSA (or the UTSA’s statute of limitations provision) in interpreting the DTSA. See George v. McDonough , 596 U.S. 740, 746 (2022) (“Where Congress employs a term of art obviously transplanted from another legal source, it brings the old soil with it.” (internal quotation marks omitted)).

Circuits analyzing the UTSA statute of limitations have applied the same access-plus-similarity framework. See Seatrax, Inc. v. Sonbeck Int’l, Inc. , 200 F.3d 358, 365 (5th Cir. 2000) (applying Texas law); Heraeus Med. GmbH v. Esschem, Inc. , 927 F.3d 727, 735–36 (3d Cir. 2019) (applying Pennsylvania law); Sokol Crystal Prods., Inc. v. DSC Commc’ns Corp. , 15 F.3d 1427, 1430 (7th Cir. 1994) (applying Wisconsin law).⁶ For example, in Seatrax , a Page 13crane manufacturer knew that individuals who were 'given access to technical information, manuals, service bulletins, computer software and . . . manufacturer drawings” formed another company to sell crane parts that were similar to the trade-secret owner's designs. Seatrax , 200 F.3d at 362-63. The Fifth Circuit held that these events “put [the trade secret owner] on notice of possible misappropriation,” id. at 365, and therefore the trade secret owner’s claim was time-barred as a matter of law because it brought suit more than 'three years after the misappropriation[ ] is discovered or by the exercise of due diligence should have been discovered,” id. at 364 (emphasis removed) (quoting Tex. Civ. Prac. & Rem. Code Ann. § 16.010(a)). Further, our court has held that if a plaintiff also id. at 1430 (“Sokol did not know that DSC was misusing (and therefore misappropriating) its trade secret until DSC sold an ECS–3 system (containing a VCXO that resembled Sokol's) to a third party.'); id. at 1432 (holding that “once the jury concluded that (1) DSC had access to Sokol’s trade secrets, and (2) DSC’s product was similar to Sokol’s, it was entirely reasonable for it to infer that DSC used Sokol’s trade secret”).

The question was whether the plaintiff had knowledge that a similar device was being sold (and thus misappropriated) before the critical date. While the defendant sent sample components (VCXO boards) to the plaintiff before the critical date, the plaintiff did not then know that the defendant would sell a similar device containing the VCXO boards, i.e. , it did not “know the use to which these boards were going to be put.” Id. at 1430. The plaintiff did not have that knowledge of misappropriation until it learned (after the critical date) that the defendant “sold [a similar product] (containing a VCXO that resembled Sokol’s) to a third party.” Id. Sokol directly supports the access-plussimilarity framework for pleading and proving trade secret misappropriation claims.

Page 14has knowledge that (1) the alleged misappropriator has access to the alleged trade secrets through a former employee of the trade secret owner and (2) there is similarity between the trade secrets and the alleged misappropriator’s competing product, then the plaintiff possesses sufficient facts to state a claim for trade secret misappropriation. See Leggett & Platt, Inc. v. Hickory Springs Mfg. Co. , 285 F.3d 1353, 1361 (Fed. Cir. 2002) (applying Illinois law).

Here, in its brief, Insulet argued that the access-plussimilarity test does not apply and that the limitations period began to run only when it acquired detailed and specific knowledge of the misappropriation. For example, it argued that it could not ascertain the 'nominal dimensions” of the designs and prototypes of the EOPatch from EOFlow’s displays at trade shows, presentations to Insulet, and public disclosures. Dkt. No. 60 at 1 (citing J.A. 21035). In this respect, Insulet is mistaken. Because the standard is substantial similarity, detailed knowledge of all of the relevant facts underlying the misappropriation is not required. Knowledge of such details as the nominal dimensions of EOPatch 2 is unnecessary to plead a trade secret misappropriation claim. Significantly, despite the arguments in its brief, Insulet agreed at oral argument that knowledge of facts demonstrating access and similarity are sufficient to plead a trade secret misappropriation claim. Oral Arg. at 28:23-29:18. The relevant question is ultimately whether Insulet had sufficient knowledge to state a claim for trade secret misappropriation before the critical date, not whether it had sufficient evidence to prove misappropriation of its trade secrets.

The next question is whether, in order to trigger the running of the statute of limitations, the plaintiff must have discovered (or should have discovered) the misappropriation of each alleged trade secret or if it is sufficient that Page 15it discovered (or should have discovered) misappropriation of a single trade secret.

Over EOFlow’s objection, the jury was instructed that it must “separately consider the statute of limitations defense as to each of [the] trade secrets,” J.A. 21392; see also J.A.19977-78, J.A. 130, 6:1-12, J.A. 184, 60:16-22. We have recognized, however, that whether a party is entitled to JMOL “does not depend on how the jury was instructed.” Harris Corp. v. Ericsson Inc. , 417 F.3d 1241, 1252 (Fed. Cir. 2005). The inquiry is whether “under the controlling law,” the defendant is entitled to JMOL. Fed. R. Civ. Proc. 50(a)(1)(B).

To resolve this dispute, we must interpret the DTSA’s provision that “[f]or purposes of this subsection [the statute of limitations], a continuing misappropriation constitutes a single claim of misappropriation.” 18 U.S.C. § 1836(d). This provision is essentially copied from a provision in the UTSA statute of limitations—“[f]or the purposes of this section, a continuing misappropriation constitutes a single claim.' Uniform Trade Secrets Act § 6 (amended 1985) (Nat'l Conf. of Comm'rs on Unif. State L. 1985) ('1985 UTSA”). While there is no First Circuit authority on this issue, given the relevant provisions of the DTSA and the UTSA are substantively identical, we consider the history of the UTSA and cases decided under the UTSA.

Before the UTSA, there was a “conflict of authority as to whether trade secret misappropriation is a continuing wrong.” 1985 UTSA, § 6 cmt. Some courts, most notably the D.C. Circuit in Underwater Storage, Inc. v. U.S. Rubber Co. , 371 F.2d 950 (D.C. Cir. 1966), concluded that each act of trade secret misappropriation is a distinct wrong with a separate limitations period because trade secrets are “valuable property” and, as such, “a cause of action for each invasion of the plaintiff's interest arose at the time of that invasion.” 371 F.2d at 954–55. “Underlying this theory is the concept that a trade secret is in the nature of property, Page 16which is damaged or destroyed by the adverse use.” Monolith Portland Midwest Co. v. Kaiser Aluminum & Chem. Corp. , 407 F.2d 288, 293 (9th Cir. 1969) (discussing Underwater Storage ). Other courts, like the Ninth Circuit in Monolith , did not “treat trade secrets as if they were property,” and instead determined that the protected interest in a trade secret misappropriation claim was “the relationship between the parties at the time the secret is disclosed.” Monolith , 407 F.2d at 293. Because “[t]he fabric of the relationship once rent is not torn anew with each added use or disclosure, although the damage suffered may thereby be aggravated,” as the court in Monolith reasoned, “[t]he cause of action arises but once” and is subject to a single limitations period. Id.

