Lufthansa v. Astronics: Disgorgement of Profits, Part 2

This second post on the Court of Appeals for England and Wales’ recent decision in Lufthansa Technik AG v. Astronics Advanced Electronic Systems, [2026] EWCA Civ 964, will focus on Lord Justice Arnold’s analysis of the lower court’s award of a portion of the profits earned by defendants Astronics, Panasonic, and Safran.  For the facts of the case, see my first post in this series, published last Thursday.

The principal issue under consideration, as the court states, is how to “distinguish between ‘the profits derived . . . from the infringement’ and profits which are not ‘derived . . .  from the infringement’” (para. 78).  To answer this question, the opinion sets out to analyze the relationship between a “differential profits” approach and apportionment.  Paragraphs 80 through 82 state the parties’ competing positions:

80. Lufthansa contends that, in most cases, the correct way in which to distinguish between “profits derived … from the infringement” and profits which are not “derived … from the infringement” is by ascertaining the difference between the profits which the defendant made from its infringing activities and the profits which the defendant would have made from the nearest available NIA [noninfringing alternative]. This type of analysis is most commonly referred to in the case law and the academic literature as a “differential profits” analysis, although other terms (such as “incremental profits”) are sometimes used. Once a differential profits analysis has been undertaken, Lufthansa contends that there is no legal or economic justification for applying an apportionment of the profits: these are alternative approaches to the isolation of profits caused by the infringement. Lufthansa accepts that there may be cases in which apportionment is appropriate, but it argues that a differential profits analysis is usually preferable. What is not legitimate is to apply both. Lufthansa also accepts that, having carried out a differential profits analysis, it remains necessary to ask whether any of the resulting profits were too remote (in the broad sense) from (i.e. not legally caused by) the defendant’s acts of infringement. Lufthansa argues that this addresses the problem of long chains of causation, and does not justify apportionment.

 

81. Lufthansa contends that, since the judge found that the Defendants had failed to establish any NIA, the result of the differential profits analysis is that all of the profits in question “derived … from the infringement”. The Defendants’ failure to establish an NIA demonstrates the Patent was a “gateway” patent which controlled access to the relevant market. Accordingly, Lufthansa says, it is just that Lufthansa should recover all of the profits generated as a result.

 

82. Astronics and Panasonic contend that apportionment is a well-established approach in English law to determining what profits are “derived … from the infringement”, and that the use of apportionment is not precluded by a finding that all of the profits in issue were factually caused by the infringing acts. . . . It would be unjust for Lufthansa to recover all of the profits in issue, because there were multiple causes for the generation of those profits. Although infringement of the Patent was necessary for those profits to be generated, it was not sufficient.

The opinion then spends several pages (paras. 84-132) discussing the relevant common-law precedents, including (among many others) Celanese International Corp v BP Chemicals Ltd [1999] RPC 203, a case I discuss and critique at some length in my book Comparative Patent Remedies, and Nova Chemical Corp v Dow Chemical Co [2022] SCC 43 (Can.), a decision that Norman Siebrasse discusses and critiques in two papers, Nova v. Dow: Intuition or Principle in the Accounting of Profits Remedy, Part I, 35 I.P.J, 249 (2023); Norman V. Siebrasse, Nova v. Dow: Intuition or Principle in the Accounting of Profits Remedy, Part II, 36 I.P.J. 81 (2023).  The opinion also discusses and critiques some academic works that Siebrasse and I, among others, have either authored or coauthored (paras. 133-42).  The opinion’s principal critique of some of this work is that it doesn’t answer the question of how a court should proceed when the evidence does not indicate what the NIA would have been.  (Later in the opinion, as noted below, the opinion rejects the argument that courts should resolve this issue simply by allocating the burden of production or burden of proof to one party or the other.)  I commend all of this to readers’ attention,  but in the interest of conciseness I will cut to Lord Justice Arnold’s resolution of the issue, which begins at paragraph 143:

143. Analysis of the law. As is common ground, factual causation is not enough on an account of profits any more than it is on an inquiry as to damages. The reason is simple: factual causation proves too much. Applying the usual “but for” test of factual causation, the infringer may have made profits which it would not have made but for the infringement, but which are not in truth derived from the infringement as opposed to other factors. . . . Accordingly, as is also common ground, legal causation is required as well as factual causation. . . .

