Eight Mile Style, LLC v. Spotify U.S. Inc.

Full title: EIGHT MILE STYLE, LLC and MARTIN AFFILIATED, LLC, Plaintiffs, v. SPOTIFY…

Court: United States District Court, M.D. Tennessee, Nashville Division

Date published: Aug 15, 2024

3:19-cv-0736 (M.D. Tenn. Aug. 15, 2024)

Opinion

3:19-cv-0736

08-15-2024

EIGHT MILE STYLE, LLC and MARTIN AFFILIATED, LLC, Plaintiffs, v. SPOTIFY USA INC. and HARRY FOX AGENCY, LLC, Defendants. SPOTIFY USA INC., Third-Party Plaintiff, v. KOBALT MUSIC PUBLISHING AMERICA, INC., Third-Party Defendant

ALETA A. TRAUGER, UNITED STATES DISTRICT JUDGE

MEMORANDUM

ALETA A. TRAUGER, UNITED STATES DISTRICT JUDGE

There are four currently pending motions for summary judgment. The Harry Fox Agency (“HFA”) has filed a Motion for Summary Judgment (Doc. No. 317), to which Eight Mile Style, LLC and Martin Affiliated, LLC (collectively, “Eight Mile Style”) have filed a Response (Doc. No. 438), and HFA has filed a Reply (Doc. No. 503). Spotify USA Inc. (“Spotify”) has filed a Motion for Summary Judgment (Doc. No. 320), to which Responses have been filed by Kobalt Music Publishing America Inc. (“Kobalt”) (Doc. No. 441) and Eight Mile Style (Doc. No. 436), and Spotify has filed a Reply to each Response (Doc. Nos. 508, 511). Kobalt has filed a Motion for Summary Judgment (Doc. No. 328), to which Spotify has filed a Response (Doc. No. 525-1), and Kobalt has filed a Reply (Doc. No. 501). Finally, Eight Mile Style has filed a Motion for Summary Judgment (Doc. No. 348), to which Responses have been filed by Spotify (Doc. No. 426), HFA (Doc. No. 414), and the United States as an interested party (Doc. No. 550), and Eight Mile Style has filed a Reply to each Response (Doc. Nos. 514, 515, 516). The court heard oral arguments on July 12, 2024, after which the litigants filed additional briefing. (See Doc. Nos. 691, 693, 698, 703). For the reasons stated herein, HFA's motion will be granted, Spotify's motion will be granted in part and denied in part, Kobalt's motion will be granted in part and denied in part, and Eight Mile Style's motion will be denied.

I. INTRODUCTION

The 21st Century has seen a radical transformation in the ways that recorded music is offered, distributed, obtained, and enjoyed. That transformation-driven, in chief, by advancements in digital information technology-first took the form of widespread illicit file sharing, but that practice fell out of prominence once the Supreme Court firmly established, in MGM Studios, Inc. v. Grokster, Ltd., 545 U.S. 913 (2005), that the operators of services that enabled that unlawful sharing of others' copyright-protected works faced potentially astronomical liability based on their complicity in the actions of their users. In the wake of that holding, an ostensibly licit economy in music downloads and streaming arose that included, as a dominant player, Spotify. Spotify was founded in 2006 and is now reportedly the world's most popular audio streaming subscription service. (Doc. No. 437 ¶¶ 1-5.)

The fact that companies like Spotify avoided the overt, obvious infringement associated with peer-to-peer file sharing, however, did not mean that everything that they did was actually legal. In fact, such companies, in their jostling for market position, faced powerful incentives to play fast and loose with copyright law in the name of building the most comprehensive-and, therefore, enticing-music libraries possible. The traditional licensing scheme that governed musical compositions in the United States was not designed with modern streaming services in mind, and assembling the licenses necessary for a comprehensive library required a great deal of information gathering and paperwork that had to be completed, painstakingly, on a track-by-track basis, for every composition at issue. In companies' rush to build their businesses, many necessary rights were obtained but many were not, and many recordings that should not have been streamed were.