The comments to the UTSA make clear that the UTSA’s statute of limitations specifically follows Monolith and “rejects a continuing wrong approach.” 1985 UTSA, § 6 cmt. Likewise, under the DTSA statute of limitations, “a continuing misappropriation constitutes a single claim of misappropriation.' 18 U.S.C. § 1836(d). This means that “the first discovered (or discoverable) misappropriation of a trade secret commences the limitation period, placing the focus . . . on the breach of the relationship between the parties at the time the secret is disclosed.' Kehoe Component Sales Inc. v. Best Lighting Prods., Inc. , 796 F.3d 576, 583 (6th Cir. 2015) (omission in original) (citation omitted) (applying Ohio law).

A variety of cases in California state and federal courts have interpreted this provision to mean that '[w]hen a plaintiff brings suit against an individual defendant for separate misappropriations of related trade secrets, or for multiple misappropriations of a single secret, the date of accrual is the date of the initial misappropriation.” HiRel Connectors, Inc. v. United States , 465 F. Supp. 2d 984, 988 (C.D. Cal. 2005) (applying California law). California’s approach, which is the most well-developed, appears to have originated in federal district court cases, most notably in Page 17Intermedics, Inc. v. Ventritex, Inc. , 822 F. Supp. 634 (N.D. Cal. 1993) (applying California law). The court in Intermedics held that 'when the statute of limitations begins to run on some of a plaintiff's trade secret claims against given defendants, the statute also begins to run at that same time as to other trade secret claims against those same defendants . . . at least when the plaintiff shared all the trade secrets with defendants during the same period and in connection with the same relationships and when the trade secrets concern related matter.' 822 F. Supp. at 657; see also Forcier v. Microsoft Corp. , 123 F. Supp. 2d 520, 525-26 (N.D. Cal. 2000) (applying California law); HiRel , 465 F.Supp.2d at 988. This approach 'impose[s] on plaintiffs a responsibility to take prompt and assertive corrective action with respect to all of plaintiffs' interests whenever plaintiffs detect a fracture in a once confidential relationship.' Intermedics , 822 F. Supp. at 654.

California state appellate courts have since approved the Intermedics approach to determine if related trade secrets relate to the same claim of misappropriation, most explicitly in MGA Ent., Inc. v. Mattel, Inc. , 254 Cal. Rptr. 3d 314 (Cal. Ct. App. 2019). There, the court rejected the plaintiff’s attempt to limit the accrual of its trade secrets claim to two specific trade secrets that were displayed at a particular toy fair, which the court described as a “compartmentalization of its suspicion of wrongdoing.' MGA , 254 Cal. Rptr. 3d at 321. It concluded that the statute of limitations had run against all 114 of plaintiff's alleged misappropriations, which involved multiple trade secrets, once the plaintiff had notice of an initial misappropriation, as taking a different approach would allow the plaintiff to “don blinders to avoid the accrual of the statute of limitation” for related acts of misappropriation. Id. at 320–21; see also Glue-Fold, Inc. v. Slautterback Corp., 98 Cal. Rptr. 2d 661, 666 (Cal. Ct. App. 2000) (applying California law) (“We agree with the federal court which stated that 'California law assumes that once a plaintiff knows or Page 18should know that a particular defendant cannot be trusted with one secret, it is unreasonable for that plaintiff simply to assume that that defendant can be trusted to protect other secrets.'' (quoting Intermedics , 822 F. Supp. at 654)).

The Sixth Circuit's opinion in B&P Littleford, LLC v. Prescott Machinery, LLC , No. 20-1449, 2021 WL 3732313 (6th Cir. 2021) (nonprecedential), is the only authority that Insulet cites in support of its trade-secret-bytrade-secret statute of limitations analysis. But that case recognizes the same “controlling principle consistent with the purposes of the statutory accrual rule” as Intermedics . B&P Littleford, 2021 WL 3732313, at *6 (quoting Gognat v. Ellsworth , 259 P.3d 497, 502 (Colo. 2011) (applying Colorado law)). There, to discern whether trade secrets should be subject to separate statute of limitations analyses, the Sixth Circuit described how a court “might ask whether the same relationship has been ruptured in the same way, looking, for example, at who made the disclosure, to whom the disclosure was made, and the nature, timing, and reasons for the disclosure.' Id. (citing Gognat , 259 P.3d at 503). While B&P Littleford characterized the inquiry to be whether the alleged trade secrets should be considered to be a single trade secret for limitations purposes, this is the substantially same test that the Intermedics court articulated. As such, there is no material difference in the controlling law on this issue that is advanced by either party.

Applying the Ohio Uniform Trade Secrets Act’s discovery rule in Maatuk v. Emerson Electric, Inc. , 781 F. App’x 1002, 1005 (Fed. Cir. 2019) (nonprecedential), we reached a similar conclusion. There, the plaintiff unsuccessfully brought a trade secret misappropriation suit against certain parties stemming from an alleged breach of a confidential relationship between the plaintiff and those parties. Maatuk , 781 F. App’x at 1003–04. Over a decade later, the plaintiff again alleged trade secret misappropriation stemming from the same breach of confidential relationship, but Page 19this time asserted that different trade secrets had been misappropriated. Id. at 1004. We agreed with the district court that the plaintiff’s misappropriation claim was timebarred because the plaintiff was “aware of a breach of a confidential relationship” when he filed his original lawsuit sixteen years prior. Id. at 1005. Differentiating Amalgamated Indus. Ltd. v. Tressa, Inc. , 69 F. App’x 255 (6th Cir. 2003) (nonprecedential), we held that “where a party alleges misappropriation of different trade secrets years after discovery of the breach of confidentiality, the claim is barred by the statute of limitations . . . .' Id.

Here, each of the factors outlined in Intermedics and B&P Littleford points toward a single accrual date for Insulet’s trade secret misappropriation claim against EOFlow. While serving as EOFlow’s consultant between March and May 2018, Mr. DiIanni, in a single course of alleged misappropriation, disclosed each asserted trade secret to EOFlow for the purpose of designing an improved EOPatch 2. EOFlow’s alleged misappropriations of the asserted trade secrets thus constitute a single claim of misappropriation and, therefore, are subject to the same statute of limitations analysis. See Gognat , 259 P.3d at 504 (concluding that a body of information that was “shared with the same person, in substantially the same time period, as part of the same joint commercial venture” was subject to a single statute of limitations analysis).

