Guest Post by Prof. Jorge Contreras of the University of Utah S.J. Quinney School of Law.  Disclosure statement: in 2019, the author served as an expert for HTC in an unrelated, non-U.S. case.

In August 31, 2021, the Court of Appeals for the Fifth Circuit ruled in HTC Corp. v. Telefonaktiebolaget LM Ericsson, 2021 U.S. App. LEXIS 26250, __ F.4^th __ (Fed. Cir. 2021), affirming the judgment of the District Court for the Eastern District of Texas, 2019 U.S. Dist. LEXIS 170087 (E.D. Tex. 2019).  The decision is significant as it is the first by the Fifth Circuit to address the licensing of standards-essential patents and the meaning of “fair, reasonable and nondiscriminatory” (FRAND) licensing terms, adding to the growing body of jurisprudence already issued by the Third, Ninth and Federal Circuits in this area.  It is also significant because the court addresses several issues that have become increasingly important in standards-related litigation including (1) the apportionment of value among components of a multi-component product, (2) the proper choice of law for FRAND disputes, and (3) the interpretation of the “nondiscrimination” prong of the FRAND commitment.  These issues all arose in connection with HTC’s challenge to District Judge Rodney Gilstrap’s jury instructions regarding FRAND. His vague charge appears to reflect the general uncertainty in this area, not only of the Texas district court, but of the entire judicial system. One can almost hear the weariness permeating the final sentence of Judge Gilstrap’s charge to an admittedly perplexed jury: “Ladies and gentlemen, there is no fixed or required methodology for setting or calculating the terms of a FRAND license rate.”

Background

Though it was never a household name in the U.S., HTC — founded in Taiwan in 1997 — was an early smartphone pioneer.  In 2005 HTC released the world’s first Windows 3G smartphone (the clamshell HTC Universal) and followed in 2008 with the first smartphone running Google’s Android operating system (branded as the T-Mobile G1).  Most significantly, HTC was the developer and manufacturer of Google’s Nexus One Android phone, which was released in 2010.

The 2/3/4/5G wireless communication standards published by the European Telecommunications Standards Institute (ETSI) are incorporated into almost all smartphones (and many other devices) that have been sold over the past two decades.  These standards are covered by tens of thousands of patents around the world (standards-essential patents or SEPs), a respectable number of which are held by Swedish equipment manufacturer Ericsson.  Most standards-development organizations (SDOs) today recognize the potential leverage that the holders of SEPs may hold over manufacturers of standardized products. FRAND licensing commitments are designed to alleviate the risk that SEP holders will prevent broad adoption of a standard by asserting their patents against manufacturers of standardized products. As explained by the Fifth Circuit, “To combat the potential for anticompetitive behavior, standard setting organizations require standard-essential patent holders to commit to licensing their patents on … FRAND terms” (slip op. p. 4). Thus, under ETSI’s intellectual property policy, which dates back to 1993, Ericsson and other holders of patents that cover ETSI’s standards agree to grant manufacturers licenses on FRAND terms.

Ericsson and HTC entered into three such licensing agreements in 2003, 2008 and 2014. Under the 2014 agreement, HTC paid Ericsson a lump sum of $75 million for a 2-year license to use Ericsson’s 2/3/4G SEPs. In 2016, HTC and Ericsson began negotiations to renew the license. Ericsson proposed a rate of $2.50 per 4G device, which was based on the lump sum paid by HTC in 2014 divided by the number of phones sold by HTC over the license period. HTC did not accept this offer and instead conducted an assessment of the value of Ericsson’s SEPs. It made a counteroffer of $0.10 per device in March 2017.  Negotiations stalled, and in April, HTC brought an action in the Eastern District of Texas seeking a declaration that Ericsson had breached its obligation to offer HTC a license on FRAND terms.

The trial focused largely on the proper method for determining a FRAND royalty for Ericsson’s SEPs.  Because the actual FRAND royalty or royalty range in a given case is generally viewed as a question of fact, the case was tried a jury, and each party submitted a set of draft jury instructions to the court. As summarized by the Fifth Circuit, “HTC proposed highly detailed instructions to help the jury interpret FRAND, but Ericsson objected to most of these instructions and proposed a more general FRAND instruction. The district court considered the two proposals, but ultimately gave the following instruction to the jury: ‘Whether or not a license is FRAND will depend upon the totality of the particular facts and circumstances existing during the negotiations and leading up to the license. Ladies and gentlemen, there is no fixed or required methodology for setting or calculating the terms of a FRAND license rate.’ (slip op. at 6). The jury returned a verdict “finding that HTC had not proven that Ericsson had breached its FRAND duties and that both parties had breached their obligations to negotiate in good faith.” (slip op. at 9]. On the basis of the jury verdict, Ericsson moved for declaratory judgment that, in its dealings with HTC, it complied with its FRAND obligations.  The court granted Ericsson’s motion.

HTC appealed the district court’s ruling on three grounds:  it failed to adopt three of HTC’s proposed jury instructions, it incorrectly concluded that Ericsson’s licensing offer complied with its FRAND commitment, and it improperly excluded expert testimony as hearsay.  The Fifth Circuit, in an opinion authored by Judge Jennifer Walker Elrod, affirmed the lower court’s ruling on all three grounds.  Judge Stephen A. Higginson entered a separate concurring opinion, arguing that the district court erred by omitting HTC’s apportionment jury instruction, but this omission did not rise to the level of reversible error.

Apportionment

In patent infringement cases, it is well-established that a patentee’s damages should reflect only the value of the patented features of an infringing product. Thus, in assessing damages, courts routinely “apportion” the infringer’s profits between the infringing and noninfringing features of its product. See Garretson v. Clark, 111 U.S. 120, 121 (1884).  Accordingly, HTC argued that the district court erred by failing to give the jury specific instructions on apportionment.

However, this case did not sound in patent infringement, but in breach of contract.  As my co-authors and I have previously observed (see p. 162), it is a peculiar coincidence of U.S. law that both the statutory measure for patent damages under 35 USC § 284 and the FRAND commitment call for the imposition of a “reasonable” royalty.  For this reason, several U.S. courts that have calculated FRAND royalty rates (see, e.g., Ericsson v. D-Link (Fed. Cir. 2014), Microsoft v. Motorola (9th Cir. 2015); and Commonwealth Sci. & Indus. Research Org. (CSIRO) v. Cisco (Fed. Cir. 2015)) have looked to traditional methodologies for determining reasonable royalty patent damages, including the 15-factor Georgia-Pacific framework.  But while this methodology may be useful by analogy, it is not strictly required, as the damages flowing from a breach of the contractual FRAND commitment must, by their nature, be determined under applicable principles of contract law.  As the Fifth Circuit explains, “while the Federal Circuit’s patent law methodology can serve as guidance in contract cases on questions of patent valuation … it does not explicitly govern the interpretation of contractual terms, even terms that are intertwined with patent law” (slip op. at *15).  As such, the Fifth Circuit in this case held that the district court’s omission of a specific jury instruction on apportionment – a patent law doctrine – was not error.

Judge Higginson disagreed.  Citing D-Link and CSIRO, he observed the Federal Circuit’s “unmistakable command that a jury assessing patent value must be instructed on apportionment” and that “the failure to instruct a jury on proper apportionment is error when the jury is asked to assess patent value” (slip op. at *29-30, Higginson, J., concurring). And while he acknowledges that the instant case concerns breach of contract rather than patent infringement, he argues that the Federal Circuit’s precedent regarding apportionment is instructive “because a jury assessing patent infringement damages undertakes the same task of assessing whether an offered rate is FRAND” (citing Realtek Semiconductor, Corp. v. LSI Corp. (N.D. Cal., 2014) (slip op. at *30).  Moreover, Judge Higginson points out that both HTC and Ericsson, as well as the United States as amicus curiae, requested jury instructions on apportionment.  For all of these reasons, he concludes that the district court erred by omitting an instruction on apportionment.