Being a high-volume infringer, though, can get expensive. The Copyright Act permits a rights owner to choose between recovering its actual damages, plus the infringer's profits, or receiving statutory damages up to $150,000 for each work at issue. 17 U.S.C. § 504(b)-(c). Given the size of streaming services' libraries, those statutory damages could add up. In 2019, for example, a jury awarded film and television studio plaintiffs a statutory damages judgment of over $61 million against a small streaming provider based on its infringement of 819 works.¹Such statutory damages create a floor, not a ceiling; a successful streaming provider might face even more substantial ramifications from a judgment based on its profits. Any company that built its streaming empire on the back of its willingness to infringe, therefore, risked accumulating liabilities capable of swamping everything that it had accomplished.

1 See Top Story: Jury Finds Streaming Video Service Vidangel Liable For $62m In Damages, Wolter Kluwer Intellectual Property Law Daily, 2019 WL 2572671 (June 20, 2019).

Principles of copyright, however, are not immutable laws of nature; they are established, and can be changed, by Congress. In 2018, Congress, “amid lobbyists' shouts,” Flo & Eddie, Inc. v. Sirius XM Radio, Inc., 9 F.4th 1167, 1172 (9th Cir. 2021), enacted the Orrin G. HatchBob Goodlatte Music Modernization Act (“MMA”), Pub. L. No. 115 264, 132 Stat. 3676 (2018), which attempts to both address the morass that had arisen during the early days of the streaming economy and establish a better-suited framework going forward. In so doing, it both establishes a new, simpler framework through which digital streaming providers can obtain compulsory licenses and provides a liability limitation regarding past infringement for digital streaming providers who complete certain required steps.

Eight Mile Style owns the copyrights to a number of compositions (“EMS Compositions”) made popular, for the most part, by the rapper Marshall Mathers, who performs under the name Eminem. The most widely listened-to of those compositions-a song called “Lose Yourself”-has been streamed by Spotify users more than one billion times. (Doc. No. 437 ¶ 36.) Eight Mile Style alleges that the EMS Compositions were among those streamed unlawfully by Spotify, for years, without a valid license. Eight Mile has sued Spotify for that infringement, and, while its claims were brought after the January 1, 2018 effective date of the MMA's liability limitation, Eight Mile Style seeks its full traditional copyright damages, arguing that the MMA liability limitation is unconstitutional, and that, in any event, Spotify did not comply with the requirements of the limitation.

Eight Mile Style's right to recovery, however, faces a legal obstacle far older than the MMA. The law has long disfavored plaintiffs who strategically exploit regimes of civil liability to maximize their own recoveries at the expense of the public good and in contravention of basic principles of fairness. Accordingly, the well-established principle of equitable estoppel-which, the Supreme Court has expressly acknowledged, retains its vitality in the context of federal copyright law, see Petrella v. Metro-Goldwyn-Mayer, Inc., 572 U.S. 663, 684 (2014)-bars recovery for claims that, though nominally supported, were tainted by the plaintiff's knowing, culpable role in fostering the defendant's unlawful behavior. The evidence in this case shows that Eight Mile Style was not a hapless victim, but, rather, a sophisticated steward of its copyrights that was aware that the licensing status of the EMS Compositions had fallen into confusion and simply allowed its rights to be violated in a way that would be entirely inexplicable other than as a strategic choice to manufacture infringement damages. The doctrine of estoppel disfavors that practice and, at least in this instance, forbids it.

While Spotify's handling of composer copyrights appears to have been seriously flawed, any right to recover damages based on those flaws belongs to those innocent rights holders who were genuinely harmed-not ones who, like Eight Mile Style, had every opportunity to set things right and simply chose not to do so for no apparent reason, other than that being the victim of infringement pays better than being an ordinary licensor. Accordingly, the court will grant Spotify summary judgment and will leave any decision regarding the MMA for a future case involving an appropriate plaintiff. For reasons discussed herein, the court will also enforce Spotify's right to indemnification from Kobalt.

II. BACKGROUND

A. Music Streaming, Copyrights, and the MMA

1. Pre-MMA Law

The Copyright Act recognizes two distinct sets of rights in connection with recorded music: first, rights in musical compositions, which the Act refers to as “musical works,”; and, second, rights in the recorded performances of compositions, which the Act refers to as “sound recordings” distributed, historically, in the form of “phonorecords.” See 17 U.S.C. §§ 114-115 Although music streaming involves the transmission of sound recordings, it implicates both sets of rights, because an entity cannot lawfully transmit an audio recording of a performance of a copyright-protected composition to its users without a so-called “mechanical license”² to do so. See 17 U.S.C. § 115. Indeed, copyrights in compositions have historically been the more prominent of the two sets of rights, because, among other things, “pre-1972 recordings . . . were not protected under federal copyright law” until fairly recently. Flo & Eddie, Inc. v. Pandora Media, Inc., No. CV147648PSGGJSX, 2020 WL 6336124, at *1 (C.D. Cal. Oct. 22, 2020). This case solely involves copyrights in compositions.