With these principles in mind, we must determine whether, under the access-plus-similarity standard and Merck discovery standard, Insulet knew or should have known of facts sufficient to plead access and similarity as to one of the asserted trade secrets before the critical date. This is a legal question with underlying fact issues. See Merck , 559 U.S. at 649, 652; see also Epstein , 460 F.3d at 188–89 (dismissing trade secret misappropriation claim on motion to dismiss as time-barred when parties only Page 20disputed “legal significance” of allegations); Skwira v. United States , 344 F.3d 64, 72 (1st Cir. 2003) (explaining that when parties “only disagree over the legal significance of [facts] . . . the parties’ dispute ‘focuses on pure (or nearly pure) questions of law [and] engenders de novo review'' (quoting Gonzalez v. United States , 284 F.3d 281, 287 (1st Cir. 2002)).

The district court instructed the jury that, once EOFlow proved misappropriation first occurred outside the limitations period, Insulet had the burden of proof to establish that it did not discover, or by the exercise of reasonable diligence could not discover, the misappropriation before the critical date. J.A. 21392. Neither party challenges this instruction on appeal, or that the first misappropriation of the CAD files, soft cannula, and occlusion-detection algorithm trade secrets occurred outside the limitations period, 7 so we proceed with the assumption that Insulet has the burden of proof on the discovery issue.

EOFlow points to Insulet’s August 3, 2023, complaint and argues that it “relied heavily on the same combination of access plus similarity that was fully available to Insulet” before the critical date. Appellants’ Br. 36 (citing J.A. 306– 367 ¶¶ 6–8, 40–42). We agree that we may appropriately consider the allegations in Insulet’s original complaint to determine what knowledge was necessary to sufficiently allege trade secret misappropriation.

As to access, Insulet’s complaint alleged that its former employee Steve DiIanni was 'entrusted with details of Page 21every aspect of the research and development efforts” at Insulet. J.A. 321 ¶ 40. According to Insulet, Mr. DiIanni “was in possession of detailed technical information about Insulet's Omnipod product, including but not limited to product specifications, material compositions, coating materials, plating specifications, operational software requirements and details, product functional requirements, dimensional tolerances, drive stroke and force requirements, and other product and manufacturing details” during his employment. J.A. 322 ¶ 41.

By March 15, 2019, which was more than three years before Insulet filed its original complaint on August 3, 2023, internal Insulet emails showed that Insulet knew Mr. DiIanni worked with EOFlow on the EOPatch 2. In those emails, Dave Nazzaro, Insulet’s Director of Mechanical and Materials COE, told Insulet’s Executive Vice President Eric Benjamin that “[t]he people running [EOFlow], Luis M[alave], Ian W[elsford], and their designer Steve D[iIanni] are past Insulet folks.” J.A. 43524. The undisputed facts thus demonstrate Insulet, by March 2019, had sufficient knowledge to allege that EOFlow hired a former employee to work on the EOPatch 2 who had knowledge of and access to Insulet’s trade secrets, which satisfies the access prong as a matter of law.

This case is similar to Seatrax , where the Fifth Circuit held that a trade secret owner’s knowledge that a former employee like Mr. DiIanni worked for the alleged misappropriator puts the trade secret owner on notice of access as a matter of law, as “‘it is not unexpected that a former employee will go to work for a competitor and that the competitor might thereby acquire trade secrets.'' Seatrax , 200 F.3d at 365 (quoting Comput. Assocs. Int’l, Inc. v. Altai, Inc. , 918 S.W.2d 453, 457 (Tex. 1996)).⁸ An individual like Page 22Mr. DiIanni is “a veritable repository of information on [his former employer’s] products and designs,” and, if “hired by [the alleged misappropriator] to design directly competing products, [the employee] would be hard pressed to avoid disclosing such information.' Leggett & Platt , 285 F.3d at 1361.

With respect to similarity, the plaintiff must possess knowledge of sufficient facts to allege that the defendant’s competing product is similar to at least one of the alleged trade secrets.

Evidence from Insulet’s own employees showed that Insulet understood that EOFlow had developed a similar, competing product before the critical date. Specifically, both Insulet and EOFlow attended a June 2018 American Diabetics Association ('ADA') Conference, at which attendee EOFlow displayed samples of EOPatch 2. Insulet took a photograph of the EOFlow booth, which included samples both inside and outside of a glass case. Following the conference, Insulet employee Jeffrey Miller sent an email to Insulet management stating that 'EOFlow has cloned our product . . . Our legal team is aware of this. Someone should request samples.” J.A. 43583. Similarly, Insulet engineer Ashutosh Zade stated in an email immediately after the conference that “[w]e need to see if this solution is based on our IP.” J.A. 54333. Another report from the conference by Insulet Senior Test Engineer Joe Gragido stated that EOFlow “make[s] a product called ‘EO Patch' that looks almost identical to Omnipod from the PDM to the GUI to the Pod . . . except it has 200 U insulin Page 23and it is ‘AP Ready.’” J.A. 43585. And an intelligence summary sent to Insulet management concluded that EOFlow had a “[v]ery similar product to Insulet’s OmniPod as it was started by a former Insulet engineer; similar form factor and reservoir size.” J.A. 43595. At trial, Eric Benjamin, Insulet's Chief Product and Customer Experience Officer, testified that the EOPatch 2 presented at the June 2018 conference “bore a stunning resemblance” to the Omnipod. J.A. 20589.

These admitted conclusions amount to more than “mere suspicion of misappropriation.' Dissent 3. The record shows that these conclusions caused Insulet to start investigating. Shortly after the 2018 ADA conference, Insulet personnel visited the EOFlow website twenty times. Insulet and EOFlow attended another ADA conference in 2019, where EOFlow similarly displayed samples. Insulet’s Deputy General Counsel, Jonathan Paris, attended the 2019 ADA conference. In the course of its investigation, Insulet discovered or should have discovered product similarity suggesting trade secret misappropriation.

The CAD-files trade secret concerns the design and manufacture of the Omnipod device and was defined as “CAD files for the Omnipod, including the file 13800-PODASSEMBLY-ASM.” J.A. 21387. In Insulet’s original complaint, Insulet alleged trade-secret misappropriation based on visual comparisons of the Omnipod device with a sample EOPatch 2 that Insulet obtained at a January 2023 trade show. Insulet relied on side-by-side images of “the similarities in physical components between Insulet's Omnipod and EOFlow’s EOPatch.” J.A. 310–12 ¶ 9; J.A. 347 ¶ 134. The illustrative examples of similarities in Insulet’s complaint included: (1) a drive system using a walking-man hook and ratchet gears with a large number of teeth per wheel; (2) a tube nut configuration to engage a drive rod in a plunger; (3) the insulin reservoir with a single O-ring for Page 24the plunger; and (4) a “nail head” cannula seal. Uncontradicted testimony and documents that were either known to Insulet or publicly available showed that each of these features was known or should have been known by Insulet before the critical date.