Whether or not the district court erred, none of the Fifth Circuit judges felt that the omission of a specific apportionment instruction constituted reversible error because HTC had the opportunity to, and did, discuss apportionment during its closing argument before the jury.  This observation raises interesting questions regarding the basis on which juries are expected to make decisions in complex cases.  During trials that often last for weeks, the members of the jury hear hours upon hours of conflicting expert testimony and argumentation.  Under Rule 51 of the Federal Rules of Civil Procedure, a trial judge must instruct the jury regarding the substantive law governing the verdict. Its instructions are intended to distill the legal standards that the jury must apply during its deliberations.  Given that members of the jury are not permitted to take notes or make recordings during trial, they must rely heavily on these court-sanctioned instructions.  And contrary to the Fifth Circuit’s holding, it does seem problematic for a court to omit instructions that are germane to the jury’s deliberations concerning an area of law as complex as FRAND royalty determinations.  In such cases, irrespective of what courts and commentators believe the law to be, the results of cases on the ground depend heavily on the instructions given to the jury (we discuss the importance of jury instructions, especially model jury instructions, in FRAND cases in Jorge L. Contreras & Michael A. Eixenberger, Model Jury Instructions for Reasonable Royalty Patent Damages, 57 Jurimetrics J. 1 (2016) (“if … increasingly complex damages calculations are to remain in the hands of the jury, it is now more essential than ever that the instructions given to jurors be as clear, accurate, and understandable as possible.”)

 The French Connection

In considering HTC’s proposed jury instruction on apportionment, Judge Elrod also concluded that the proposed instruction was inaccurate because it summarized U.S. patent law with no reference to French contract law.  ETSI was formed in France in 1988 and approved its first patent policy in 1993.  That policy, and all subsequent ETSI policies, state that they are governed by the laws of France.  Judge Elrod observed that “HTC’s proposed jury instructions are based on United States patent law. HTC did not even attempt to justify its proposed instructions under French contract law or to argue that French contract law and United States patent law are equivalent” (slip op. *14).

This last point is particularly interesting.  In chastising HTC for failing to argue the equivalency of U.S. and French law, Judge Elrod refers to Apple v. Motorola, 886 F. Supp. 2d 1061, 1081 (W.D. Wis. 2012), in which a U.S. district court seemingly concluded, without any specific references, that there are no material differences between the laws of France and Wisconsin when it comes to interpreting a company’s FRAND commitment (discussed here at p.8).  In my experience that conclusion is widely ridiculed by European lawyers, and rightly so.  Yet Judge Elrod seems to imply that merely making a conclusory statement about the equivalency of French and U.S. law ought to suffice, or at least overcome one hurdle to the acceptance of a jury instruction grounded in principles of U.S. law.

Nondiscrimination

HTC also argued that the district court should have instructed the jury with respect to the nondiscrimination prong of Ericsson’s FRAND commitment.  HTC’s proposed jury instruction reads as follows (Appellant’s Opening Brief, p. 40):

The non-discrimination requirement of FRAND requires an SEP holder to provide similar licensing terms to licensees that are similarly situated. The financial terms do not have to be precisely identical, because the difference might be explained by other offsetting adjustments in other terms in the license. But, at a minimum, if the difference in terms creates a competitive disadvantage for a prospective licensee, then the offered royalty terms are discriminatory. Discrimination may exist even if preferential treatment is accorded to only one or a few companies.

The non-discrimination prong of FRAND serves to level the playing field among competitors, and to foster entry and innovation from new market participants, by prohibiting preferential treatment that imposes different costs to different competitors. Thus, for purposes of the non-discrimination prong of FRAND, licensees are “similarly situated” if they compete for the purchase or sale of a product or service. It would defeat the purpose of FRAND if a licensor could draw a distinction between entrenched and emerging firms.

To my eye, this is an accurate statement of the law and the general understanding of the nondiscrimination requirement under FRAND (see Jorge L. Contreras & Anne Layne-Farrar, Non-Discrimination and FRAND Commitments in Cambridge Handbook of Technical Standardization Law: Competition, Antitrust, and Patents, Ch. 12 (Jorge L. Contreras, ed., 2017)).  This formulation is also consistent with the most extensive analysis to-date of FRAND nondiscrimination by a U.S. district court (TCL v. Ericsson, 2017 U.S. Dist. LEXIS 214003 (C.D. Cal. 2017), rev’d on other grounds, 943 F.3d 1360 (Fed. Cir. 2019)).

Yet Judge Elrod writes that “HTC’s proposed instruction would transform the non-discrimination element of FRAND into a most-favored-licensee approach, which would require Ericsson to provide identical licensing terms to all prospective licensees” (slip op. *16). This statement is plainly incorrect.  The first sentence of HTC’s proposed jury instruction, which mirrors the model instruction published by the Federal Circuit Bar Association in 2020 and cited by the court (slip op. *17 n.3), refers to “similar licensing terms [for] licensees that are similarly situated”, not identical licensing terms for all prospective licensees.  HTC further clarifies that “The financial terms do not have to be precisely identical.” In explaining the meaning of “similarly situated”, HTC follows the reasoning of the district court in TCL, stating that a distinction should not be drawn between “entrenched and emerging firms”.  None of this suggests that all licensees should pay identical royalties.

Moreover, the parties clearly disputed at trial whether various “comparable” licensees identified by Ericsson were “similarly situated” to HTC (“HTC further argued that Ericsson’s licenses with companies like Apple, Samsung, and Huawei were much more favorable, but Ericsson presented additional evidence indicating that those companies were not similarly situated to HTC due to a variety of factors” (slip op. *22)). Thus, both parties appear to acknowledge the relevance of the “similarly situated” test for nondiscrimination, making it all the more puzzling why the district court and the Fifth Circuit found HTC’s proposed jury instruction on this point to be inaccurate.

Conclusion

Regrettably, the Fifth Circuit’s decision in HTC v. Ericsson does little to clarify the scope or nature of FRAND commitments for standards-essential patents.  Rather, by diverging from the guidance of the Federal Circuit in terms of the apportionment of value among patented and unpatented products, and misinterpreting the scope of the nondiscrimination prong of the FRAND commitment, the Fifth Circuit has substantially muddied the already turgid waters of this increasingly important area of law.  Unfortunately, Judge Gilstrap was right when he told the jury in Texas, “Ladies and gentlemen, there is no fixed or required methodology for setting or calculating the terms of a FRAND license rate.”  It might be helpful to the industry, however, if there were.

Tiffany Cunningham: Nominee for the United States Court of Appeals for the Federal Circuit

Tiffany P. Cunningham has been a partner at Perkins Coie LLP in Chicago, Illinois since 2014. She is a member of the Patent Litigation practice and serves on the 17-member Executive Committee of the firm. Ms. Cunningham serves as trial and appellate counsel for large multinational companies, as well as small enterprises, and individuals in complex patent and trade secret disputes.

Ms. Cunningham is a registered patent attorney before the U.S. Patent and Trademark Office. From 2002 to 2014, she worked in the Chicago office of Kirkland & Ellis LLP as an associate until she was elevated to partner in 2007. Ms. Cunningham began her legal career as a law clerk to Judge Timothy B. Dyk on the U.S. Court of Appeals for the Federal Circuit from 2001 to 2002. Ms. Cunningham received her J.D. from Harvard Law School in 2001 and her S.B. from the Massachusetts Institute of Technology in 1998.

The following short statement was written by Prof. Bernard Chao (Denver) and then joined by 20+ additional professors whose names are listed below.  Chao was a patent attorney and patent litigator for 20 years before becoming a professor and I have long valued his insight. You’ll note that I also signed below – DC. 