2 The term “mechanical license” is an anachronism from the time when such licenses involved “‘mechanically' recording a song on media such as phonographic records or piano rolls.” EMI Ent. World, Inc. v. Karen Recs., Inc., 603 F.Supp.2d 759, 762 (S.D.N.Y. 2009) (citing 3 Nimmer on Copyright § 8.04[A], at 8-58-4 (2008)).

Prior to the MMA, mechanical licenses could be obtained on a composition-by-composition basis through a compulsory licensing process-that is, without the affirmative consent of the rights holder-if certain procedural requirements were observed. Specifically, “[o]nce a copyright owner distribute[d] the musical work to the public, . . . anyone [could] obtain a compulsory license in the musical work by serving [Notice of Intention to Obtain a Compulsory License for Making and Distributing Phonorecords (‘NOI')] on the copyright owner within the applicable time frame and following other specific requirements set out in the copyright regulations.” Yesh Music, LLC v. Amazon.com, Inc., 249 F.Supp.3d 645, 651 (E.D.N.Y. 2017) (citing 17 U.S.C. § 115(a), (b)) (internal quotation omitted). A compulsory license entitled the notice sender to use the underlying composition at a so-called “statutory rate” set by the Copyright Royalty Board. The first such rate for interactive streaming was set in 2009. (Doc. No. 438-1 ¶ 5.) For most of this system's existence, it was required that an NOI be sent prior to any distribution of the recording at issue, meaning that an NOI sent after a track had already been streamed would be ineffective. See 17 U.S.C. § 115 (2017).

2. Streaming Services Struggle to Comply

As others have observed, “th[e] compulsory licensing system became unworkable in the digital music streaming era,” as digital streaming providers “struggled to serve or file a notice of intention for every one of the millions of works available on their services.” Lowery v. Rhapsody Int'l, Inc., 75 F.4th 985, 989 (9th Cir. 2023). As Spotify explained in a May 23, 2014 statement to the Copyright Office, “[i]dentifying and locating the co-authors of each of millions of copyrighted musical works is a daunting task that is hampered significantly by” the fact that there was no singular, official public database of rights holder information. (Doc. No. 517-4 at 4.)

One of the ways that Spotify and others sought to obtain the rights they needed was by entering into so-called “blanket mechanical licensing agreements” with major music publishers who controlled large catalogs of musical works. (Doc. No. 509 ¶ 9.) For example, in 2011, Spotify entered into such an agreement with Warner/Chappell Music, Inc.-one of the largest music publishers in the world, see Broad. Music, Inc. v. Pandora Media, Inc., 140 F.Supp.3d 267, 285 (S.D.N.Y. 2015)-governing, broadly speaking, Warner/Chappell's full available catalog. (See Doc. No. 606-13.) That agreement, however, did not include a schedule actually identifying the compositions at issue, and Spotify concedes that it had only “general industry knowledge as to the musical works owned, controlled, or administered by major music publishers and publishing administrators.” (Doc. No. 509 ¶ 115.) Such blanket licenses, accordingly, introduced an additional element of uncertainty regarding whether any given composition had been licensed.

Of course, a streaming service could still comply with the law simply by erring on the side of caution and streaming only tracks for which it knew it had a valid license. That, though, was not always what happened-which, given business realities, is not surprising. As one of Spotify's experts in this litigation explained, “streaming services had to offer vast catalogs of music to satisfy consumer demand and thereby remain commercially viable.” (Doc. No. 604-3 at 30.) That pressure was likely especially acute with regard to star musical artists whose absence from a platform would risk alienating many customers or potential customers. A digital streaming provider's short-term business incentives, therefore, were in direct tension with the need to ensure that all of the underlying copyrights had been duly licensed.

There is room for debate regarding how scrupulous or sloppy companies like Spotify were, but there is little reasonable dispute that at least some works were added to streaming libraries when they should not have been. Such unauthorized uses have already been the subject of substantial litigation. Most prominently, on March 17, 2016, Spotify and the NMPA reached a settlement agreement allowing publishers to claim and receive royalties in connection with works for which Spotify “may have been unable to provide a proper accounting of usage.” (Doc. No. 437 ¶ 311.)