Walking-man hook and ratchet gears : Insulet's complaint describes similarities between the drive systems in the Omnipod and EOPatch 2, which include a so-called “walking-man” hook and ratchet gears. Insulet representatives visited the EOFlow booth at both the 2018 and 2019 ADA trade shows and took a picture of EOFlow’s display. At trial, EOFlow introduced physical samples of the EOPatch clear case prototypes available at the 2018 and 2019 ADA conferences. Referring to the physical exhibit of the sample from the 2018 ADA conference, Mr. Kim, the chief executive officer of EOFlow, testified that “[s]o what you see here, you can see very easily through the clear case [features] such as the walking man.” J.A. 20749–52 (emphasis added). Mr. Kim also identified 'the arm of the walking man' and 'the foot . . . that works between the sensors,” and stated “if you look at it carefully, you can see the teeth of the gears. So those are actually the gears in that ratchet gear.” J.A. 20751. Insulet did not provide contrary testimony.

EOFlow also introduced undisputed evidence that a prospectus created in connection with EOFlow’s IPO in Korea, which was publicly available by July 16, 2020, showcased various design features of the EOPatch. Eric Benjamin, Insulet’s Chief Product and Customer Experience Officer, testified that Insulet was “paying consistent attention” to EOFlow, and because “the IPO was a public announcement, our attention increased.” J.A. 20600. While the parties dispute whether Insulet reviewed the prospectus before the critical date, a reasonable company that is 'paying consistent attention' to its competitor should have known of major public announcements like the IPO and accompanying prospectus before the critical date.

Page 25J.A. 20600; see Ruth v. Unifund CCR Partners, 604 F.3d 908, 911 (6th Cir. 2010 ) ('When individuals have 'the means of discovery in [their] power,' they generally are 'held to have known it.'' (alteration in original) (quoting Wood v. Carpenter , 101 U.S. 135, 141 (1879))). Mr. Kim testified without contradiction that an image in the prospectus fully disclosed the walking man drive system, including the ratchet gears, in the EOPatch 2. J.A. 20766 (referring to J.A. 40205).

Tube nut configuration : The complaint also alleged similarities in the screws and nuts used to engage the plungers in the respective devices. Mr. Kim testified that 'the tube nut and the screws . . . eventually drives the plunger forward” in the insulin patch pump. J.A. 20773. In reference to the clear case sample of the EOPatch 2 from the 2018 ADA conference, Mr. Kim, again without contradiction, testified that the sample showed these “screws and the nuts.” J.A. 20751.

Reservoir and plunger o-ring : Another similarity between the Omnipod and the EOPatch 2 identified in Insulet’s complaint was an insulin reservoir with a single Oring for a plunger. Mr. Kim testified that the clear case sample showed “the reservoir [and] the plunger head Oring.” J.A. 20751. EOFlow employee Luis Malave also confirmed that the clear case sample showed the insulin reservoir. J.A. 20877. While Insulet argued that “there’s a white piece that blocks the view of the bottom of the reservoir,” J.A. 20877, and that the reservoir’s metal cap was removed on the clear case sample, it did not provide testimony to contradict Mr. Kim’s testimony that the white piece did not obstruct the view of any particular allegedly similar component of the reservoir, including the plunger.

“Nail-head” soft cannula seal : Finally, Insulet alleged in its complaint that the EOPatch 2 had a similar “nail-head” cannula seal to the Omnipod. Mr. Kim testified that EOFlow “display[ed] a poster at the 2018 ADA trade Page 26show . . . highlighting the soft cannula.” J.A. 20746–47 (referring to J.A. 55335). And, in particular, he testified that the prospectus had “images of [the EOFlow] soft cannula in EOPatch 2. . . . [T]his one has the flange or the nail head and the tipping that goes into the body.” J.A. 20765 (emphasis added) (referring to J.A. 40208). Indeed, Insulet’s technical expert admitted that the prospectus showed “features on the soft cannula.” J.A. 21042. Again, there was not contrary testimony.

In summary, the undisputed record shows that there were no material facts in dispute about the similarity between the CAD files and the EOPatch 2. The record shows that what was disclosed by examining the January 2023 sample was previously known as a result of the earlier disclosures. Therefore, the statute of limitations expired with respect to the CAD-files trade secret before Insulet brought suit, and the jury’s contrary conclusion was not supported by substantial evidence.

We next turn to the soft cannula and design history file trade secrets. Insulet asserted the soft cannula (which has its design included in the CAD-files trade secret) as a separate trade secret, defined as the “design and manufacturing process of the Omnipod soft cannula.' J.A. 21387. Having concluded that the statute has run as to the soft cannula design contained in the CAD-files trade secret, the statute of limitations has also run as to the specific soft cannula trade secret.

In addition, Insulet also included the CAD files in the scope of a broader compilation trade secret, the design history file. The design history file trade secret is defined as 'the design history file or DHF for the Omnipod Eros.' J.A. 21387. When 'the processes, procedures, or other pieces of information used in each separate act of misappropriation are related to each other as portions or phases of a single trade secret,' each act relates to the same Page 27misappropriation claim. Gognat , 259 P.3d at 503 (interpreting Colorado Uniform Trade Secrets Act). The statute of limitations thus begins to run for the entire trade secret whenever misappropriation of a portion of that trade secret is discovered or should have been discovered. Id. at 50405. If the statute of limitations has run on a trade secret, the statute of limitations has also run with respect to related, broader trade secrets of which it is a part. The statute of limitations therefore began to run with respect to the design history file trade secret whenever Insulet knew, or should have known, of sufficient information to plead a claim as to the CAD-files trade secret. This occurred before the critical date.⁹

Finally, we address the occlusion-detection algorithm trade secret. Occlusion detection is a standard feature of all insulin pumps. But a particular occlusion-detection algorithm may be a trade secret. The Omnipod’s occlusiondetection algorithm is a “built-in method for quickly and accurately identifying any occlusion” that prevents delivery of insulin, Appellee’s Br. 6, and is comprised of subcomponent algorithms.

Before the critical date, as discussed above, Insulet possessed or should have possessed sufficient knowledge of (1) EOFlow's access to Insulet's trade secrets through Mr. DiIanni; and (2) striking similarities between the CAD-files trade secret and the EOPatch 2, which was disclosed by Mr. DiIanni to EOFlow during the same time period and for the same purpose as the other asserted trade secrets (to design the improved EOPatch 2). Mr. DiIanni’s Page 28unauthorized disclosures to EOFlow, including his disclosure of the occlusion-detection algorithm trade secret, form the breach of confidence at the heart of Insulet’s trade secret misappropriation claim. See Monolith , 407 F.2d at 293; Kehoe , 796 F.3d at 583. As Insulet’s claim is timebarred with respect to the other trade secrets that were part of this disclosure, Insulet’s trade secret claim is also time-barred with respect to the occlusion-detection algorithm.

Importantly, this conclusion reflects a “controlling principle consistent with the purposes of the statutory accrual rule” for trade secrets that do not outwardly manifest in a misappropriator’s product. Gognat , 259 P.3d at 502. Such is the case when trade secrets are used internally to develop a product, like the occlusion-detection algorithm here. Otherwise, absent direct proof of a misappropriator’s access to the trade secret, 10 the statute of limitations would never expire for a trade secret that could not be discovered based on publicly available information or other information known to the plaintiff. The statute of limitations begins to run with respect to all trade secrets disclosed during substantially the same time to that defendant by the same person for the same purpose.