= = = [PDF Version] = = =

Judge Alan Albright’s court in the Western District of Texas is rapidly becoming the latest hot spot for patent litigation. While only a total of two patent cases were filed in 2016 and 2017 in Judge Albright’s Waco federal courthouse, the number of patent filings rose to seven hundred and ninety-three (793) in 2020.  See J. Jonas Anderson & Paul R. Gugliuzza, Federal Judge Seeks Patent Cases, (forthcoming Duke L. J 2021). Professors Anderson and Gugliuzza provide a thorough explanation (and critique) of this sudden ascent. I have a smaller, but nonetheless important, point to make. If the Western District of Texas is going to hear some of patent law’s most important cases, it should not do so in secret. Unfortunately, that appears to have just what happened in one the highest dollar value patent trials in recent history.

The patent world has been abuzz about the $2.18 billion verdict that the Waco jury handed down on March 2, 2021 in VLSI Technology v. Intel. The public debate in patent law has often focused on whether courts and juries are getting patent damages right. Looking at relevant filings on damages provides critical information for this important discussion. In high stakes cases, parties typically file summary judgment motions on damages and Daubert motions attempting to exclude certain theories. These motions often attach expert reports and deposition testimony as exhibits. Together these documents illustrate how patent doctrine shapes damage awards. For example, filings often explain how the parties seek to apportion damages between the value of the infringing features and the product as a whole. This is not an easy task and parties have taken many approaches to apportionment with varying levels of success.

However, these documents cannot be retrieved from the VLSI Technology v. Intel docket. To be clear, sealing is appropriate in some instances. Companies should be able to keep their confidential technical and financial information under wraps. But at least as of March 15, 2021, the following docket entries were wholly unavailable (i.e. not even redacted copies were available).

Two short orders granted motions to seal these filings. However, the sealed filings appear to go far beyond damages. The first order (dated October 8, 2020) granted roughly thirty motions (Docket Entries 214-244) and the second order (dated February 18, 2021) granted even more (Docket Entries 287-89, 293-319, 321-345, 374-381, 398, 404, 405, 410, 416. 418, 420, 424, 425). Many of the sealed motions are motions in limine that do not have a descriptive title.  They are simply numbered (e.g. Motion In Limine #3).  So, there is no way to even know what subjects they cover.

Judge Albright did give a small nod to transparency in his February 18, 2021 order which required “[t]he filing party shall file a publicly available, redacted version of any motion or pleading filed under seal within seven days.” Unfortunately, the redacted versions do not appear to have been filed.  A few days earlier, Judge Albright also issued a February 12, 2021, standing order that requires parties to do the same in all cases pending in his court.

But these orders are inadequate in several ways. First, they explicitly permit parties to file purportedly confidential information under seal without a motion. But the parties have no incentive to be transparent. It is the court’s job to protect the public interest, and this order abdicates that duty.  Second, there also appears to be no safeguard for parties that redact too much information in their filings, a problem we have seen before. See In re Violation of Rule 28(D) (Fed Cir. 2011)(sanctions for redacting information that was not confidential including case citations).  Third, the order does not require parties to redact exhibits. But some of the most critical information like deposition testimony and expert reports are exhibits. Finally, the order is not retroactive allowing most of the VLSI v. Intel case to remain in the dark.

This is not the first time that patent litigation has suffered from transparency problems. Other district courts have allowed too many documents to be sealed in previous high stakes cases like Broadcom v. Qualcomm (S.D. CA) and Monsanto v. Dupont (E.D. MO).  See, Bernard Chao, Not so Confidential: A Call for Restraint in Sealing Court Records, 2011 Patently-O Patent L.J. 6 & Bernard Chao & Derigan Silver, A Case Study in Patent Litigation Transparency, 2014 Journal of Dispute Resolution 83.

While these stories appear to paint a bleak picture for transparency in patent litigation, some courts are making an effort to keep their dockets accessible to the public. There have been notable decisions to ensure that filings are accessible. For example, in 2016 the Electronic Fronter Foundation successfully intervened and obtained an order from the E.D. of Texas to unseal records in a patent case, Blue Spike LLC, v Audible Magic Corp.  Earlier this month, the Federal Circuit, affirmed a lower decision rejecting attempts to seal specific filings in another patent dispute. See DePuy Synthes Products Inc. v. Veterinary Orthopedic Implants (Fed. Cir. Mar. 12 2021).

Moreover, individual court rules are now requiring greater transparency. The Federal Circuit’s latest rules clearly seek to maximize public disclosure. In each filing, Rule 25.1(d) only allows parties to mark “up to fifteen (15) unique words (including numbers)” as confidential.  A party seeking to exceed that limit must file a motion. As Silicon Valley’s home venue, the Northern District of California entertains numerous patent cases.  The court’s local rules say that material may only be sealed when a request “establishes that the the document, or portions thereof,  are privileged, protectable as a trade secret or otherwise entitled to protection under the law.”  Moreover, the request must “must be narrowly tailored to seek sealing only of sealable material.”  Ironically, even Texas state courts, which obviously do not hear patent cases, treat requests to seal far more seriously than the W.D. of Texas.  Specifically, Texas Rule of Civil Procedure 76a states that court records are “presumed to be open to the general public”, and only allows records to be sealed upon a showing that “a specific, serious and substantial interest which clearly outweighs” various interests in openness.

In short, the Western District of Texas should join these other courts and take its duty to ensure transparency seriously. As the Fifth Circuit recently put it , “[w]hen it comes to protecting the right of access, the judge is the public interest’s principal champion.” Binh Hoa Le v. Exeter Finance Corp. (5^th Cir. Mar. 5, 2021). Accordingly, we make a few basic recommendations.  First, parties should not be able to file material under seal without judicial scrutiny. They should be required to file motions and justify their requests. Of course, a court also cannot rubber stamp these requests. If resources are a problem, the court can appoint a special master in larger cases. Second, parties should have to redact exhibits too. There can be valuable non-confidential information in those exhibits. Finally, a quick aside, even though judges bear the primary responsibility for making their dockets transparent, that does not mean the parties should not show more restraint. To the extent that either side is a repeat player, and thinks the patent system needs reform, they should be careful not to over seal. Their filings might end up being important contributions to the public debate.

Signatories (more…)

Australian Therapeutic Supplies Pty. v. Naked TM (Fed. Cir. 2020) (en banc)

NAKED TM holds the registration for the mark NAKED that it uses to sell its luxury condoms.  However, by the time NAKED TM started its business, Australian Therapy was already selling its NAKED condoms to US customers over the internet.  In the early 2000s, the companies reached some form of a tacit agreement — although without an express contract.  Since 2003, Australian has continued to sell NAKED condoms in the US, but the TTAB found that it was never more than 48 consumers per year.

The appeal here stems from Australian Therapy’s cancellation proceeding that ended when the TTAB found the petitioner lacked standing under 15 U.S.C. § 1064 (petition to cancel may be filed by “any person who believes that he is or will be damaged … by the registration of a mark …”).  The TTAB particularly found a binding contract between the parties where Australian Therapy gave-up its rights to use the mark in the US: 

The evidence shows that the parties reached an agreement. The mutuality of intent to contract is satisfied because the parties recognized their trademark issue and they communicated and exchanged offers to resolve it. The consideration for the contract is Petitioner’s agreement not to use or register the NAKED trademark for condoms in the United States and Respondent’s agreement not to use or register the NUDE trademark for condoms.

The Board explained also that Australian must show a “proprietary rights in its unregistered mark” in order to have standing under the statute.

On appeal, the Federal Circuit reversed — holding a cancellation petition may be filed by any party who demonstrates “a real interest in the cancellation proceeding and a reasonable belief of damage regardless.” It does not require evidence of a “proprietary interest in an asserted unregistered mark.”  Here, the court found a real-interest based on the fact that Australian had filed to register its marks, and had demonstrated a belief of damage because the USPTO refused registration of both the ’237 and ’589 applications based on a likelihood of confusion with Naked’s registered mark, U.S. Registration No. 3,325,577.”

At that point, NAKED TM petitioned for en banc rehearing on two questions:

I. Whether a petitioner who has forfeited its rights in a mark pursuant to a valid settlement agreement has a real interest in the proceedings and a reasonable belief of harm for the registration of said mark so as to support a finding of statutory standing under 15 U.S.C. § 1064 to challenge the registration of the mark; and

II. Whether a petitioner in the TTAB is required to establish the pleaded basis for its standing at the commencement of an action, or whether this Court may rely on an argument that has been waived by a party in finding a basis for standing in the TTAB.