 

145. Lufthansa argues that a correct application of the “but for” test of factual causation involves the identification of a counterfactual world in which the infringements did not take place. That requires identification of the nearest NIA available to the defendant. The difference between the profit the infringer in fact made and the profit the defendant would have made had it adopted that NIA (i.e. the differential profit) represents the profit derived from the infringement. On this argument, a differential profit analysis identifies with precision the profits attributable to infringement as opposed to other factors. Thus it accounts for the causal potency of the different factors. Accordingly, the only role for legal causation is to police the length of the causation chain, and to exclude profits which are too remote from the infringing acts (e.g. profits made by reinvestment of the profits from the infringing acts into a cryptoasset which performs very well). . . .

Lord Justice Arnold, however, expresses agreement “with Astronics and Panasonic that the ‘but for’ test of factual causation does not necessarily require the identification of a specific NIA,” and “that the role of legal causation in this context is not limited in the manner contended for by Lufthansa” (para. 148).  In my view, the next several paragraphs are the most important portion of the opinion, so I’m going to quote them with only minimal editing for concision. 

149. As Astronics and Panasonic accept, there may be cases in which a differential profit analysis is useful for this purpose. As they submit, however, differential profit analysis is fraught with difficulty. . . .

 

150. The reason why differential profit analysis is fraught with difficulty is that, in order to identify with precision the profits attributable to the infringement, as opposed to other factors, it is necessary to identify a counterfactual in which all other factors are held constant and the minimum change is made to ensure that the patent is no longer infringed, so that the economic impact of the infringement can be isolated from the economic impact of other factors. In theory, this should present no difficulty. In the real world, the opposite is true.

 

151. The first question is who bears the burden of identifying and proving the NIA. The Canadian courts have held the burden rests on the defendant, but it is not clear to me why this should be so. One could argue that the burden should lie on the claimant, since the claimant is claiming the profits derived from the infringement, and if it relies upon differential profits analysis to quantify those profits, then the claimant must prove the NIA which should be used for that purpose. A potential difficulty with that approach is that it would enable the claimant to skew the differential profits analysis by selecting a very unprofitable NIA. As I understand the jurisprudence of the US courts on this question, they apply a shifting burden of proof under which the claimant must first identify an NIA, and then the onus is upon the defendant if it wishes to rely upon a different NIA as being a better one. That is a principled approach, but in many cases it will lead to an evidential burden on both parties.

 

152. If the defendant bears the burden of proof either in full or in part, the next question is what happens if the defendant proposes an NIA which the claimant contends is not an NIA because it also infringes the patent in suit? This is what happened in the present case. The first problem with this is that it required the judge to undertake a patent infringement trial as part of the account of profits, with all the attendant complexity and expense.

 

153. The next problem is what happens if the claimant turns out to be correct, and the proposed NIA actually infringes. Lufthansa argues that, because the Defendants failed to prove their chosen NIA, Lufthansa can claim all the profits. But all this shows is that the supposed NIA is inapposite for a differential profit analysis because it is not actually an NIA. It should not mean that the court is relieved from the burden of identifying an NIA, because differential profit analysis requires an NIA. One answer to this would be for the defendant to plead and prove a series of alternative potential NIAs, each further away from the claimed invention than the last, but that would simply compound the first problem.

 

154. The next difficulty is that a question may arise as to whether the defendant could have undertaken the NIA. Suppose that the NIA requires access to a particular raw material or part, but the defendant did not have access to that material or part at the relevant time. As I understand Lufthansa’s argument, this means that the defendant cannot rely upon the NIA, but I question why not. The availability of that material or part does not alter the inventive contribution of the patent. The object of a differential profit analysis is to identify what profits are caused by the use of invention, not what profits are caused by adventitious commercial factors. A similar problem arises if the defendant is prevented from using the NIA by regulatory factors unrelated to the invention. . . .  

 

155. Furthermore, a question may arise as to whether, even if the defendant could have undertaken the NIA, it would probably have done so. This seems to me to even less relevant, since ex hypothesi we are considering a counterfactual. A counterfactual is a thought experiment whose purpose is objectively to identify the consequences of what the defendant actually did. It does not depend on the probability of the defendant doing the alternative in the counterfactual world. . . .

