Carnegie v. Fenix Diamonds: When Patent Litigation Becomes an Exceptional Case
A patent case can become expensive for reasons that have little to do with the damages originally sought. In Carnegie Institution of Washington v. Fenix Diamonds LLC, the U.S. Court of Appeals for the Federal Circuit affirmed an award of more than $3.24 million in attorney fees, non-taxable expenses, and post-judgment interest after the district court concluded that the plaintiffs continued pursuing infringement positions that the evidence and the court’s claim construction could not support.
The September 17, 2026 Federal Circuit opinion is expressly designated nonprecedential. It therefore does not announce a new binding rule. It is nevertheless a useful warning for patent owners, exclusive licensees, and accused infringers: litigation positions must be reassessed when claim construction, discovery, or expert work changes the factual landscape. A theory that may have been colorable at filing can create substantial fee exposure if it becomes objectively baseless and the party keeps pressing it.
What happened in the Carnegie patent case?
Carnegie Institution of Washington owned two patents relating to methods for making lab-grown diamonds. Carnegie had licensed the patents to M7D Corporation, including enforcement rights. Carnegie and M7D sued Fenix Diamonds in the Southern District of New York, alleging that diamonds manufactured by Fenix’s supplier infringed the patents.
One patent concerned chemical-vapor-deposition methods for growing single-crystal diamonds. The other concerned a high-temperature, high-pressure annealing process intended to improve a diamond’s optical clarity. As often happens in patent litigation, the dispute turned on both the language of the asserted claims and technical evidence about the accused manufacturing process.
The district court construed disputed claim terms, including “growth surface” and “single-crystal diamond.” During discovery, Fenix produced an affidavit, images, schematics, manufacturing information, and other materials from its supplier. According to the Federal Circuit’s opinion, that evidence showed extensive non-monocrystalline growth in the accused process and showed that the supplier’s annealing process operated outside a pressure range required by the other asserted patent.
The district court ultimately entered summary judgment of noninfringement. It later found the case exceptional under 35 U.S.C. § 285 and awarded Fenix $2,547,594.94 in attorney fees and $693,119.72 in non-taxable expenses. With post-judgment interest, the judgment exceeded $3.24 million. The Federal Circuit affirmed.
Section 285 does not require misconduct in every case
Section 285 states that a court may award reasonable attorney fees to the prevailing party in an “exceptional” patent case. Under the Supreme Court’s framework, a case may be exceptional because it stands apart from others based on the substantive weakness of a party’s litigating position, the unreasonable manner in which the case was litigated, or the totality of the circumstances.
That distinction matters. A court does not necessarily need to find fraud, intentional misconduct, or independently sanctionable behavior before shifting attorney fees. In Carnegie, the district court found both objectively baseless infringement positions and unreasonable litigation conduct. The Federal Circuit affirmed the exceptional-case determination based on the substantive weakness of the infringement positions and did not need to reach the separate unreasonable-conduct ground for the Section 285 award.
The review posture also matters. Exceptional-case determinations are reviewed for abuse of discretion. That deferential standard gives trial judges significant room to evaluate the strength of the positions, the record developed before them, and the parties’ conduct over the life of the case.
The critical question was when the infringement theories stopped being reasonable
The decision did not treat the filing date as the only relevant checkpoint. The district court awarded fees and expenses for work performed after July 14, 2020—one month after the plaintiffs received evidence about the supplier’s manufacturing process. It found that, by that point, the infringement positions had become objectively baseless.
For the first patent, the problem was the fit between the evidence and the court’s construction of the claims. The plaintiffs’ expert advanced a view of “growth surface” that, in the courts’ assessment, effectively excluded material the claim-construction order had said must be included. The Federal Circuit agreed that the infringement position remained contrary to the controlling construction even after the expert attempted to clarify it.
For the second patent, the accused process did not operate at the pressure required by the claim. The opinion noted that the plaintiffs presented no infringement expert report for that patent and eventually stopped asserting it. On that record, the Federal Circuit held that the district court did not abuse its discretion in identifying a point at which continued pursuit of the claims was no longer reasonable.
This is why patent claim construction cannot be treated as a self-contained phase of litigation. A Markman ruling changes the legal framework for infringement and invalidity. The parties then need to map the evidence to the claims as construed—not to the construction they preferred or expected to obtain.