The Federal Circuit has now denied the rehearing, although the case includes an interesting dissent from Judge Wallach. Wallach argues that the decision in Australian Therapy:

1. Conflicts with our case law requiring a “legitimate commercial interest” to have a valid cause of action under 15 U.S.C. § 1064, see Empresa Cubana Del Tabaco v. Gen. Cigar Co., 753 F.3d 1270, 1274 (Fed. Cir. 2014) (following Lexmark Int’l, Inc. v. Static Control Components, Inc., 572 U.S. 118 (2014)., noting that a petitioner must have a “legitimate commercial interest sufficient to confer standing”);
2. Undermines our case law favoring the enforcement of settlement agreements, see Wells Cargo, Inc. v. Wells Cargo, Inc., 606 F.2d 961, 965 (C.C.P.A. 1979) (“If there [is] a policy favoring challenges to trademark validity, it too has been viewed as outweighed by the policy favoring settlements.”); and
3. Raises questions as to the impact of Supreme Court precedent on our statutory standing jurisprudence, see Lexmark, 572 U.S. at 128 n.4 (noting that statutory standing does not implicate Article III subject matter jurisdiction), 134 (providing “a direct application of the zone-of-interests test and the proximate-cause requirement [to] suppl[y] the relevant limits on who may sue”).

AusTherapy v. Naked.

Guest post by Associate Professor Yang Yu of Shanghai University of International Business and Economics and Professor Jorge L. Contreras.

As one of us has recently observed (here and here), courts in the U.S. and UK adjudicating cases involving standards-essential patents and FRAND licensing commitments[1] have increasingly issued injunctions preventing one party from pursuing parallel litigation in another jurisdiction (anti-suit injunctions or ASIs).  In response, courts in other countries – namely Germany and France – have begun to issue anti-anti-suit injunctions (AASIs) to prevent parties from seeking ASIs that would hinder their own proceedings.  And even anti-anti-anti suit injunctions (AAASIs) have emerged as more than theoretical possibilities.  Amidst this international jockeying for jurisdiction, commentators have wondered where it will end, and whether every country involved in complex, multi-jurisdictional disputes will get into the ASI game.  In two recent cases, China has now stepped forward as another major jurisdiction in which ASIs are available to litigants.

Conversant v. Huawei Link to translation

The first case involves Chinese smartphone manufacturer Huawei and Texas-based patent assertion entity Conversant Wireless Licensing (formerly Core Wireless). Conversant holds a globe-spanning portfolio of patents that it acquired from Nokia and which it claims are essential to the 2G, 3G and 4G wireless telecommunications standards developed under the auspices of the European Telecommunications Standards Institute (ETSI).  Conversant sought to license these patents to Huawei, but negotiations broke down and in 2017 and 2018 Conversant asserted the patents against Huawei in multiple jurisdictions.  The UK suit was recently decided by the UK Supreme Court, [2020] UKSC 37, which held that a UK court has the authority to set a global FRAND royalty rate between the parties, notwithstanding its lack of jurisdiction over patents outside the UK (see discussion here).  (This case also involved an unsuccessful 2018 bid by Conversant to obtain an ASI from the UK Court of Chancery, [2018] EWHC (Ch) 2549, against the prosecution of a related suit by ZTE, Huawei’s co-defendant, in the Shenzhen Intermediate People’s Court in China – discussed here).

In response to Conversant’s initial patent assertions, in January 2018 Huawei sought declarations from the Nanjing Intermediate People’s Court of Jiangsu Province, China, that it did not infringe three of Conversant’s Chinese patents and, if it did, that it was entitled to a license on FRAND terms. On September 16, 2019, the Nanjing court declined to issue a declaration regarding infringement, but established a “top down” FRAND royalty for the three Conversant patents (discussed at length in IAM, 23 Sept. 2019 and IAM, 5 Aug. 2020). The decision is currently on appeal at the Supreme People’s Court (SPC) of China.

Meanwhile, in April 2018, Conversant sued Huawei in the District Court in Düsseldorf, Germany, alleging infringement of several European patents. On August 27, 2020, the Düsseldorf court granted Conversant an injunction against Huawei’s sale, use or importation in Germany of devices infringing patent EP1797659 (the German Patent).

On the same day, Huawei applied to the SPC seeking an “act preservation” ruling (a judicial action roughly equivalent to an ASI) to prevent Conversant from enforcing the Düsseldorf injunction until the conclusion of the Chinese proceedings.  The SPC granted the ASI after an ex parte hearing in which Conversant did not participate.  In its decision, the SPC cited a number of factors weighing in favor of granting the ASI, many of which resemble the factors considered by U.S. courts when considering ASIs (see Contreras, The New Extraterritoriality, pp. 265-67):

• the parties to the Chinese and German actions are the same;
• there is partial overlap between the subject matter of the cases;
• Conversant’s enforcement of the German injunction would interfere with the Chinese action;
• in order to avoid the effects of the German injunction, Huawei must either withdraw from the German market or accept the license offered by Conversant in Germany, which sets a global FRAND royalty rate that is 18.3 times higher than the rate set by the Nanjing court;
• imposition of the ASI will not materially prejudice Conversant’s rights in Germany with respect to the merits decisions (as opposed to the injunction decisions) of the German courts;
• imposition of the ASI will not harm the public interest;
• imposition of the ASI will not affect international comity, as the Chinese actions were brought before the German action and the ASI will not affect the subsequent trial in the German case or detract from the legal validity of the German judgment.

Also relevant to the SPC’s decision, though not specifically included in its consideration of the ASI, was the fact that in August, 2018, the Chinese State Intellectual Property Office (SIPO) ruled that the Chinese counterpart to the German Patent was invalid.  Conversant brought an administrative lawsuit challenging that ruling in the Beijing Intellectual Property Court.  That matter is still pending.

Upon entry of the ASI, the SPC established a penalty of RMB 1 million per day (approximately US$150,000) if Conversant violates the ASI.  Because the ASI was issued ex parte, Conversant was entitled to apply for reconsideration within five days.  It did so, and its request was denied on September 11, leaving the ASI in force.

Interdigital v. Xiaomi Link to translation

The dispute between InterDigital, a U.S.-based patent assertion entity, and Xiaomi, a Beijing-based electronics firm, relates to five InterDigital patents covering the 3G and 4G standards.   When negotiations for a license failed, on June 9, 2020 Xiaomi sought a declaration by the Wuhan Intermediate People’s Court of the appropriate FRAND royalty rate for the patents.  On July 29, InterDigital sued Xiaomi for infringement in the Delhi High Court in India, seeking monetary damages and an injunction. In response, on August 4, Xiaomi asked the Wuhan court for an “act preservation” ruling (ASI) preventing InterDigital from enforcing the injunction while the Wuhan proceeding was in progress.

The Wuhan court granted Xiaomi’s request, reasoning that:

• InterDigital’s filing of parallel litigation in India during the pendency of the Wuhan case showed a lack of respect for the Wuhan court and the case before it;
• The Indian action was intended to interfere with and obstruct the Wuhan case;
• The Indian action is likely to lead to a ruling in conflict with that of the Wuhan court;
• The entry of an injunction in India will likely cause significant harm to Xiaomi;
• Because InterDigital is a non-practicing entity that does not make products, but relies solely on licensing income, it would not be harmed substantially by the inability to seek an injunction against Xiaomi

Accordingly, on September 23, the Wuhan court (a) ordered InterDigital to withdraw its requests for injunctive relief in India, (b) prohibited InterDigital from seeking injunctive relief in China or any other country with respect to the 3G/4G patents currently at issue in the Wuhan case, and (c) prohibited InterDigital from asking another court in China or any other country to determine a FRAND royalty rate or resolve a FRAND dispute relating to the 3G/4G patents at issue.