In summary, we conclude that the undisputed evidence establishes that the statute of limitations expired with respect to each trade secret before Insulet filed its complaint on August 3, 2023, and that the jury’s verdict to the contrary was unsupported by substantial evidence.

Largely through unexplained citations, Insulet argues that the record supports its position that it was not able to Page 29discover facts sufficient to bring its trade secret claim until after the critical date. We have examined each one of Insulet’s references cited in its brief, at oral argument, and in its letter to the court following oral argument. None of its citations contradicts the evidence that Insulet discovered or should have discovered facts sufficient to allege trade secret misappropriation before the critical date.

Nominal dimensions : First, Insulet points to testimony that, before the critical date, it could not discover the precise, nominal dimensions contained in the CAD files because the nominal dimensions shown in the CAD files “could not be gleaned from examining a physical sample.” Appellee’s Br. 5; see also Oral Arg. at 27:10-36:17.¹¹

As discussed earlier, knowledge of the precise nominal dimensions of a device are not needed to plead trade secret misappropriation, and Insulet itself did not plead facts related to the nominal dimensions of the EOPatch 2 in its original complaint based on the sample it examined in February 2023. Insulet did not need to know the dimensions of the EOPatch 2 to plead misappropriation.

Manufacturing processes of the cannula seal : Insulet relies on testimony from Luis Malave, an EOFlow employee who formerly worked for Insulet, who stated that one could not see “the method by which the cannula was Page 30manufactured' by looking at the 2018 ADA conference sample. J.A. 20877.¹² The manufacture of the soft cannula is only a portion of the soft cannula trade secret. Insulet could have discovered the similarity of the design of the EOPatch 2 cannula seal to its trade secrets before the critical date. In its complaint, Insulet alleged misappropriation based on the visual similarity between these features of the cannula seal on the Omnipod and EOPatch 2, rather than manufacturing processes. There is no testimony that knowledge of the misappropriation of the design of the soft cannula seal was not available prior to the critical date.

“Cut off” soft cannula : Insulet relies on testimony from several EOFlow witnesses who testified that the soft cannula was 'cut off' of the physical samples of the EOPatch 2 at the ADA conference in June 2018. J.A. 20780 (Jesse Kim); J.A. 20877 (Luis Malave). However, as discussed earlier, other disclosures, including EOFlow’s poster at the 2018 ADA conference and the prospectus, disclosed the relevant features of the soft cannula on the EOPatch 2 before the critical date. J.A. 20746–47 (referring to J.A. 55335); J.A. 21042.

Conclusory testimony : Finally, Insulet relies on generalized conclusory testimony from its own witnesses. Referring to the prospectus, an Insulet expert testified that, “[M]y opinion [is] [t]here is nothing there that would indicate . . . that a trade secret had been misappropriated.' J.A. 21042. Insulet also refers to testimony from its diligence expert that, in his opinion, Insulet could not send a cease-and-desist letter to EOFlow “until they had obtained a product and were able to see the inner workings of [the Page 31EOPatch 2].' J.A. 21321-22. Another Insulet witness testified that he would not “think that EOFlow had stolen Insulet’s trade secrets” until he saw “the inner workings” of the EOPatch 2, and that “as of 2020 there was still nothing to suggest that EOFlow was doing something different than following a different technology path to produce a product that was a lot like Omnipod.' J.A. 20540-41, 20587; accord J.A. 20524, 20583-84, 20586-88, 20598.

In light of the specific testimony as to the features discerned before the critical date, such generalized, conclusory testimony reflecting the subjective view of the witness (the opinion or “thinking” of the witness) is insufficient evidence to support the verdict. See Trs. of Boston Univ. v. Everlight Elecs. Co., Ltd. , 896 F.3d 1357, 1363 (Fed. Cir. 2018) ( “ [C]onclusory statements by an expert . . . are insufficient to sustain a jury's verdict.' (quoting MobileMedia Ideas LLC v. Apple Inc. , 780 F.3d 1159, 1172 (Fed. Cir. 2015)) (alterations in original)). The district court here should have granted JMOL because the statute of limitations expired before Insulet filed its complaint on August 3, 2023.¹³

Page 32The dissent does nothing to undermine this analysis. First, the dissent suggests that somehow this approach “induces an unfortunate, and premature, rush to the courthouse.' Dissent 3. Quite the contrary, the analysis is entirely in terms of the original complaint that was actually filed by Insulet (not the amended complaint that incorporated information secured through discovery) and simply asks whether the information alleged in Insulet’s own complaint to support Insulet's misappropriation claim had been discovered or should have been discovered before the critical date. That focus hardly prompts premature filing.

Second, the dissent questions whether allegations under the access-plus-similarity framework may sufficiently state a claim for trade secret misappropriation because several of the relevant cases do not involve the motion-todismiss stage. While the relevant cases arise in different procedural postures, they each address and confirm the relevant standard to plead and prove trade secret misappropriation claims.

Third, the dissent suggests that “the Opinion flips the [JMOL] standard on its head and accepts EOFlow’s view of the facts over Insulet’s,” Dissent 7, and “assumes . . . similarities were known before the critical date.” Dissent 9. We do no such thing. Rather, we conduct a detailed analysis of the Insulet testimony that allegedly supports a lack of access and similarity, and we conclude that it demonstrates a lack of substantial evidence to support the jury’s verdict.

Page 33As to access, the dissent simply suggests that Insulet was unaware before the critical date whether Insulet’s former employees took relevant documentation with them when they left the company. Dissent 7–8. As outlined in Oakwood Laboratories and Leggett & Platt , the relevant inquiry is the 'competitor's access to the plaintiff's former employee,” Oakwood Lab’ys , 999 F.3d at 912 n.19, particularly when that employee, like Mr. DiIanni, is a “veritable repository of information on [plaintiff]’s products and designs,” Leggett & Platt , 285 F.3d at 1361. Under the relevant cases, knowledge that specific documentation has been misappropriated is unnecessary. In any event, the original complaint said nothing about Mr. DiIanni or any other former Insulet employee's misappropriation of Insulet documents, which are facts that Insulet admittedly only learned after discovery and incorporated in the amended complaint. See Appellee’s Br. 16; Appellee’s Principal Br. at 9–12, Insulet I (No. 24-1137), Dkt. No. 29 (discussing evidence of document retention and disclosure produced during expedited preliminary discovery). Insulet itself thought this evidence was unnecessary to allege trade secret misappropriation in its original complaint.

As to similarity, the dissent recognizes that the “evaluation of the similarity prong must be a fact-intensive inquiry that carefully considers the context and type of trade secret at issue.” Dissent 4. But the dissent largely fails to conduct this fact-intensive analysis or to address the factual analysis necessary for substantial evidence review.