3. The MMA

In 2018, Congress enacted the MMA in an effort to address the growing problems associated with copyrights related to music streaming. The MMA empowers the Register of Copyrights, with the approval of the Librarian of Congress, to designate a “nonprofit entity . . . that is created by copyright owners” but “not owned by any other entity” as the official “mechanical licensing collective” for the purposes of the MMA. 17 U.S.C. § 115(d)(3). Although the MMA does not require that the chosen entity use “Mechanical Licensing Collective” (“MLC”) as its formal name, that is what happened. The MMA provides that, once the MLC system is in place, “[a] digital music provider may obtain a blanket license” from the MLC “by submitting a notice of license” to it. 17 U.S.C. § 115(d)(2)(A). Subject to an exception for certain works covered by negotiated (that is, non-compulsory) licenses, an MLC blanket license “covers all musical works (or shares of such works) available for compulsory licensing.” 17 U.S.C. § 115(d)(1)(B). The MLC process, therefore, was designed to eliminate the need for conventional track-by-track licensing through NOIs.

The task that the MMA sets for the MLC, however, is a demanding one, and there was no way that the MLC could have started its work immediately upon the MMA's enactment. Indeed, the MMA does not even require an official mechanical licensing collective to be selected until, at the latest, 270 days after the enactment date. 17 U.S.C. § 115(d)(3)(B)(a). When this litigation began, the MLC system had not yet gone into effect; the effective date was January 1, 2021. Congress could simply have left the old NOI-based structure in place until the MLC could be up and running and, in many ways, it did. However, the MMA expanded the window for service of an NOI to include the first 30 days after a digital music provider's first distribution of a track- rather than requiring an NOI to precede that first distribution. See 17 U.S.C. § 115(b)(1).

The MMA also established a liability limitation potentially applicable to claims for infringement against digital music providers that were filed “on or after January 1, 2018” but “prior to the license availability date”-that is, January 1, 2021. 17 U.S.C. § 115(d)(10)(A). If the limitation is found to apply to an instance of infringement, then “the copyright owner's sole and exclusive remedy against the digital music provider” is that the owner is “eligible to recover the royalty prescribed under” the ordinary compulsory licensing structure-that is, the amount of money that the digital streaming provider would have owed if it had filed a timely NOI, even though it did not. Id.

The MMA's liability limitation, however, is not automatic. Rather, if a digital music provider wishes to rely on the limitation, it must take a number of steps, which it must initiate “[n]ot later than 30 calendar days after first making a particular sound recording of a musical work available through its service via one or more covered activities, or 30 calendar days after the enactment date, whichever occurs later.” 17 U.S.C. § 115(d)(10)(B)(1).

B. Spotify's Streaming of the EMS Compositions

There are 243 EMS Compositions. Eight Mile Style now concedes that one of those compositions-the song “My Name Is”-belongs to someone else. Eight Mile Style, however, continues to assert that it possesses interests in the other 242 compositions. (Doc. No. 436-1 ¶¶ 33-34.) While Spotify has taken issue with aspects of Eight Mile Style's documentation regarding that ownership, it has identified no substantive reason to disregard Eight Mile Style's assertion of those interests based on the registration of those compositions with the Copyright Office.

Spotify has openly streamed recordings embodying at least some of the EMS Compositions since its 2011 U.S. launch. (Id. ¶ 60.) For the entirety of that period, Spotify has paid royalties associated with that streaming to Eight Mile Style's collection agent, Kobalt, as if a license had been in place (Doc. No. 436-1 ¶ 281; Doc. No. 437 ¶ 64), and Kobalt provided Eight Mile Style with a quarterly document summarizing the royalties being paid. (Doc. No. 436-1 ¶ 284.) A 2012 prospectus confirms that Eight Mile Style was aware that the songs were being streamed on Spotify and that Eight Mile Style was being paid royalties for those streams. (Doc. No. 437 ¶¶ 67-72.) Eight Mile Style concedes that, “[a]t no time prior to initiating this litigation did [it] tell Spotify that its use of the [EMS Compositions] was unauthorized or instruct Spotify to cease and desist its use of the [EMS Compositions].” (Doc. No. 436-1 ¶ 63.)