 

156. The next difficulty is the one I mentioned when discussing the academic literature. In adversarial litigation courts depend on the parties to adduce evidence. What happens if the evidence does not enable the court to identify a suitable NIA? Counsel for Lufthansa argued that it is always possible to postulate an NIA, even if it is simply not producing the product in question at all. The problem with this argument is that an NIA only serves the purpose of the differential profit analysis if it enables the court to distinguish between the profits derived from the infringement from the profits derived from other factors. As explained above, this requires the identification of an NIA in which the minimum change necessary to avoid infringement is made, but all other factors are held constant. As the present case illustrates, the evidence may not permit this satisfactorily to be done.

 

157. The final problem I will mention is the one touched on by Laddie J in Celanese v BP at [43] (paragraph 102 above). In the real world, it is often the case that a complex product or process is covered (or arguably covered) by multiple patents relating to different aspects of the product or stages of the process. How does differential profits analysis work on a claim for an account of profits for infringement of just one of those patents? It cannot be correct to treat all of the profits generated by the manufacture and sale of the complex product or process as attributable to that infringement and none as attributable to the use of the other inventions. It could be argued that this depends on whether the other patents are (a) valid and (b) infringed, but that raises the spectre of determining the validity and infringement of each of those patents, without the participation of the owners of those patents, for the purposes of an account of profits. . . .

 

158. In short, while differential profit analysis has much to be said for it in terms of legal and economic theory, applying it in real world litigation is at best difficult, costly and uncertain.

 

159. The conclusions which I draw from this discussion are as follows. First, as Astronics and Panasonic accept, there are some cases in which it is possible to say that all of the profits in issue are derived from the infringement. As Laddie J explained in Celanese v BP at [47] (paragraph 104 above) and Lewison LJ noted in Abbott v Design & Display at [28] (paragraph 108 above), these are cases where, without the infringement, the infringer’s product or process would not have existed at all or where the invention was the essential ingredient in the creation of the infringer’s whole product or process. It may be possible, as discussed above, to reconcile such cases with differential profit analysis on the basis that the NIA is not manufacturing products at all, but that does not seem satisfactory. In any event, that is not how they have been analysed in the English or Australian case law. I shall return to this question in the context of ground 2.

 

160. Secondly, as Astronics and Panasonic also accept, there may be some cases in which differential profit analysis is a useful tool to identify the profits derived from the infringement, rather than from other factors. These will be cases where there is a well-defined and uncontested NIA which only changes the defendant’s product or process to the minimum extent necessary to avoid infringement and holds all other factors constant. For the reasons given in paragraphs 149-157 above, I am sceptical as to whether there are likely to be many such cases.

 

161. Thirdly, the English and Australian case law demonstrates that there is a well-established alternative to differential profit analysis, which is for the court to make a fair apportionment of the profits in issue. This approach is to be adopted when the case does not fall into either of the two categories discussed in paragraphs 159-160 above, that is to say, it is not a case where all of the profits are derived from the infringement or where a differential profit analysis can readily be undertaken. I will discuss how apportionment is to be carried out when I come to ground 4.

 I’ll stop there for now, and comment a bit on the preceding paragraphs.

First, I agree with Lord Justice Arnold that, in the real world, and particularly in cases involving complex products, it often may be unduly difficult or impossible to calculate the profit attributable to the infringement by means of the differential profits approach; and that, in recognition of these difficulties, what courts often tend to do is to apportion, as best they can, the profit attributable to the invention in comparison with the other features of the accused product.  This is a point I do make when I teach my IP remedies course, but I probably have not made it sufficiently clear in my scholarship, which has tended to emphasize why I believe that, in general, the differential profits approach is correct in theory and should be employed when it is feasible to do so.  A related point that I have tried to emphasize, however, in papers such as Patent Damages Heuristics, is that there often is a tradeoff between (theoretical) accuracy and administrability; and that sometimes the net benefits of a more easily administered rule outweigh the net benefits of (what might seem to be) a more precise, but also more costly to implement, approach.