As in Conversant v. Huawei, the Wuhan court established a penalty of RMB 1 million per day (approximately US$150,000) if InterDigital violates the ASI.  Because the ASI was issued ex parte, InterDigital was entitled to apply for reconsideration within five days.  It did so, and its request appears to have been denied, leaving the ASI in force.

In response to the entry of the ASI by the Wuhan court, InterDigital disclosed to the U.S. Securities and Exchange Commission (first reported by IAM) that on September 29, it filed for an anti-anti-suit injunction (AASI) in the Delhi High Court to prevent Xiaomi from enforcing its Chinese ASI.   This is the first instance of which we are aware of an AASI being sought in a FRAND case in India (prior examples being observed in Germany, France and the UK, as discussed in Contreras, Anti-Suit Injunctions All the Way Down).

Discussion

The ASIs issued by Chinese courts in Conversant v. Huawei and InterDigital v. Xiaomi signal a new willingness of Chinese courts to vie for jurisdictional authority in global battles over standard-essential patents.  To some degree, this development was entirely predictable.  Courts in the U.S. and UK have, for several years, sought to resolve global FRAND disputes being litigated on the international stage.  But there is no reason that U.S. and UK courts should be the exclusive venues for such proceedings.  The courts of every country in which the parties have sufficient assets to care about adverse rulings can exert similar authority.  As one of us has previously observed, this inter-jurisdictional competition is rapidly leading to a “race to the bottom” and a “race to the courthouse” in international FRAND disputes.

Yet even more seems to be at play in these two cases than China’s adoption of jurisdictional measures already deployed in the U.S. and UK. The ruling of the Wuhan court is far broader than the Supreme People’s Court’s ruling in Conversant v. Huawei in two major respects that are worth considering.  First, its geographic scope is not limited to the country in which InterDigital sought injunctive relief (India), but extends to all jurisdictions in the world.  As an anti-suit measure, this is more sweeping than any ASI issued in U.S. or other courts in FRAND cases.  Under U.S. and UK law, a court considering a request for an ASI must compare the action in the domestic court with the parallel action in the foreign court and determine whether they address the same matter and whether the resolution of the domestic action would dispose of the foreign action.  This type of analysis is not possible with an ASI that is not directed to a particular foreign action, but prospectively prohibits any attempt to seek an injunction anywhere in the world.

The second major addition by the Wuhan court is the prohibition on InterDigital’s ability to ask any other court in the world to determine a global FRAND rate for its 3G/4G patents.  As such, China has clearly joined the international race to be the jurisdiction of choice for determining FRAND royalty rates in global disputes involving standard-essential patents.

[1] A FRAND commitment is an obligation undertaken by a participant in a standards development organization (SDO) to grant licenses of its patents that are essential to implementation of one of the SDO’s standards on terms that are fair, reasonable and nondiscriminatory.  For a primer on standards development and FRAND commitments, see this chapter.

Guest post by University of Utah College of Law Professor Jorge L. Contreras. 

The Ninth Circuit’s recent decision in FTC v. Qualcomm (9th Cir., Aug. 11, 2020) is generally viewed as a resounding victory for Qualcomm.  In a strongly worded opinion, the Ninth Circuit reversed the entirety of the district court’s holding, which found that Qualcomm violated Sections 1 and 2 of the Sherman Act.  The Ninth Circuit exonerated Qualcomm with respect to each of its allegedly anticompetitive practices, concluding that these practices merely reflected the flexing of Qualcomm’s “economic muscle” with admirable “vigor, imagination, devotion, and ingenuity” (slip op. at 55).

Among Qualcomm’s challenged practices was its refusal to license rival chip makers under patents that are essential to one or more wireless telecommunications standards (standards-essential patents or SEPs).  While the District Court found that this refusal violated Qualcomm’s antitrust duty to deal under Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985), the Ninth Circuit disagreed.  It reasoned that Qualcomm did not violate any duty to deal because it uniformly refused to grant patent licenses to chip makers and did not “single[] out any specific chip supplier for anticompetitive treatment” (slip op. at 35).

In praising Qualcomm’s egalitarian approach toward rival chip makers, the Ninth Circuit points out that instead of granting licenses to these rivals, Qualcomm merely “declines to enforce its patents” against them “even though they practice Qualcomm’s patents” (id). As such, the Ninth Circuit quips that Qualcomm’s “policy toward rival chipmakers could be characterized as ‘no license, no problem’” (id., emphasis added).  Yet, as I discuss below, this approach could actually be a very big problem, not only for Qualcomm, but for all patent licensors seeking to extract revenue from the most lucrative point in the supply chain.

The Patent Exhaustion Doctrine and Chip Sales

As the Supreme Court explained in Quanta Computer, Inc. v. LG Elecs., Inc., 553 U.S. 617, 625 (2008), “The longstanding  doctrine of patent exhaustion provides that the initial authorized sale of a patented item terminates all patent rights to that item.”  That is, once the patent holder or its authorized licensee sells a product covered by a patent, that patent can no longer be asserted against a downstream buyer or user of the product. The patent is “exhausted” with respect to that particular product.

In Quanta, LG licensed three patents to Intel.  Intel manufactured chips allegedly covered by the patents, then sold the chips to Quanta for incorporation into Quanta’s PCs. LG then attempted to assert the patents against Quanta.  The court held that so long as the Intel chips “substantially embodied the patent[s]”, they were exhausted upon Intel’s sale of the chips to Quanta (553 U.S. at 633).  LG had no right to assert the patents against Intel’s customer Quanta.

Level Discrimination and SEPs

To grossly oversimplify, the supply chain for standardized wireless telecommunications functionality can be divided into three relevant tiers: (1) standards developers, (2) chip manufacturers, and (3) end user device (e.g., smartphone) manufacturers.  Standards developers like Qualcomm cooperate within standards-development bodies to create telecommunications standards like 4G LTE. Chip manufacturers then implement these standards in chipsets, which they sell to device manufacturers for incorporation into smartphones and other consumer devices.

What happens, however, when a standards developer like Qualcomm holds patents (SEPs) that cover a standard like LTE?  In theory, both the chips embodying the standard and the smartphones incorporating those chips infringe its SEPs.  Thus the SEP holder could choose to license those SEPs at either Tier 2 (chip manufacturers) or Tier 3 (device manufacturers). How to choose?

If a SEP holder licenses a chip manufacturer, then its SEPs covering a particular chip will be exhausted as soon as the manufacturer sell that chip to a device manufacturer, just as LG’s patents were exhausted in Quanta.  This means that if the SEP holder licenses a Tier 2 chip manufacturer, it cannot separately license, or collect royalties from, Tier 3 smartphone manufacturers for the same SEPs.  Qualcomm was keenly aware of the risk of patent exhaustion, which is why it refused to grant “exhaustive” licenses to chip makers like Intel. 411 F.Supp.3d at 748, 761.

If SEP royalties were standardized on a per-unit basis (e.g. $0.50 per product embodying the standard), then it would not matter whether the SEP holder licensed its SEPs at Tier 2 or Tier 3.  In either case it would receive the same payment.  However, due to longstanding industry practice, that is not how SEP royalties are calculated. Instead, they are usually based on some percentage (say 2.5%) of the price of the product embodying the standard.  So for a 4G LTE wireless radio chipset priced at $30, the royalty would be $0.75.  But for a $600 iPhone incorporating that chipset, the royalty would be $15.  For this reason, SEP holders strongly prefer to license their SEPs to end device makers (Tier 3).  As explained by one Ericsson licensing executive, “we choose to license the patents as late in value chain as possible …. One big advantage with this strategy is also that it is likely that the royalty income will be higher since we calculate the royalty on a more expensive product.” Or, as more succinctly expressed by a Qualcomm attorney at trial, licensing SEPs to device makers is “humongously” more lucrative than licensing  them to chip makers. 411 F.Supp.3d at 754, 758, 796.  The practice by which a SEP holder licenses its SEPs at only one tier of the supply chain is sometimes called “level discrimination.” (Courts and commentators disagree whether level discrimination is permitted under the nondiscrimination prong of a FRAND commitment – see this article for a discussion).