C. Relationships Central to this Case

Although the core intellectual property dispute in this case is between Eight Mile Style and Spotify, there were a number of other parties who played a significant role in the underlying events-including two entities, Kobalt and HFA, that have become defendants in their own right. The complex web of underlying intellectual property agreements means that the nature of the parties' relationships-including, in particular, whether Kobalt “administered” the EMS Compositions-is key to determining whether or not Spotify had a license and, if not, whether it is entitled to indemnification. Those relationships, broadly speaking, can be grouped into two clusters: those on the side of the copyright holders, and those on the side of Spotify, the potential licensor.

1. On the Rights Holder Side (Eight Mile Style, Bridgeport, and Kobalt)

The plaintiffs, whom the court is collectively calling “Eight Mile Style,” are music publishing companies founded and run by Joel Martin. (Doc. No. 427 ¶ 1.) The day-to-day management of Eight Mile Style is performed by Martin and Sarah Catlett, whom Martin characterizes as an independent contractor, but who has worked with Martin since 1995. (Doc. No. 438-1 ¶¶ 54-56; Doc. No. 436-1 ¶¶ 44-45.) Eight Mile Style also admits that it “retained the services of an accountant experienced with auditing books and records of companies that distribute royalties.” (Doc. No. 438-1 ¶ 60.)

Martin's longtime friend Armen Boladian founded and operates a third, formally distinct music publishing company, Bridgeport Music, Inc. (“Bridgeport”), over which Martin exercises practical control. (Doc. No. 436-1 ¶ 39.) Bridgeport's and Eight Mile Style's operations are closely intertwined. For example, since July 1, 2009, Martin Affiliated has performed rights administration services in connection with compositions owned or controlled by Bridgeport, and Joel Martin was the Rule 30(b)(6) witness for both entities in this litigation. (Id. ¶¶ 32, 40-43.) The companies also share personnel, including Catlett, and a physical office. (Id. ¶¶ 44-46.) Eight Mile Style concedes that Martin and Catlett handle Bridgeport's day-to-day operations. (Doc. No. 438-1 ¶ 68.)

Kobalt is an independent music publisher and rights administrator founded in 2000 by Willard Ahdritz.³(Doc. No. 437 ¶¶ 75-77.) Kobalt administers hundreds of thousands of musical works. That precise roster of administered works, however, is ever-changing, as new clients are added and existing contracts expire. (Id. ¶¶ 79-80.) “Joel Martin and/or his companies” were among Kobalt's first U.S. clients, and Kobalt has discussed purchasing the EMS Compositions in the past-although no such purchase has taken place. (Id. ¶¶ 81-83.) Mark Levinsohn, an attorney who has represented Eight Mile Style for about fifteen years, has also served as outside counsel to Kobalt, and he has had a relationship with Kobalt's now-General Counsel, Jim Arnay, for over two decades. (Id. ¶¶ 53-56.) Kobalt long touted its relationship with Eight Mile Style publicly. It listed Eight Mile Style among its clients on its website continuously from September 2004 to November 2019. (Doc. No. 437 ¶ 85.)

3 The court is simplifying this somewhat for clarity. There are actually multiple affiliated Kobalt entities using similar names and owned, together, by parent Kobalt Music Group, Ltd. The distinctions between these entities are mostly irrelevant to the issues before the court. (Doc. No. 437 ¶¶ 75-77.)

The first agreement between Eight Mile Style and Kobalt was executed by the parties on June 29, 2004, under the header “ADMINISTRATION AGREEMENT.” (Id. ¶ 82; Doc. No. 332-15.) The agreement identifies Kobalt as “the Administrator” and grants it a range of rights in connection with the EMS Compositions-including the right to grant mechanical licenses-but it identifies the United States, Canada, and Japan as “Excluded Territories,” exempting them from much of the authority granted in the agreement. (Doc. No. 332-15 at 2, 4, 6-9.) However, Eight Mile Style did grant Kobalt the right to grant what are typically referred to as “synchronization licenses”-essentially, licenses for use in works such as films and advertisements-within the Excluded Territories, subject to Eight Mile Style's “consent in each instance.” (Id. at 7.) Eight Mile Style and Kobalt entered into a similar agreement in 2007, which continued to include the U.S. and Canada as Excluded Territories. (Doc. No. 339-1 at 4.) Kobalt admits that, under the 2007 agreement, it had the right to register the EMS Compositions with collection societies, issue licenses to the EMS Compositions in the covered territories, and collect royalties relating to the EMS Compositions. (Doc. No. 437 ¶ 90.)