Second, I agree with much, though not all, of Lord Justice Arnold’s analysis of the problems that can arise when trying to carry out a differential profits analysis.  It is, of course, often difficult to determine precisely what the NIA was (or to disentangle how its use might have affected other aspects of the accused product).  There is also the difficulty of determining how to proceed if a proposed NIA itself turns out to be patented, which is a matter my coauthors and I briefly noted at pages 20-22, 62 of Patent Remedies and Complex Products, but didn’t attempt to resolve. Where I might respectfully disagree with Lord Justice Arnold is in his discussion at paragraphs 154-55 of whether a differential profits analysis should take into account “adventitious commercial factors” or what the defendant would have done but-for the infringement as opposed to what it could have done.  More generally, I tend to agree with Norman Siebrasse, in his critique of the Canadian Supreme Court decision in Nova v. Dow (a case discussed in the present decision at paras. 127-32), that the correct approach when applying the differential profits analysis is to identify what the defendant would have done but-for the infringement, even if that noninfringing option consists of deploying its resources to make an entirely different product.  The goal of the disgorgement remedy should be to determine how much the defendant benefited from the infringement, and that means taking into consideration what action the defendant would have taken had it not infringed and estimating what benefits, adventitious or not, it would have derived from doing so.  For further discussion of this issue, see Siebrasse's article on Nova v. Dow, Part 1, particularly pp. 299-301 (arguing, inter alia, that "[a]n effort to determine the true value of the invention, apart from such happenstance, is akin to an effort to determine the true harm from negligent driving by awarding damages according to some ideal or average harm that would be caused by a negligent driving accident, rather than the accident that actually happened"). 

Third, and related to the preceding points, the differential profit approach can run into problems if there are two or more patents that are essential to the production of the product in question.  Imagine, for example, that two patentees each own an essential patent that is infringed by the maker of the accused product, and that each patentee files its own independent infringement action.  Each might claim that, absent the use of the patented technology in suit, the defendant would have earned zero profits; but surely it would make no sense to award each patentee the entire profit earned from sales of the infringing product.  Some sort of apportionment would therefore seem to be necessary instead.  (Perhaps the correct theoretical approach solution in such a case would be to apply some version of Shapley Pricing to isolate the inventive contribution of each essential patent, as Siebrasse and I proposed in The Value of the Standard--though we cautioned there that our analysis was not “intended to describe how we think a real world royalty setting process should work,” but rather as presenting “conceptual benchmarks for assessing a FRAND royalty” (p.1199).  I would also call interested readers’ attention to a somewhat analogous issue discussed by Jason Reinecke in his article Lost Profits Damages for Multicomponent Products:  Clarifying the Debate, 71 Stan. L. Rev. 1621 (2019), in the context of lost profits and multiple essential patents.) 

Returning to the Lufthansa decision, the next portion of Lord Justice Arnold’s opinion concludes, in brief, that the trial court was correct to apportion profits, despite some language in the lower court opinion suggesting that it was doing so despite having found that those profits were not legally (proximately) caused by the infringement (see paras. 162-76).  In fact, there was sufficient evidence that an allocable portion of the profits were legally caused by the infringing use of the patent (paras. 177-90).

That leaves for discussion Lord Justice Arnold’s analysis of the method of apportionment, the double recovery issue, and interest, as well as the two short concurring opinions by Lord Justices Nugee and Lewison.  I will return to these in a subsequent post or posts.    

Lufthansa v. Astronics: Disgorgement of Profits, Part 1

Earlier this month I noted the decision of the Court of Appeal for England and Wales in Lufthansa Technik AG v. Astronics Advanced Electronic Systems, [2026] EWCA Civ 964, and said that I would return with a more detailed, possibly two- or three-part analysis, focusing on the court’s handling of disgorgement of profits, double recovery, and prejudgment interest.  This post is the first part of that series, in which I will lay out the essential facts and set the stage for discussion of the general principles for calculating the profit to be disgorged.