The dissent suggests there is a conflict in testimony that preserves the jury’s verdict. Dissent 9–11. There is no such conflict. The specific testimony on which the dissent relies fails to show Insulet lacked knowledge of similarity between the Omnipod and the EOPatch 2 before the critical date. The dissent identifies Insulet’s claimed lack of knowledge as to manufacturing details of the EOPatch 2, Dissent 9 (citing J.A. 20583–84); id. at 11 (citing J.A. 21041–42), but knowledge of these details is Page 34unnecessary to state a claim for misappropriation. See supra section III. The dissent also relies on testimony that the soft cannula was “cut-off” of the ADA conference samples. Dissent 10 (citing J.A. 20780, 20877). As discussed above, the EOPatch 2’s soft cannula was fully disclosed before the critical date, with Insulet’s own expert admitting that the features of the soft cannula were disclosed in the prospectus. See supra section V.B.2; see also J.A. 21042. The remaining testimony relied upon by the dissent, all of which came from Insulet employees, Dissent 9–10 (citing J.A. 20541, 20583–84, 20586–88) is conclusory and subjective. See supra section V.B.5

In summary, the record shows that Insulet personnel were close enough to the EOFlow samples to take photographs at the 2018 ADA conference and observed striking similarities between the products; that Insulet personnel visited the EOFlow website twenty times in the days after the 2018 ADA conference; that Jonathan Paris, Insulet’s Deputy General Counsel, examined the EOFlow booth at the 2019 ADA conference; and that images of features of the EOPatch 2 were publicly available in the prospectus before the critical date. None of the testimony raised in the dissent rebuts the undisputed evidence that Insulet knew or should have known of the similarities between the asserted trade secrets and the EOPatch 2 before the critical date. Insulet’s claim is thus time-barred.

For the foregoing reasons, we reverse the judgment because the statute of limitations for Insulet’s DTSA claim expired before Insulet brought suit. We therefore decline to address the other issues EOFlow raises on appeal.

COSTS

Costs to EOFlow.

Page 35United States Court of Appeals for the Federal Circuit

2025-1807

Appeal from the United States District Court for the District of Massachusetts in No. 1:23-cv-11780-FDS, Judge F. Dennis Saylor, IV.

PROST, Circuit Judge , dissenting.

After a five-week trial in which over twenty witnesses testified and more than 300 exhibits were admitted into evidence, the jury found that none of the misappropriated trade secrets were time-barred. Yet, the majority overturns the jury's verdict based on legal theories that EOFlow neither presses nor develops. The majority errs in four main ways: (1) it conflates its application of the discovery rule with the inquiry-notice standard, (2) it adopts a framework not supported by the case law, (3) it encourages plaintiffs to race to the courthouse with undeveloped claims based on mere suspicion, and (4) it improperlyPage 36 substitutes our own fact findings for those of the jury. I respectfully dissent.¹

The debate in this case begins with which standard applies to the DTSA’s statute-of-limitations provision: the inquiry-notice standard or the discovery rule? That answer matters because it determines when the clock starts for the statute of limitations. Under the inquiry-notice standard, the clock starts when a reasonably diligent person would have begun to investigate a potential misappropriation. Under the discovery rule, the clock starts when a reasonably diligent person discovered (or should have discovered) the misappropriation.

Rather than address this issue head on, the majority sidesteps the parties’ dispute and concludes that “even under the more demanding Merck standard [(i.e., discovery rule)], the statute of limitations expired before Insulet filed its complaint.” Maj. 10. My concern with the majority’s approach is two-fold: (1) it is clear from the plain text of 18 U.S.C. § 1836(d) that the discovery rule applies, not the inquiry-notice standard, and (2) the majority’s application of the discovery rule does not appear meaningfully different from the inquiry-notice standard.

As to my first concern, a claim under the DTSA “may not be commenced later than 3 years after the date on which the misappropriation with respect to which the action would relate is discovered or by the exercise of reasonable diligence should have been discovered .' 18 U.S.C. § 1836(d) (emphasis added). This is the discovery rule, not the inquiry-notice standard. There should be no reasonable debate or hedging about which standard governs here.

Page 37As to my second concern, the majority focuses its analysis on when Insulet should have begun its investigation (i.e., when the clock starts under the inquiry-notice standard), not when Insulet reasonably should have discovered the misappropriation (i.e., when the clock starts under the discovery rule). See, e.g. , Maj. 23. Those are two separate events. The majority’s apparent application of the discovery rule raises questions as to whether it sees a meaningful difference between those standards. And, to be sure, there is a difference. As I explain below, under proper application of the discovery rule, Insulet’s complaint was timely filed (or so the jury might have reasonably found).

I agree with the majority that the access-plus-similarity framework helps determine when a reasonably diligent plaintiff should have discovered the misappropriation, but I do not agree with how the majority defines the accessplus-similarity framework and applies it to the facts in this case. First, the majority’s characterization (and application) of the access-plus-similarity framework is not supported by the case law and induces a premature, and unfortunate, rush to the courthouse based on mere suspicion. Second, the majority’s analysis overturning the jury’s statute-of-limitations findings is based on a mistaken view that those findings were unsupported by substantial evidence. By deeming all asserted trade secrets time-barred, the majority substitutes its own view of the facts for that of the jury.²

The majority adopts an access-plus-similarity framework not supported by the case law. It announces that wePage 38 have held “the plaintiff possesses sufficient facts to state a claim for trade secret misappropriation” “if a plaintiff has knowledge that (1) the alleged misappropriator has access to the alleged trade secrets through a former employee of the trade secret owner and (2) there is similarity between the trade secrets and the alleged misappropriator’s competing product.' Maj. 13-14 (citing Leggett & Platt, Inc. v. Hickory Springs Mfg. Co. , 285 F.3d 1353 (Fed. Cir. 2002)). That framework is not supported by the case law. And, practically speaking, that can’t be the framework.

“Access” must require more than just the fact that a former employee who had access to trade secrets now works for the defendant (otherwise, premature suits will proliferate), and evaluation of the similarity prong must be a fact-intensive inquiry that carefully considers the context and type of trade secret at issue.

Contrary to the majority’s suggestion, Leggett does not address the pleading standard for trade secret cases; rather, it concerned the district court's grant of summary judgment of no misappropriation of trade secrets.³ We also Page 39did not affirmatively hold that “access to the alleged trade secrets through a former employee of the trade secret owner” meets the access prong. Contra Maj. 14. Rather, we concluded that 'the district court prematurely dismissed [the trade-secret holder’s] evidence on access and similarity” because there were genuine issues of material fact. Leggett , 285 F.3d at 1361.