Pseudo-Licensing Deals with Chip Makers

If a SEP holder licenses its SEPs at Tier 3, what happens to the Tier 2 chip manufacturer? Does the chip that embodies the standard infringe the SEPs?  Yes, probably. Patent exhaustion only works downstream, not upstream.  That is, a smartphone manufacturer can’t infringe a SEP if it purchases a chipset from a licensed chip maker.  But a chip manufacturer can infringe a SEP even if its customer (the smartphone maker) has a license to use it.  Without a license, the Tier 2 chip maker is exposed to infringement claims by the SEP holder.

So what’s a chip maker to do?  Should it manufacture and sell chipsets that embody a standard even though it knows that it is infringing a host of SEPs?  Wouldn’t this infringement be willful, subjecting the chip maker to a risk of treble damages (see Sec. 5.2.1(1) of this chapter for a discussion of willful infringement of SEPs)?  It seems like an untenable situation for a chip maker.

To address this situation, Qualcomm appears to have developed various strategies.  In the 1990s, it granted chip makers purportedly “non-exhaustive licenses” that permitted them to manufacture chipsets covered by Qualcomm’s SEPs (in exchange for a royalty), but which explicitly excluded any license rights for the purchasers of those chipsets (9^th Cir., slip op. at 14 n.7).  In Quanta, the Supreme Court rejected such a “non-exhaustive” arrangement between LG and Intel, holding that LG’s patent rights were exhausted upon Intel’s sale of covered chips to Quanta.  After this, Qualcomm amended its practices and began to enter into “CDMA ASIC Agreements” with chip makers. Under these agreements, “Qualcomm promises not to assert its patents in exchange for the company promising not to sell its chips to unlicensed [smartphone manufacturers]” (9^th Cir., slip op. at 14, emphasis added).  According to the Ninth Circuit, these agreements “allow Qualcomm’s competitors to practice Qualcomm’s SEPs royalty-free” (id.).  Or, as the court pithily observed, Qualcomm’s “policy toward rival chipmakers could be characterized as ‘no license, no problem’” (id. at 35).

The Ninth Circuit found that because Qualcomm applied its “no license, no problem” policy uniformly toward all rival chip makers, it did not violate the antitrust laws.  But did Qualcomm, instead, open the door to a finding that its patents are exhausted at the chip maker level?

Do SEP Makers Inadvertently Grant Exhaustive Licenses to Chip Makers?

As observed by the Ninth Circuit, Qualcomm “promises not to assert” its SEPs against chip makers.  Its CDMA ASIC Agreements allow chip makers “to practice Qualcomm’s SEPs royalty-free”.  Ericsson, which employs a similar form of level discrimination, has referred to the result as “indirect licensing” of chip manufacturers (see Ericsson v. D-Link, 2013 U.S. Dist. LEXIS 110585, *80 (E.D. Tx. 2013)).

In assessing whether a patent has been licensed, courts have generally looked beyond the language used by the parties.  As the Supreme Court reasoned in De Forest Radio Telephone Co. v. United States, 273 U.S. 236, 241 (1927), “No formal granting of a license is necessary in order to give it effect. Any language used by the owner of the patent, or any conduct on his part exhibited to another from which that other may properly infer that the owner consents to his use of the patent in making or using it, or selling it, upon which the other acts, constitutes a license”.

A number of lower court cases have equated a license to a ‘covenant not to sue’.  As the Federal Circuit held in Ortho Pharmaceutical Corp. v. Genetics Institute, Inc., 52 F.3d 1026, 1031 (Fed. Cir. 1995), “A license may amount to no more than a covenant by the patentee not to sue the licensee for making, using or selling the patented invention.”

Given this precedent, SEP holders’ practice of tacitly permitting chip manufacturers to operate under their patents, whether by promising not to assert or “indirectly” licensing, looks suspiciously like licensing.  And, if SEP holders are granting chip manufacturers licenses to make and sell chips under their SEPs, then those SEPs should, by rights, be exhausted upon the sale of those chips to smartphone and other device manufacturers.  And this exhaustion should thereby prevent SEP holders from seeking to license and collect royalties from Tier 3 device manufacturers who incorporate those chips into their smartphones and other products.

This result should come as no surprise to anyone, least of all Qualcomm.  According to the District Court, a Qualcomm executive admitted to the IRS in 2012 that “if Qualcomm licensed a rival [chip manufacturer] … ‘[W]hen [the rival] sell[s] that chip to somebody who’s going to put the chip in a cell phone, okay, the licensee’s sale of that chip will exhaust our rights and then we won’t be able to collect a royalty on a cell phone that’s based on the price of the cellphone’” (411 F.Supp.3d at 796).  When Huawei apparently asserted that Qualcomm’s SEPs were exhausted after selling chips to Huawei, Qualcomm allegedly “threatened to cut off [Huawei’s] chip supply” (id. at 712).

These statements and actions indicate that Qualcomm was well-aware of the threat of patent exhaustion, and actually took measures to avoid the appearance of exhaustion (e.g., by converting its chip maker license agreements into CDMA ASIC Agreements).  Yet in trying to rebut the antitrust allegations made against it, and to overturn the District Court’s antitrust holdings, Qualcomm seems to have persuaded the Ninth Circuit that it effectively grants licenses to rival chip manufacturers. And, in doing so, Qualcomm may have armed its next smartphone licensee with a potent exhaustion defense to any claim of infringement.  Ultimately, “no license, no problem” may cause big problems for Qualcomm and other SEP holders that seek to license only at the most lucrative level of the supply chain.

Guest post by University of Utah College of Law Professor Jorge L. Contreras. 

In its Judgment of 26 August 2020, [2020] UKSC 37, the UK Supreme Court affirms the lower court decisions ([2017] EWHC 711 (Pat) and [2019] EWCA Civ 38) in the related cases Unwired Planet v. Huawei and ZTE v. Conversant [I discuss the High Court’s 2017 decision here].  The judgment largely favors the patent holders, and holds that a UK court may enjoin the sale of infringing products that incorporate an industry standard if the parties do not enter into a global license for patents covering that standard. The court covers a lot of important ground, including the parties’ compliance with EU competition law under Huawei v. ZTE (CJEU, C-170/13, 2015) (¶¶ 128-158) and the appropriateness of injunctive remedies under UK law (¶¶ 159-169). But in this post, I will focus on what I consider to be the most significant aspect of the court’s judgment – its interpretation of the patent policy of the European Telecommunications Standards Institute (ETSI), an interpretation that largely determines the outcome of the case and could have far-reaching ramifications for the technology sector.

Background

The case began in 2014 when Unwired Planet, a U.S.-based patent assertion entity (PAE), sued Huawei and other smartphone manufacturers for infringing UK patents that it acquired from Ericsson (other suits were brought elsewhere). The patents were declared essential to the 2G, 3G and 4G wireless telecommunications standards developed under the auspices of ETSI, an international standards-setting organization (SSO).  A companion case was brought by another PAE, Conversant, with respect to similar patents that it acquired from Nokia.

Because Ericsson and Nokia participated in standards-development through ETSI, they were bound by ETSI’s various policies, including its patent policy.  Accordingly, when the patents were acquired by Unwired Planet and Conversant, these policies continued to apply.  The ETSI patent policy requires that ETSI participants that hold patents that are essential to the implementation of ETSI standards (standards-essential patents or SEPs) must license them to implementers of the standards (e.g., smartphone manufacturers) on terms that are “fair, reasonable and non-discriminatory” (FRAND).