On June 22, 2009, Kobalt and Plaintiffs executed a Letter Agreement amending the 2007 Administration Agreement to include the U.S. and Canada. (Id. ¶ 91; Doc. No. 332-15.) Kobalt concedes that, following that amendment, it “had worldwide administration rights for” the EMS Compositions. (Doc. No. 437 ¶ 92.)

About a year and a half later, however, in February 2011, Eight Mile Style and Kobalt executed a new Administration Agreement reinstating the status of the U.S. and Canada as Excluded Territories. (Id. ¶ 94; Doc. No. 332-19.) Around the same time, Eight Mile Style entered into an Administration Agreement with Bridgeport, pursuant to which Bridgeport “agreed to administer the [EMS] Compositions throughout” the United States and Canada, under which Bridgeport would receive commissions. (Doc. No. 332-20 at 2, 5.)

As part of this reorganization, Kobalt entered into a separate, contemporary letter agreement with Bridgeport. The letter agreement frames itself as an adjunct to Kobalt's non-U.S. administration agreement with Eight Mile Style, which it refers to as the “the ‘Administration Agreement.'” (Doc. No. 335-1 at 1.) Pursuant to the letter agreement, Kobalt “acknowledged that [Bridgeport] own[s] and/or control[s] all rights in the Compositions for the Excluded Territories,” and Kobalt “agreed [to] provide certain services to [Bridgeport] in respect of the Compositions.” (Doc. No. 335-1 at 2.) Most of the listed services were related to receiving, processing, and allocating royalties. (Id. at 3.) The letter agreement, however, also granted Kobalt “the non exclusive right in the Excluded Territories to pitch the Compositions for synchronization uses in motion pictures, adverts, [interactive] games and television programs originating in the Excluded Territories,” subject to Bridgeport's prior written approval. (Id. at 23.) Kobalt concedes that the “2011 agreement with Bridgeport was unlike any other agreement Kobalt had with a client.” (Doc. No. 437 ¶ 97.)

The 2011 agreement between Bridgeport and Eight Mile Style included, as an attachment, a form “NOTICE AND DIRECTION TO THIRD PARTIES”-commonly referred to as a “letter of direction,” or “LOD”-that could be sent to parties to inform them of the change in the underlying rights. (Id. ¶ 108; Doc. No. 332-20 at 50.) Eight Mile Style does not dispute that, when the rights to license a composition change hands, “[i]t is customary for [the] incoming publishing administrator to send an LOD to advise third parties, including mechanical rights organizations and performing rights organizations, of its assumption of rights.” (Doc. No. 436-1 ¶ 106.) Bridgeport, however, never sent the LOD to any party, never registered interests in any of the EMS Compositions in its name with any U.S. mechanical rights organization, and, in fact, never “formally notified” any third party that it was taking over the mechanical licensing of the EMS Compositions in the U.S. (Id. ¶¶ 111-15.; Doc. No. 438-1 ¶ 83.)

Kobalt, in contrast, did send an LOD of its own to parties in early February 2011, informing them that its “exclusive administration agreement” with Eight Mile Style meant that “all statements, payments, license requests, correspondence and inquiries” involving the EMS Compositions “should be remitted to” Kobalt. (Doc. No. 335-19.)

Bridgeport does not appear to have performed much actual work in its capacity as the party with formal licensing authority for the EMS Compositions. It “did not issue any licenses in its name” for the compositions and, consistent with the formal division of responsibilities in the parties' agreements, it “did not collect royalties from third-party licensees, mechanical rights organizations, or performing rights organizations” for those compositions. (Doc. No. 437 ¶¶ 124-25.) In an October 2011 email exchange between Kobalt executives, one described the parties' underlying relationship as “Bridgeport ‘admining' North America, with us actually doing it.” (Id. ¶ 149.) Martin testified that he continued to believe, even after the Bridgeport agreement, that any license requests should have been sent to Kobalt, despite the fact that Kobalt was no longer the party with the formal authority to grant a license. (Id. ¶ 131.) Kobalt concedes that it did, in fact, continue to receive and pass along licensing requests related to the Eight Mile Style Compositions for years after the assignment of rights to Bridgeport. (Id. ¶ 225.) Kobalt also concedes that it actually “issued, delivered to third parties, or executed a handful of mechanical licenses” for the EMS Compositions in the early years after it formally lost its right to do so- issuing the last of those on March 28, 2013. (Id. ¶ 227; see Doc. No. 338-20 at 7.)