The facts, in brief, are as follows.  Lufthansa owned European Patent (UK) No. 0 881 145, relating to “an apparatus that supplies electrical power to a socket that can be used by aircraft passengers for equipment such as laptops, mobile phone chargers and the like” (para. 2).  “The invention comprises a safety feature which detects whether a plug has been properly inserted before supplying power from a remote unit to the socket,” and “is useful where it is desired to supply passengers with high-voltage AC power rather than low-voltage DC power” (id.).  “The infringing acts concern the supply of EmPower Fusion systems incorporating the patented safety feature,” which “feature is implemented by an outlet unit or socket, an inseat power supply unit . . . and a cable that connects the two together (“the Primary Components”). The Primary Components can be used in conjunction with a number of additional components (“the Secondary Components”), the most important of which is the master control unit,” which “is usually located within the electronics bay of the aircraft and controls the distribution of power to a variety of downstream systems” (para. 5).  Astronics made the Primary and Secondary Components, which it sold to providers such as Panasonic, who “in turn sold integrated IFE [In-Flight Entertainment] systems incorporating the Primary and Secondary Components to airlines” (para. 6).  Astronics also sold the Primary Components directly to airlines, e.g. for use on shorter-haul flights, and in such cases companies such as Safran incorporated those components into airline seats (id.).  

Although the patent expired in 2018, litigation concerning the English, German, and French designations of EP `145 has been going on for some time.  The judgment of the Patents Court finding the defendants to have infringed issued in 2020, and in 2022 Lufthansa elected for an account of profits.  In 2024, the Patents Court held that Astronics was liable in the amount of US$4.42 million, Panasonic in the amount of US$7,384, and Safran (as subsequently revised) in the amount of $US 567,800.  In 2025, the Patents Court issued another decision awarding interest and declining “to make an order requested by the Defendants with a view to preventing double recovery by Lufthansa in proceedings in France and Germany” (para. 8).  Lufthansa appeals the profits awards against Astronics and Panasonic (it argues they should be higher).  The three defendants appeal the award of interest and the order concerning double recovery, and Astronics and Panasonic appeal an aspect of the 2024 judgment concerning whether a 1998 license Lufthansa granted to a third party (KID Systeme) was exclusive (which would affect the amount of profit to be awarded for a portion of the relevant time period).

There are additional facts set forth early on in Lord Justice Arnold’s opinion, concerning the development of an in-seat power supply unit by, and the issuance of related patents to, Olin Aerospace (later spun off as Primex Technologies, which was later acquired by General Dynamics, which later sold the unit to Astronics); the development by Lufthansa of the patented technology in suit; Lufthansa’s partnering with KID Systeme, a division of Airbus, to build its own in-seat power supply unit, as memorialized in a 1998 Teaming Agreement and a later, 2014 Teaming Agreement; and the 2003 settlement of litigation between Airbus/KID and General Dynamics concerning possible infringement of what were originally the Olin patents by Airbus/KID, and possible infringement of Airbus/KID’s “right of exclusive use” of the patent in suit by General Dynamics (paras. 12-35).  During the relevant time period, Astronics’ AC power system outsold KID’s, and by 2018 the former’s market share was 87.9% (para. 36).

On then to the principles concerning disgorgement of profits.  Early on, Lord Justice Arnold sets out the rationale for the remedy as follows:

53.  It is debatable whether the remedy of an account of profits for infringement of an intellectual property right is accurately described as a restitutionary one, although it can certainly be viewed as a remedy for unjust enrichment. For this reason, some scholars prefer to term the remedy “disgorgement”. Either way, the remedy should be distinguished from using the infringer’s profits as either evidence of, or a proxy for, the damages suffered by the right owner, as some legal systems do. It is well established that the purpose of the remedy is not to punish the infringer. On the other hand, it has been recognised by some courts that deterrence is part of the rationale, since an award of damages may in some circumstances leave the infringer better off than if it had not infringed, whereas an account of profits cannot have that effect: see, for example, Nova Chemical Corp v Dow Chemical Co [2022] SCC 43, [2022] 3 SCR 352 at [44]-[48] (Rowe J).

He then moves on to causation:

55. As [Patents Act] section 61(1)(d) makes clear, the profits for which an infringer must account are the profits “derived by him from the infringement”. Thus the profits must have been caused by the infringing acts. It is common ground that this involves both factual causation and legal causation. It is also common ground that, to this extent, the same principles are applicable to accounts of profits as to damages for patent infringement: see Imperial Oil Ltd v Lubrizol Corp (1996) 71 CPR (3d) 26 at 30 (Canadian Federal Court of Appeal), Celanese International Corp v BP Chemicals Ltd [1999] RPC 203 at [37] (Laddie J) and OOO Abbott v Design and Display Ltd [2016] EWCA Civ 98, [2016] FSR 27 at [14] (Lewison LJ). . . .