The majority’s reliance on Seatrax to support its application of the access-plus-similarity framework fares no better. In Seatrax , the defendant not only previously had access to the trade secrets, but there was also evidence that the trade-secret holder knew the defendant had wrongly taken the trade secrets. Seatrax, Inc. v. Sonbeck Int’l, Inc. , 200 F.3d 358, 365 (5th Cir. 2000) (noting trade secret drawings were missing from premises). And while the majority cites Sokol , that case undermines the majority's characterization (and application) of the access-plus-similarity framework. There, even though the trade-secret holder knew the misappropriated product “was similar to its own,” 'these apprehensions were still just concerns and suspicions rather than knowledge of misuse.' Sokol Crystal Prods., Inc. v. DSC Commc’ns Corp. , 15 F.3d 1427, 1430 (7th Cir. 1994) (emphasis in original) (concluding the statute of limitations did not start accruing until the misappropriated products were sold to a third party).

The majority also, in my view, does not accurately characterize Insulet's alleged concessions during oral argument. Maj. 14. The majority cites Insulet’s agreement thatPage 40 the access-plus-similarity framework applies in this case, but that agreement does not extend to how the majority defines and applies that framework. According to Insulet, the access prong could not be met here because Insulet did not know that its former employees had taken the tradesecret documents with them and given them to EOFlow before the critical date. Oral Arg. at 28:27–29:49.

Finally, from a practical standpoint, the majority’s access-plus-similarity framework would arguably produce absurd results. See Sokol , 15 F.3d at 1430 ('At least in the commercial setting at issue here, suspicion and fear are not sufficient predicates for launching a lawsuit, and to file an action with no other basis would offend norms articulated in rules like Federal Rule of Civil Procedure 11.” (cleaned up)). Employees frequently move between companies working on similar products. Under the majority’s framework, a plaintiff’s statute-of-limitations clock starts prematurely whenever a former employee joins a competitor that makes a superficially similar product. Clearly, that cannot be correct.

Applying what I consider to be the proper access-plussimilarity framework, I would conclude that ample evidence supports the jury’s finding that Insulet did not discover, nor reasonably should have discovered, the misappropriation of the asserted trade secrets before the critical date—August 2020. Yet, rather than view the evidence in the light most favorable to Insulet—as the governing standard of review here requires, Dimanche v. Massachusetts Bay Transportation Authority , 893 F.3d 1, 8 (1st Cir. 2018)—the majority flips the standard on its head and accepts EOFlow’s view of the facts over Insulet’s.

As to the access prong, the majority concludes that “undisputed facts” demonstrate that by March 2019, InsuletPage 41 “had sufficient knowledge to allege that EOFlow hired a former employee to work on the EOPatch 2 who had knowledge of and access to Insulet’s trade secrets.” Maj. 21. That conclusion is doubly flawed. First, starting the statute-of-limitations clock by March 2019 is too early. As discussed above, the mere fact that Insulet’s former employee had knowledge of and access to Insulet’s trade secrets and subsequently worked on EOFlow's EOPatch 2 alone is insufficient to satisfy the access prong. And the two paragraphs of the complaint the majority cites, Maj. 21 (quoting J.A. 321-22 ¶¶ 40-41), only show that Insulet's former employees knew about the trade secrets while at Insulet .

Second, there are disputed facts concerning the access prong that were properly before the jury. For example, Insulet’s senior executive Eric Benjamin testified that “[i]t’s common practice to hire folks who have relevant industry experience. We hire folks for their experience, and if they join us, we expect them to abide by whatever confidentiality requirements they had from a prior employer.” J.A. 20584 (119:12–19); see also J.A. 20584 (119:20-22).⁴ And it was not until after the critical date that Insulet knew its former employees had wrongly transferred thousands of confidential Insulet documents to their personal computers and later provided them to EOFlow. See J.A. 20540 (148:17–24); J.A. 20584 (119:12–22); J.A. 20586 (126:21-127:3); J.A. 20587 (130:25-131:7); J.A. 20588 (134:17-136:11); J.A. 20598 (20:13-15); J.A. 20618-19 (100:4–103:2); J.A. 21119 (238:20–239:2); J.A. 21327 (61:22–62:2). Subsequent revelations prompted Insulet to amend its complaint further to allege that during the course of Insulet's former employees' employment andPage 42 consultancy agreements, 'Defendants DiIanni and Welsford improperly transferred thousands of confidential Insulet documents to their personal computers' and later provided those documents to EOFlow. Amended Complaint ¶¶ 50, 61, Insulet Corp. v. EOFlow, Co. , No. 23-11780 (D. Mass. Oct. 27, 2023), Dkt. No. 159. Those allegations were not in the original complaint.⁵

As to the similarity prong, the majority concludes that the 'undisputed evidence establishes that the statute of limitations expired with respect to each trade secret” before the critical date and “that the jury’s verdict to the contrary was unsupported by substantial evidence.” Maj. 28. To reach that conclusion, the majority conducts its own evaluation of the credibility of witnesses and the weight of the evidence by treating all evidence in favor of EOFlow as undisputed (it was not) and entirely credible (which is for the jury to decide), and any contrary evidence presented by Insulet as allegedly conclusory or too general to support the jury’s verdict. On this record, a reasonable jury could have found in favor of Insulet.

The majority liberally cites Insulet’s complaint and appears to accept the (unproven) premise that the allegations in Insulet's complaint were known or should have been known to Insulet more than three years before it was filed. See Maj. 23-26; see also Appellants’ Br. 36. Yet the allegations in the complaint concerning events before the critical date do not necessarily reflect the moment when Insulet should have known that the misappropriation occurred (absent some statement in the complaint suggesting otherwise). Rather than drawing unwarranted inferences from the complaint , I would instead defer to the results of the Page 43jury trial (including the facts developed at trial) on this issue. And rather than draw inferences from the complaint that are unfavorable to Insulet, I would view the facts developed at trial in the light most favorable to Insulet-the prevailing party—as is required by our standard of review.

For example, the majority relies on side-by-side images of Insulet’s Omnipod and EOFlow’s EOPatch 2 in the complaint and assumes those similarities were known before the critical date. Maj. 23-26. Facts developed at trial, however, show that EOFlow was evasive when Insulet approached EOFlow’s trade show booth and that EOFlow prevented Insulet from inspecting the small device and its intricate inner workings up close. J.A. 20583 (116:12–15), J.A. 20584 (117:19–118:3); see also J.A. 41453 (“[n]eed[ed] to keep [Insulet] guessing on our corporate plans”); J.A. 41605 ('[j]ust need[ed] to find out what [Insulet] want[ed] while keeping a low profile on our own plans”). I therefore don’t see how the majority can credibly state that “[u]ncontradicted testimony and documents that were either known to Insulet or publicly available showed that each of these features [in the complaint] was known or should have been known by Insulet before the critical date.” Maj. 24.