Attorney fees and non-taxable expenses required different legal grounds
The award included more than ordinary attorney fees. The district court also shifted non-taxable expenses, including expert-related expenses, under its inherent authority. That required a finding of bad faith or vexatious conduct, a higher threshold than the Section 285 exceptional-case standard.
The Federal Circuit upheld that exercise of inherent power. The district court had found, among other things, that the plaintiffs prolonged the case after they must have known their infringement positions were baseless, made a misleading representation concerning access to the supplier’s facility, waited until late in the case to abandon one patent, and improperly asserted privilege over a presentation to third-party lenders.
The appellate court also approved using July 14, 2020 as the dividing line for expenses. Under the causation principles discussed in the opinion, a court may shift all fees or expenses from a particular point when it finds that the case would have ended then but for the misconduct. The court did not have to connect each individual billing entry to a separate act of misconduct.
A patent owner may share liability with its licensee
One of the decision’s most practical lessons concerns the relationship between a patent owner and an enforcement licensee. Carnegie argued that it was merely a passive owner joined for standing purposes and should not be liable for conduct attributed to M7D. The Federal Circuit rejected that position.
The owner and licensee had participated together from the beginning, shared counsel, filed joint submissions, and presented unified arguments. Their patent license also gave Carnegie a share of litigation recoveries, required consultation, required its written consent before a voluntary final disposition, and required cooperation in litigation. Those facts supported the conclusion that the parties had a close, intertwined relationship and could be held jointly and severally liable.
For licensors and licensees, enforcement provisions are not boilerplate. Control, consultation, settlement authority, cooperation duties, cost allocation, indemnification, and responsibility for adverse fee awards should be addressed deliberately. A party that expects to benefit from enforcement while retaining meaningful control may not be able to characterize itself as a detached observer when the litigation produces an adverse fee judgment.
Practical lessons for patent owners and accused infringers
Reevaluate the case at defined checkpoints
A pre-suit claim chart is the beginning of the analysis, not the end. The infringement and validity positions should be revisited after claim construction, key document productions, source-code or product inspections, important depositions, and opening expert reports. The question is not simply whether some argument remains available. It is whether the argument remains grounded in the controlling claim construction and admissible evidence.
Make the expert theory match the court’s construction
Experts apply claim constructions; they do not get to rewrite them. Counsel should test every expert opinion against the precise language of the court’s order and identify any place where the analysis depends on a rejected interpretation. If the evidence does not satisfy a required limitation, narrowing the case can be less costly than preserving a weak theory through summary judgment.
Document the decision to continue, narrow, settle, or stop
When significant new evidence arrives, the legal team and client should make a disciplined assessment of what changed. A contemporaneous record of the claim analysis, factual support, open questions, and business objectives can improve decision-making. It does not immunize a party from a fee motion, but it helps prevent momentum from substituting for analysis.
Treat discovery representations and privilege assertions carefully
The fee and expense rulings were not based only on the ultimate noninfringement result. The district court also evaluated representations about access to evidence, the timing of abandoning a patent, and a privilege assertion. Accuracy and candor in discovery can affect exposure far beyond the immediate dispute.
Plan owner–licensee enforcement before filing
A patent owner and exclusive licensee should understand who controls the lawsuit, who approves settlement, who pays litigation costs, and who bears the risk of fee shifting. Those terms should be considered when negotiating the license, not for the first time after a dispute arises. Tucker Law’s strategic counseling work connects litigation decisions with licensing and business objectives.
What the decision does—and does not—mean
Carnegie does not mean that losing a patent case automatically makes it exceptional. Section 285 calls for a case-specific assessment, and reasonable positions can fail without justifying fee shifting. Nor does the decision establish a new binding rule, because the Federal Circuit designated it nonprecedential.
Its value lies in showing how courts may examine the life of a case. A claim can begin with a plausible basis, then become unsustainable after a construction or production of evidence. Fee exposure may be measured from that turning point. The decision also shows that a patent owner’s contractual and practical relationship with an enforcement licensee can affect responsibility for the resulting judgment.
Patent litigation requires continual alignment among the claim language, the technical record, the expert analysis, and the client’s commercial goal. Tucker Law helps patent owners and accused businesses evaluate those connections and make informed decisions as the record develops. To discuss an infringement claim, defense, licensing arrangement, or litigation-stage reassessment, contact Tucker Law.
This article provides general information and is not legal advice. The outcome of any patent dispute depends on its facts, claims, evidence, agreements, and procedural posture.