 

57.  It was common ground before the judge that the “but for” test of factual causation required consideration of a hypothetical counterfactual world in which the Defendants had adopted the nearest non-infringing alternative (“NIA”) to the arrangement claimed in the Patent. As explained in slightly more detail below, the Defendants argued that, had they not infringed, they could and would have made an alternative outlet unit referred to as the 1171M. The judge found that the Defendants had established this on the balance of probabilities, but that the 1171M would still have fallen within the claims of the Patent. Thus the Defendants failed to establish the only NIA for which they contended. Accordingly, the judge found that the profits claimed by Lufthansa had been factually caused by the Defendants’ infringing acts.

 

58. There is no challenge by the Defendants to the judge’s finding of factual causation. It should nevertheless be noted that the Defendants no longer support the proposition that factual causation in an account of profits for patent infringement requires consideration of an NIA. The Defendants consider that factual causation is a much simpler question: what profits were the consequence of the pleaded acts which were held to infringe? In the present case the infringing acts consisted of the supply of the Primary Components in the United Kingdom. Those supplies led to the profits being made which the parties’ accounting experts were able to identify as being consequent on those acts, including profits made by the supply of Secondary Components.

The next portion of the opinion sets out the trial judge’s analysis of the causation issue.  As summarized by Lord Justice Arnold, the trial judge found “that Astronics would not have made the profits in issue but for the infringement of the Patent for which Morgan J had held Astronics liable. Secondly, he found at [452]-[465] that, but for the supply of the Components to Safran and Safran connecting them together and incorporating them into airline seats, Panasonic would not have made the profits in issue. Thirdly, he found at [466] that Safran would not have made the profits in issue but for the infringement of the Patent for which Morgan J had held Safran liable” (para. 63).  However, the trial judge also found that legal (proximate) causation was lacking, because the features of the patented invention were only some of the many other features that were necessary to the functional and commercial success of the infringing system (para. 64).  The same reasoning applied to Panasonic; as for Safran, the plaintiff conceded that apportionment would be appropriate, so the court didn’t believe it to be necessary to decide legal and proximate causation with respect to that defendant (para. 65).  The trial judge nevertheless concluded that, despite the lack of proximate causation as to Astronics and Panasonic, Lufthansa was entitled to an apportioned amount of these firms’ profits.  “He noted at [574] that the Defendants had pleaded four alternative ways of apportioning their profits: (i) patent counting, (ii) the cost of the Components, (iii) the causative effects of the Patent and other factors and (iv) by reference to the 2014 Teaming Agreement [under which Lufthansa granted KID a non-exclusive license, see para. 39] (para. 71).  Ultimately, the trial court derived an implicit royalty rate from the 2014 Teaming Agreement, of 21% deducting all overheads or 13% deducting incremental overheads only.  He applied the 13% rate to Astronics’ and Panasonic’s profits, resulting in the $4.42 million and $7.384 million figures noted above (paras. 73-75).

From these conclusions, Lufthansa raises four grounds for appeal.  “Ground 1 is that, having correctly carried out a differential profits analysis, the judge was wrong then to make an apportionment of the Defendants’ total profits because those approaches are alternatives. Ground 2 is that the judge erred in his approach to legal causation. Ground 3 is that the judge wrongly applied the test for legal causation. Ground 4 is that, even if he was right to apportion the Defendants’ profits, the judge erred in his approach to apportionment” (para. 77).

Long story short:  Lord Justice Arnold concludes that it was appropriate to carry out an apportionment analysis on the facts of the present case; that the judge’s approach to legal causation was erroneous, but that he was correct to apportion nonetheless; and that the method used by the judge for apportionment was correct, though with a correction to be made with regard to the effect of the original (1998) Teaming Agreement.  Why?  That will be the subject of my next installment. 