Huawei and ZTE are China-based smartphone manufacturers with operations in the UK.  Unwired Planet and Conversant offered to license the patents to them on a worldwide basis, but the manufacturers objected to their proposed royalty rates, claiming that they were not FRAND. Among other things, Huawei argued that any license entered to settle the UK litigation should cover only UK patents.  After numerous preliminary proceedings, in 2017 the UK High Court (Patents) held that a FRAND license between large multinational companies is necessarily a worldwide license. Moreover, if Huawei did not agree to such a worldwide license incorporating FRAND royalty rates determined by the court, the court would enter an injunction against Huawei’s sale of infringing products in the UK. The Court of Appeal largely affirmed the High Court’s ruling.

The UK Supreme Court Embraces SSO Policy Interpretation

In its judgment, the UK Supreme Court gives significant weight to the language and intent of the ETSI patent policy – far more than either the High Court of the Court of Appeal.  This approach contrasts starkly with that of the U.S. Court of Appeals for the Ninth Circuit, which decided another FRAND case, FTC v. Qualcomm (9th Cir., Aug. 11, 2020), just a fortnight earlier.  The Ninth Circuit explicitly dodged any interpretation of the SSO policies at issue in that case (those of the Telecommunications Industry Association (TIA) and Alliance for Telecommunications Industry Solutions (ATIS)), focusing solely on the antitrust issues raised by the parties.  The UK Supreme Court, in contrast, appears to have embraced the exercise of SSO policy interpretation, focusing intently on the language and drafting history of the ETSI policy, as well as its own conclusions about the intent of that language.  This judicial interpretive exercise leads to three key holdings in the case:

ETSI’s Policy Compels a Worldwide License

In determining that a FRAND license between Unwired Planet and Huawei should be global in scope, rather than limited to the UK, Mr Justice Birss of the UK High Court looked to industry practice and custom. He first noted that “the vast majority” of SEP licenses in the industry, including all of the comparable licenses introduced at trial, were granted on a worldwide basis, and both Unwired Planet and Huawei are global companies.  He then reasoned that “a licensor and licensee acting reasonably and on a willing basis would agree on a worldwide licence” (¶543).  In contrast, he regarded the prospect of two large multinational companies licensing SEPs on a country-by-country basis to be “madness” (¶543). Accordingly, the High Court held that  a FRAND license, under these circumstances, must be a worldwide license.

The UK Supreme Court acknowledges the industry practices referenced by the High Court, but bases its reasoning much more heavily on ETSI’s patent policy. First the Supreme Court recognizes the inherent territorial limitations on the jurisdiction of national courts (¶58). However, it is the ETSI patent policy, adopted by the SSO and accepted by its participants, that opens the door both to the consideration of industry practices (¶61) and the extension of national court jurisdiction to the determination of global royalty rates. The Court concludes, “[i]t is the contractual arrangement which ETSI has created in its patent policy which gives the [English] court jurisdiction to determine a FRAND licence” for a multi-national patent portfolio (¶58).  Thus, the Supreme Court affirms the decisions of the lower courts, but grounds its decision more firmly in the ETSI patent policy.

ETSI’s Policy Contemplates Injunctions

The Supreme Court also relies on the ETSI patent policy to support its conclusion that SEP holders may seek injunctions against standards implementers who do not enter into FRAND license agreements. The availability of injunctions in FRAND cases has been the subject of considerable debate in jurisdictions around the world. The UK court comes down in favor of allowing such injunctions, not on the ground that patent holders can do whatever they like, but because “[t]he possibility of the grant of an injunction by a national court is a necessary component of the balance which the [patent] policy seeks to strike, in that it is this which ensures that an implementer has a strong incentive to negotiate and accept FRAND terms for use of the owner’s SEP portfolio” (¶61).

This conclusion is striking in two regards.  First, it largely omits the analysis of EU competition law that typically accompanies the consideration of injunctive relief in EU FRAND cases.  While the Court later discusses Huawei v. ZTE at length (¶¶ 128-158), it does so while analyzing whether the parties violated applicable competition law, not whether competition law itself establishes a basis for seeking injunctive relief.

Second, and more surprisingly, the Court imputes to the ETSI patent policy an affirmative authorization to seek injunctive relief that is found nowhere in the policy itself.  From the fact that the patent policy includes provisions that are favorable to both implementers and SEP holders, the Court finds that the policy intended to establish a “balance” between these two groups, and that a “necessary component” of that balance is the ability of the SEP holder to seek an injunction against the implementer.

This is a surprising result that was not forecast in either of the decisions below. It is particularly significant because it may influence other courts’ interpretations of the ETSI patent policy.  It may also encourage other SSOs, if not ETSI itself, to adopt policy language expressly prohibiting participants from seeking injunctive relief against adopters of their standards (as IEEE has already done).

 Non-Discrimination is Not a Stand-Alone Commitment

The third significant aspect of the judgment relates to the non-discrimination (-ND) prong of the ETSI FRAND commitment.  At the High Court, Mr Justice Birss held that the -ND part of a FRAND commitment does not have a “hard edge”, which would mandate that every FRAND license must be priced at exactly the same rate. Instead, based on EU competition law, he found that differences in pricing should not be objectionable unless they distort competition.  As such, he did not fault Unwired Planet for pricing some FRAND licenses below the rates that it offered to Huawei.

I disagreed with Justice Birss’s reasoning on this point in 2017, arguing that it “conflate[s] two issues: the competition law effects of violating a FRAND commitment, and the private “contractual” meaning of the FRAND commitment itself.” I was thus pleased to see that the UK Supreme Court looks not to competition law, but to the content of the ETSI patent policy, to define the scope of the SEP holders’ non-discrimination obligation.

This being said, the Court’s interpretation of “non-discrimination” is novel and somewhat radical.  Rather than considering the -ND prong of FRAND to be an independent commitment of the SEP holder – that the licenses it grants not discriminate (however that is defined) — the Court blends the -ND prong  together with the “fair and reasonable” (FRA-) prong to form a “general” obligation.  It explains,

Licence terms should be made available which are ‘fair, reasonable and non-discriminatory’, reading that phrase as a composite whole. There are not two distinct obligations, that the licence terms should be fair and reasonable and also, separately, that they should be non-discriminatory. Still less are there three distinct obligations, that the licence terms should be fair and, separately, reasonable and, separately, non-discriminatory” (¶113 (emphasis added)).

As evidence for its interpretation, the Court points to ETSI’s rejection, in 1993, of a ‘most-favored license’ clause in its patent policy.  Interpreting the policy’s non-discrimination commitment as a ‘hard edged’ commitment would, in the Court’s view, re-introduce most-favored treatment “by the back door” (¶116).  As a result, the Court concludes that the non-discrimination prong of ETSI’s FRAND commitment merely “gives colour to the whole and provides significant guidance as to its meaning. It provides focus and narrows down the scope for argument about what might count as ‘fair’ or ‘reasonable’ for these purposes in a given context” (¶114).

As far as I am aware, the elimination of non-discrimination as a separate pillar of the FRAND obligation is at odds with both U.S. case law and the academic literature that address this issue (an overview can be found here).  The Court’s reasoning also contradicts the explicit concerns of the European Commission, which emphasized the importance of non-discrimination during debates over the ETSI patent policy in 1992:

Terms and conditions applied to participants and non-participants should not significantly discriminate against the latter. A fortiori where the standard-making body acts in an official or quasi-official standard-making capacity and where its standards are recognized and even made compulsory by virtue of legislation, access to the standard must be available to all without a precondition of membership of any organization (Communication from the Comm’n, Intellectual Property Rights and Standardization at p. 19, 27 Oct. 1992).

Clearly, the Commission did not view non-discrimination simply as giving color to the meaning of ‘fair and reasonable’.  On the contrary, non-discrimination, standing alone, is among the most important features of the FRAND commitment.  The UK Supreme Court’s interpretation to the contrary is thus highly problematic.

Conclusions

While I applaud the UK Supreme Court’s shift from a focus on competition law to the language and intent of the ETSI patent policy, I am concerned about its conclusions regarding the authority of one country’s courts to determine global FRAND rates, the availability of injunctive relief against standards implementers and the demotion of non-discrimination as an independent prong of the FRAND analysis.