Communications between Catlett and Martin in 2013 confirm that, at least broadly speaking and for some purposes, the two continued to consider Kobalt an “administrator” of the EMS Compositions. (Doc. No. 437 ¶¶ 137-40.) They made similar statements to third parties numerous times. (Id. ¶¶ 141-65; see, e.g., Doc. No. 335-17 at 3 (referring to an EMS Composition as “admin'd by Kobalt so the request will have to go through them”).) As the agreements contemplated, moreover, Kobalt continued licensing the EMS Compositions within the U.S. for synchronization purposes. Kobalt concedes that, “[i]n every year from 2011 to 2019, Kobalt issued individual synchronization licenses to [the EMS Compositions] in the U.S. with the express authorization of” Eight Mile Style and/or Bridgeport. (Doc. No. 437 ¶ 231.)

Kobalt also participated in Eight Mile Style's rights enforcement efforts, albeit to a limited degree. Kobalt concedes that, on at least one occasion, it sent a cease and desist letter regarding an EMS Composition to a U.S-based party in which it referred to itself as the “exclusive administrator[] of the composition.” (Doc. No. 437 ¶ 270.) Eight Mile Style concedes that it “often included Jim Arnay, Kobalt's general counsel, through the process of enforcement actions it undertook itself.” This practice, according to Eight Mile Style, reflects the fact that “Kobalt and Eight Mile Style worked collegially together during the entire course of their relationship,” but did not reflect any actual enforcement authority by Kobalt during the relevant period. (Doc. No. 436-1 ¶ 269.)

2. On the Licensee Side (Spotify and HFA)

Spotify is a digital music provider founded in 2006. Its service first became available to users in Europe in 2008, and it launched in the United States in 2011. (Doc. No. 437 ¶¶ 1-4.)

HFA is a nearly 100-year-old company that provides rights management, mechanical licensing, and royalty distribution services to clients in and around the U.S. music publishing industry. See, e.g., Duchess Music Corp. v. Stern, 331 F.Supp. 127, 128 (D. Ariz. 1971) (discussing HFA). Prior to 2015, HFA was a subsidiary of the National Music Publishers Association, but it is now an independent entity. (Doc. No. 437 ¶ 16.) It has been plausibly described as the “oldest and best-known mechanical rights agency in the U.S.” (Id. ¶ 15.)

HFA's traditional business, in significant part, was providing licensing and rights administration services to the owners of copyrights in musical works-that is to say, licensors. (Doc. No. 436-1 ¶¶ 17-18.) Since at least 2001, however, HFA has provided services to licensees as well, although it did not begin serving digital streaming platforms until 2009. (Id. ¶ 18.) HFA has since provided rights administration services to Apple, YouTube, and Google, among others. (Id. ¶ 20.)

HFA is also known, in the music industry, for its operation and maintenance of large, publicly available databases of information related to copyrighted musical compositions. (Id. ¶¶ 21-22.) As the court has already noted, there has not historically been a definitive database of music publishing rights in the United States. There are, however, a handful of major private databases that parties widely use, including HFA's database of musical compositions, which includes information about the works, as well the corresponding copyright owners and administrators. As a matter of industry custom, publishers and administrators of musical works typically “register” their rights with HFA for inclusion in its database, which is updated daily. (Id. ¶¶ 23-24, 28.) Eight Mile Style concedes that owners and administrators submit registrations to HFA “in order to claim licensing authority over all or a portion of those musical compositions.” (Doc. No. 438-1 ¶ 11.) When all or a portion of a composition has not been claimed, HFA lists the composition as “in Copyright Control.” (Id. ¶ 16.)