Federal Circuit Affirms Marking Requirement for NPEs Who License Their Patents

While I was away last week, the Federal Circuit handed down its decision in VDPP, LLC v. Volkswagen Group of America, Inc., precedential opinion by Chief Judge Moore joined by Judges Lourie and Cunningham.  The patent at issue had expired by the time the patentee sued Volkswagen, and so the patentee could only seek damages for infringement occurring (a) prior to patent expiration and (b) within the six-year statute of limitations.  The patentee nevertheless failed to plead with sufficient particularity, much less prove, that it had made reasonable efforts to ensure that other entities to whom it had licensed the patent in suit pursuant to settlement agreements complied with the patent marking requirement set forth in 35 U.S.C. § 287(a).  The Federal Circuit affirms that that failure is fatal to the patentee’s ability to recover damages, and since that was the only relief to which it might have been entitled, it failed to state a claim upon which relief may be granted.  In addition, the court affirms an award of attorneys’ fees, for reasons discussed below, and dismisses an appeal of an award of sanctions against the patentee’s attorney.   

Readers who are not familiar with the peculiar U.S. patent marking requirement may benefit from a brief primer on what this all about, so here is a little background.  First, § 287(a) establishes a general rule that patent owners cannot recover damages for any infringement of their patents occurring before the date on which they marked their products with the patent number, or provided actual notice to the accused infringer that their products are patented.  Second, as the Federal Circuit reaffirms in VDPP, the marking/actual notice requirement is intended to “(1) help[] . . . avoid innocent infringement; (2) encourag[e] patentees to give public notice that the article is patented; and (3) aid[] the public [in] identify[ing] whether an article is patented” (opinion at 6, citation omitted).  Third, according to the case law, if a patentee licenses its patents to another entity, it must make reasonable efforts to ensure that the licensee complies with the marking requirement, if it wishes to recover damages for infringing activity that pre-dates the date on which it puts the accused infringer on actual notice.   (Actual notice can be accomplished by means of a cease-and-desist letter or, at the latest, by the service of the complaint.)  Fourth, however, the marking requirement is riddled with so many exceptions and formalisms that, in my opinion, we would be better off without it.  As I have written previously, the three policy justifications recited above

may sound reasonable, but in practice marking does not necessarily further [them], and the manner in which the statute has been interpreted produces some strange results. In particular, absent compliance with the marking requirement damages will not begin to accrue until the owner provides the infringer with actual notice, even if the infringer has acquired actual knowledge of the patent prior to that date. That seems bad enough, but now consider what happens if the owner has failed to mark but the defendant, with actual knowledge of the owner’s patent, knowingly induces a third party to infringe that patent. Under the law of induced infringement, the defendant is liable from the moment the third party takes the bait and begins infringing, precisely because the defendant had actual knowledge of the patent; and yet damages cannot accrue until the date on which the owner provides actual notice, even though common sense would suggest that such notice is superfluous. Similarly, a defendant may be adjudicated a willful infringer, and thus be on the hook for enhanced damages, based on evidence that it knowingly infringed prior to the owner’s having put the defendant on actual notice; but if the owner failed to mark its products, no damages (enhanced or otherwise) will accrue until the date of actual notice. On the other hand, if the owner successfully asserts a process patent instead of a product patent, damages accrue from the moment of infringement regardless of whether the owner provides actual notice or the defendant has actual knowledge, because there is nothing to mark. Nevertheless, if the owner asserts both a product and a process claim of a patent containing both, and sells unmarked goods covered by the product claim and made in accordance with the process claim, a defendant who infringes both product and process claims is probably not liable for damages until the owner provides actual notice. The outcome is the same if the owner asserts both claims but only prevails on the process claim. If the owner asserts only the process claim, however, it can recover damages from the date the infringement began; similarly, if the owner has two separate patents, one reciting product and the other process claims, it can recover damages for the infringement of the process patent from the date that infringement began, and it can recover damages for the damages for infringement of the product patent from the date on which it puts the infringer on actual notice.