There was additional evidence presented to the jury showing that Insulet reasonably should not have known about EOFlow’s misappropriation before the critical date. For example, the similarities in appearance between the Omnipod and EOPatch 2 did not necessarily mean that EOFlow misappropriated any trade secrets, especially given the type and size of technology at issue here. See, e.g. , J.A. 20541 (149:8–150:6) (Mr. Benjamin testifying “the fact that EOPatch [2] had similarities to the way” the Omnipod looked did not make Insulet think EOFlow had stolen Insulet’s trade secrets because “[p]eople are trying to build patch pumps, they’re all going to be in that general size and shape'); see also J.A. 20586 (126:21-127:3), J.A. 20587 (130:25-131:7), J.A. 20588 (134:17-135:12).

Page 44And before the critical date, the publicly available information suggested that EOFlow was working on products that used technology different from the Omnipod, reinforcing Insulet’s understanding that EOFlow was lawfully developing a device using different technology. See, e.g. , J.A. 20587 (130:25–131:7) (“As of 2020 there was still nothing to suggest that EOFlow was doing something different than following a different technology path to produce a product that was a lot like Omnipod.”); see also J.A. 54333, J.A. 21041–42 (102:21–106:22). EOFlow’s witness also testified that looking at the samples provided at the 2018 and 2019 ADA conferences, he could not 'see the method by which the cannula was manufactured by looking at these devices” and “the cannula was cut off of the samples that were used at the ADA conferences.” J.A. 20877 (111:18–23); see also J.A. 20780 (137:7-14). Thus, there were disputed facts properly before the jury.

The majority belittles Insulet employees’ testimony as “conclusory and subjective.” Maj. 34. As discussed above, I disagree that the testimony is conclusory. And of course Insulet employees' testimony was 'subjective,' just as EOFlow employees’ testimony was subjective. Both sides had opportunities to cross-examine each other’s witnesses, and it was for the jury, not this court, to weigh the credibility of the witnesses’ statements.

The majority also holds that Insulet could have discovered the misappropriation if only it had pored over a predominantly Korean-language, 500-page prospectus filed with Korean financial regulators in July 2020-mere weeks before the critical date of August 3, 2020. Maj. 24-26. The standard is reasonable diligence, not maximum feasible diligence. There was also a factual dispute as to whether the prospectus contained a sufficient level of relevant details to indicate possible misappropriation. See J.A. 21041–42 (104:16–106:22).

Page 45Purportedly applying the discovery rule, the majority concludes that the events that transpired at (or shortly after) the 2018 ADA conference “caused Insulet to start investigating.” Maj. 23. The majority points to the fact that 'Insulet personnel visited the EOFlow website twenty times,” and “EOFlow attended another ADA conference in 2019.' Maj. 23. It was for the jury to weigh that evidence and determine whether these so-called investigations resulted in Insulet discovering, or if Insulet reasonably should have discovered, the trade-secret misappropriation before the critical date.

* * *

The majority’s critique of this dissent for failing to conduct the fact-intensive analysis just proves my point. Maj. 33. That’s not this court’s role. When reviewing challenges to the sufficiency of the evidence supporting the jury’s verdict, we must view the evidence in the light most favorable to Insulet and draw “all factual inferences and resolving all credibility determinations in [its] favor.” Dimanche , 893 F.3d at 4 n.2. The analysis the majority seeks from the dissent was for the jury to conduct, not this court.

I write separately to address the occlusion-detection algorithm trade secret. The majority concludes that this trade secret is time-barred on a theory not advanced by EOFlow on appeal. Maj. 14–19, 27–28. In any event, I don’t think we need to reach that theory if the majority had properly applied the discovery rule on these facts (and in this posture).

To conclude that the occlusion-detection algorithm trade secret is time-barred, the majority decides what appears to be an issue of first impression in the First Circuit: whether the statute of limitations inquiry should be evaluated on a trade-secret-by-trade-secret basis. The majority concludes that the facts here 'point[] toward a singlePage 46 accrual date for Insulet's trade secret misappropriation claim against EOFlow.” Maj. 19.

I am uncomfortable embracing a theory not advanced by EOFlow and that was not briefed before this court, particularly on an issue of first impression in the First Circuit. The district court instructed the jury to analyze the statute-of-limitations inquiry trade-secret-by-trade-secret, J.A. 21392 (173:12-13; 174:14-15), and EOFlow did not challenge those instructions on appeal. I wouldn't have gone where the parties did not.

The majority suggests it’s duty-bound to address this issue because “whether a party is entitled to JMOL does not depend on how the jury was instructed. The inquiry is whether under the controlling law, the defendant is entitled to JMOL.” Maj. 15 (cleaned up). But the majority isn’t consistent in deciding when to address the proper “controlling law.” Later in its analysis, it declines to address the burden-shifting jury instruction because neither party challenged that instruction on appeal. See Maj. 20.

In any event, I do not believe we need to reach this issue because, in my view, under the proper application of the discovery rule, none of the trade secrets are timebarred. See Part II. The occlusion-detection algorithm trade secret was not alleged in Insulet’s original complaint. Insulet discovered this misappropriation during discovery . And EOFlow only used Insulet’s occlusion-detection algorithm trade secret while developing the EOPatch 2; it was not used in the final product. J.A. 72–73. Even EOFlow’s witnesses admitted that there was 'nothing proprietary about the Omnipod algorithm [i.e., the occlusion-detectionalgorithm trade secret] reflected in” the Korean-language prospectus, J.A. 21198 (85:15-87:7), nor would someone simply looking at a sample of the EOPatch 2 at the 2018 or 2019 ADA conferences have been alerted to suspect misappropriation of the occlusion-detection-algorithm trade secret, J.A. 20877 (111:4–23). I would conclude that given Page 47the trial record before the jury, the district court did not err in denying EOFlow’s JMOL motion as to the occlusion-detection algorithm trade secret.

Because I conclude that substantial evidence supports the jury verdict for all asserted trade secrets, I would have reached EOFlow’s arguments concerning (1) the specificity of Insulet’s design history file trade secret, (2) the district court's reasonable measures jury instruction, and (3) its challenge to the damages award. On the design-historyfile-trade-secret issue, I would have agreed with EOFlow that Insulet’s design-history-file trade secret was severely overbroad. I also see no error in the district court’s jury instructions related to reasonable measures. On the damages issue, I would have held that the district court properly awarded Insulet avoided-cost damages based on the research and development costs EOFlow saved by misappropriating Insulet's trade secrets. These are costs EOFlow would have otherwise incurred to possess and use the trade secrets. That award is legally permissible and not duplicative. I also see no abuse of discretion by the district court in imposing joint-and-several liability on Mr. Kim.

The jury’s role as fact-finder lies at the heart of our judicial system. Both Insulet and EOFlow put forth evidence supporting their respective positions concerning whether Insulet should have discovered EOFlow’s misappropriation before the critical date. The majority’s characterization of EOFlow’s evidence as “undisputed” or “uncontradicted” is unsupported and would undermine the critical fact-finding role of the jury. I respectfully dissent.