One silver lining in this cloud, perhaps, is that the Court’s judgment, which relies so heavily on the particulars of the ETSI policy, is thus limited to the ETSI policy.  It is unclear how much weight its findings would have for a court, whether in the UK or elsewhere, assessing participants’ obligations under FRAND policies adopted by different SSOs such as TIA and ATIS (as in FTC v. Qualcomm), not to mention SSOs such as IEEE that have adopted language expressly contravening some of the interpretations that the Court makes with respect to ETSI.

In fact, the Court seems to invite SSOs to re-evaluate their patent policies.  Huawei objected to the UK court’s determination of global FRAND rates because, among other things, permitting a national court to resolve a global dispute could promote “forum shopping, conflicting judgments and applications for anti-suit injunctions” (¶90).  The Court tacitly agrees, but then pushes back, seeming to blame SSOs for allowing this to happen:

“In so far as that is so, it is the result of the policies of the SSOs which various industries have established, which limit the national rights of a SEP owner if an implementer agrees to take a FRAND licence. Those policies … do not provide for any international tribunal or forum to determine the terms of such licences. Absent such a tribunal it falls to national courts, before which the infringement of a national patent is asserted, to determine the terms of a FRAND licence. The participants in the relevant industry … can devise methods by which the terms of a FRAND licence may be settled, either by amending the terms of the policies of the relevant SSOs to provide for an international tribunal or by identifying respected national IP courts or tribunals to which they agree to refer such a determination” (¶90).

In this regard, I wholeheartedly agree with the Court. I have long advocated the creation of an international rate-setting tribunal for the determination of FRAND royalty rates.  I continue to believe that such a tribunal, if supported by leading SSOs, would eliminate much of the inter-jurisdictional competition and duplicative litigation that currently burdens the market.  If the UK Supreme Court’s judgment in Unwired Planet encourages ETSI and other SSOs to endorse such an approach, then this could be the most significant outcome of the case.

by Shubha Ghosh, Crandall Melvin Professor of Law and Director, Syracuse Intellectual Property Law Institute

As companies voluntarily retire their offensive trademarks, two questions tug at whatever passes for a conscience nowadays.  First, can these undesirable marks come back, revived by whomever sees a market niche for these symbols? This may seem like a ridiculous possibility, but on June 21, 2020, an Intent to Use Application was filed  on the word mark “Aunt Jemima” by Retrobrands, a Florida LLC, whose mission “is to revive ‘abandoned’ consumer iconic brands and to bring them back to the marketplace.” The second question is, do these intellectual property mea culpas do any good in the face of companies like Retrobrands and the entrenched nostalgia it represents?  After all, gallons of maple syrup were transformed into Benjamins, even more Tubmans, over the years. Should not there be some disgorgement in the form of reparations?

The proposed doctrine of vestigial use under federal trademark law can address both questions. As described below, vestigial use can prevent trademark abandonment, which would potentially allow some enterprising cultural chauvinist from appropriating the mark. Vestigial use, as applied, can also provide a new revenue source that can finance the necessary reparations.

A vestigial use is the use of a mark to maintain the memory of a brand. Instead of offensive symbols littering the shelves of your local grocery store, they can be relegated to a museum. The idea would be similar to that of Budapest’s Memento Park, where the brutalist statues  from the Soviet era have a fitting resting place, about a forty minute bus ride from the Budapest bus terminal in a rural outskirt more habitable than Siberia but just as overlooked. Memento Park is a reminder of ideas gone woefully wrong.

Vestigial use would allow a trademark owner to continue using a mark without completely abandoning it. A trademark owner can claim continued use of the mark as a curator with the once fully commercialized mark relegated to a virtual museum. Retired offensive marks would be hidden away, but still can serve as a reminder of what was once bought and sold freely in the United States.  History is neither erased nor glorified. Instead, these marks become objects of study, specimens of bigotry past and continuing. These vestigial marks would have a place on a pedestal much like Uncle Tom’s Cabin, Huckleberry Finn. The Yellow Wallpaper, Native Son, and To Kill a Mockingbird have a protected place in public libraries. Like these books, vestigial marks benefit the public through their presence, serving as the occasional fodder for those who like to ban cultural relics and as a constant memento of a persistent ugliness.

I envision vestigial use as a judge-made doctrine, although with some imagination an appropriate statute could be drafted. It would allow trademark owners to defeat a claim of abandonment by someone who seeks to take up the mark. Under section 1127 of the Lanham Act, abandonment occurs when a mark’s “use has been discontinued with intent not to resume such use.”  The provision defines use as “the bona fide use of such mark made in the ordinary course of trade, and not made merely to reserve a right in a mark.” If vestigial use were adopted, a trademark owner’s showcasing the history of the mark would count as bona fide use. For offensive marks, owners like Quaker Oats, or its licensees, could create the equivalent of a Memento Park while blocking others from adopting the mark for full profit-making branding.

Vestigial use is necessary because failure to commercialize the mark may constitute abandonment under current doctrine. If Quaker Oats used the Aunt Jemima mark solely as a museum relic, that would most likely constitute a discontinuation of the mark “with an intent not to resume such use.”  Under the Lanham Act, use requires the goods labelled with the mark “are sold or transported in commerce.” Vestigial use of the mark would not require sales or transportation in commerce in the traditional sense. In fact, vestigial use presumes the mark is removed from the national mass market associated with the establishment of trademark rights. However, an archival use entails interested viewers going towards the trademarked product much like visitors to a museum. Instead of many products transported to purchasers, vestigial use entails many purchasers travelling to view the archived mark. The mapping is not from many marked goods to many purchasers but from many viewers to the one marked product.

As far as analogies go, this is admittedly a stretch. But if trademark law is about creating and protecting associations, it is not completely clear why the law’s protection should be limited to the many products going out to many buyers. First, there is nothing sacred about many to many; trademark law protects marks on niche products, even in markets with a few consumers. It is also not clear why the transportation has to be from product to buyers. Trademark law protects marks on websites which are unique objects which attract buyers.  The transportation requirement under the statute encapsulates the many consumers going to a single branded product. Vestigial use is a possible interpretation of the statute.

But possible does not mean highly probable. Courts may be hesitant in reinterpreting the use requirement to make it ostensibly easier to obtain rights. My proposal, however, is limited to applying vestigial use as a limitation on abandonment as opposed to lowering the use requirement for registration or demonstrating ownership. Potential evisceration of the public domain may also be the concern. Vestigial use sounds like warehousing of marks. But the proposal requires some affirmative act of the mark holder, namely a memorial or curated use. Furthermore, the public domain is a strange status under trademark law. An abandoned mark is not really in the public domain. As Professor Gerhardt argues in her blogpost, Quaker Oats would have rights under 1125(a) against companies like Retrobrands. Vestigial use is designed to prevent new users from fully recapturing supposedly abandoned marks.

Picking up on the public domain point, some policy makers may turn to genericide as the appropriate doctrine to put these offensive marks to rest. Arguably the racial stereotypes underlying “Aunt Jemima” or “Eskimo” are generic signifiers for outdated and commonly understood denigrating tropes. However, genericide means that the mark refers to the genus to which a product belongs, not about the signification of marks. Consequently, genericide is not practically suited to put to rest generically offensive marks.

Even if the proposed doctrine of vestigial use is unconvincing, analyzing it reveals the underlying problem. One reason for posts like this is that the Supreme Court’s decision in Tam rules out the possibility of denying registration or other trademark protections to offensive marks. According to the Court’s logic, “Aunt Jemima” is just another viewpoint under the First Amendment, and viewpoints do not ever really disappear. Their persistence is evinced by companies like Retrobrands and the recurring terms of the current public discourse. Vestigial use acknowledges the persistence of offensive discourse, despite all good intentions. Appropriating this persistence through trademark protection for memorial uses is a reminder of the terms of the cultural and political debate, Memory may at some date lead to the substantive economic and political reforms needed to combat inequality and bigotry.  Marks are never dead, and the past is never past, even if undistinguished.