Fifty Years of Patent Remedies Case Law: Two Steps Forward, One Step Back, 50 AIPLA Q.J. 607, 621-25 (2022) (citations omitted).  Moreover, the owner of an idle patent--that is, a patentee who doesn't manufacture, sell, or license any products covered by the patent--can recover damages from the date the infringement starts, as long as it's within the statute of limitations, regardless of whether it puts the accused infringer on actual notice prior to filing suit (see opinion, p.5).  And as my former student Bernard Cryan noted in his paper Not All Licensees Are the Same: 35 U.S.C. § 287 Should Not Require Marking by Licensees that Deny Infringement, 101 J. Pat. & Trademark Off. Soc’y 531 (2021), cases involving licenses entered into pursuant to settlements are particularly fraught, because the licensee may agree to settle simply to avoid the cost of litigation, and may believe in good faith that its products do not infringe.  To be sure, in the present case the Federal Circuit states that “the subjective view of the accused infringer” does not matter (p.6); but 35 U.S.C. § 292(a) states that “[w]hoever marks upon, or affixes to, or uses in advertising in connection with any unpatented article, the word 'patent' or any word or number importing that the same is patented, for the purpose of deceiving the public,” is potentially liable for a $500 fine (or in some cases damages) for false patent marking.  Should the patentee nevertheless require the settling licensee to mark its products, even if the latter doesn’t believe that the claims of the patent in suit actually read on its products?  

My opinion is that, in practice, these rules are so arbitrary that we would ought to abolish the marking requirement altogether; or, alternatively, we could mimic practice in the U.K. and condition awards of damages on actual or constructive knowledge, such that infringers with actual knowledge would not be off the hook even if the patentee has not provided actual or constructive notice, while infringers of process patents who have no pre-suit knowledge or notice of those patents would not be liable for damages accruing prior to the filing of the complaint.  Either option would be a rational, defensible rule, but in my view our current U.S. practice is not.  (For further discussion from a comparative law perspective, see my book Remedies in Intellectual Property Law (Edward Elgar Publishing 2026), pp. 129-36.)

Coming back to the present case, as mentioned above the court also affirms an award of fees, though this is only partly related to the patent marking issue:

The district court did not abuse its discretion in determining this was an exceptional case. Specifically, the court reasonably determined that “[m]any of the positions VDPP took were frivolous and objectively unreasonable,” including seeking future damages and an injunction on an expired patent, seeking past damages despite an inability to allege patent marking, failing to disclose relevant settlement agreements, and prolonging litigation with false statements about the settlement agreements. . . . The court also noted that “VDPP made sloppy errors, over and over” ranging from obviously incorrect venue allegations to initial disclosures about an unrelated patent, which made the case “even more cumbersome and difficult to handle.” . . . Moreover, the court found a “need for meaningful deterrence” based in part on VDPP’s pattern of repeat litigation over the ’452 patent involving “settlement demands far less than the costs of defense and unrelated to any damages theory.” . . . The court further rejected VDPP’s argument that the fee award should be limited to the exceptional portion of the case because “VDPP’s misconduct infected the entire litigation” (pp. 7-8, citations omitted).

The court further notes that a case may be “exceptional” even if the losing party’s conduct is not sanctionable under Rule 11 of the Federal Rules of Civil Procedure (pp. 8-9).  Further, while “‘filing a large number of suits does not, by itself, justify an inference of . . . an improper motive,’ Thermolife Int’l LLC v. GNC Corp., 922 F.3d 1347, 1363 (Fed. Cir. 2019), and ‘[t]he mere existence of these other suits does not mandate negative inferences about the merits or purpose of this suit,” SFA Sys., LLC v. Newegg Inc., 793 F.3d 1344, 1351 (Fed. Cir. 2015). But we have also recognized that ‘a pattern of litigation abuses characterized by the repeated filing of patent infringement actions for the sole purpose of forcing settlements, with no intention of testing the merits of one’s claims, is relevant to a district court’s exceptional case determination under [35 U.S.C.] § 285,’ and ‘a district court should consider a patentee’s pattern of litigation where adequate evidence of an abusive pattern is presented.’  SFA, 793 F.3d at 1350, 1352 (emphases added)” (pp. 8-9).

Finally, the court dismisses the portion of the appeal relating to attorney sanctions, on the basis of its conclusion that the attorney “failed to timely appeal the district court’s sanctions order on his own behalf and that VDPP lacks standing to contest the sanctions for him” (p.13).

Update:  In the interest of completeness, I should also note some further issues concerning marking/notice/knowledge framework arising under Patent Act § 154(d) (relating to royalties for the unauthorized making, use, etc. of inventions claimed in pending applications) and § 287(b) (relating to the infringement of process patents by means of the unauthorized importation, sale, offer to sell, or use of products made by such processes, in violation of § 271(g)).  I'm not going to go into the details here, though.