Kobalt registered the EMS Compositions in HFA's database in its own name in 2009- which, at the time, was accurate.⁴ When the U.S. mechanical licensing rights for the compositions were reassigned to Bridgeport, however, neither Bridgeport nor Kobalt took steps to update the registration, and HFA's database listed Kobalt as the administrator of most of the EMS Compositions until 2019. (Doc. No. 436-1 ¶¶ 298-308; Doc. No. 427 ¶ 7.) Other works had been placed in Copyright Control, and still other registrations included stale information regarding Sony/ATV's rights to license the compositions. (Id. ¶¶ 214-16.) Eight Mile Style concedes, as a factual matter, that, “[w]hen a catalog of musical compositions is moved from one administrator to another administrator, usually the new administrator notifies HFA to update the registration of those compositions within HFA's records[,] and the prior administrator confirms the move.” (Doc. No. 514-1 ¶ 173.) Although Bridgeport was controlled by Eight Mile Style, it took no such steps.

4 As Kobalt points out, this is something of a simplification. In some instances, Kobalt was listed as a contact rather than as a publisher or administrator, and other parties were listed as well. (See Doc. No. 442 ¶ 203.) For the purposes of this case, however, it suffices that a significant number of the registrations directed parties to Kobalt as, seemingly, a party with licensing authority.

Eight Mile Style was aware that the Kobalt registrations were leading to some confusion and even internally discussed a potential “plan of attack” to address the issue in 2015, but it ultimately did not take any steps to correct the registrations. (Doc. No. 436-1 ¶¶ 214-15.) Eight Mile Style says that the reason for failing to update industry databases to reflect Bridgeport's role was practical: it believed that the databases in use at the time were not equipped to accurately represent situations in which one party had licensing authority but another party was empowered to receive royalties. Accordingly, Eight Mile Style argues, simply substituting Bridgeport for Kobalt in databases allegedly could have misled parties about which entity should be receiving royalty payments, causing a problem for Kobalt in its royalty collection responsibilities. HFA asserts that it was, in fact, capable of distinguishing between licensors and payees in its internal systems, but it admits that that distinction would not appear on HFA's song registration database, which would only have shown the licensing administrator. (See Doc. No. 346-17 at 12.) Accordingly, while Eight Mile Style may have been mistaken if it believed that updating HFA's information would have created issues with regard to royalties disbursed by HFA, its reasoning is plausible, insofar as other payors were relying on the HFA database or other databases that Eight Mile Style/Bridgeport failed to update.

In 2010, in advance of Spotify's U.S. launch, Spotify and HFA entered into an Administrative Services Agreement (“ASA”), pursuant to which HFA agreed to assist Spotify in obtaining mechanical licenses for compositions that Spotify desired to stream and to calculate royalties owed for Spotify's uses of those compositions. (Doc. No. 438-1 ¶¶ 21, 42.) HFA, among other things, agreed to perform track-to-work “matching” services by identifying musical works embodied within sound recordings and the holders of rights in those musical works. (Doc. No. 436-1 ¶ 370.) Spotify would send HFA electronic “request files” that included sound recording metadata, which HFA would run through bulk electronic matching software intended to identify possible matches between the information in Spotify's request file and HFA's databases. (Id. ¶¶ 371-73.) HFA began processing requests from Spotify in early 2011. (Doc. No. 514-1 ¶ 186.)

When HFA's matching efforts identified a rights holder, it would assist Spotify in obtaining a license, if one was necessary and available. Often, this took the form of sending NOIs. However, because HFA was itself an administrator of rights, sometimes the compositions identified were ones for which it already had the right to grant a license on behalf of the rights holder. For example, Sony/ATV-which had administered some of the EMS Compositions for a period that concluded prior to Spotify's launch-had an affiliate agreement with HFA that permitted HFA to issue licenses for Sony/ATV compositions. (Doc. No. 436-1 ¶¶ 386-88.) HFA documentation suggests that it granted Spotify licenses for 66 EMS Compositions pursuant to its authority related to Sony/ATV, but the dates of those ostensible licenses appear to fall outside the period in which Sony/ATV would have had any such authority. (See Doc. No. 325-10 at 26.)

HFA oversaw Spotify's distribution of royalties, including the payment of royalties based on Spotify's use of the Eight Mile Style Compositions. HFA based its royalty calculations on information provided by Spotify about the usage of the sound recordings on its service. (Doc. No. 436-1 ¶ 378.) Spotify would pay the royalties seemingly due to HFA, which would distribute them to the appropriate party-in the case of the EMS Compositions, Kobalt. (Id. ¶ 379.) HFA's royalty payments were typically accompanied by statements of account, including “detailed information concerning the royalties paid.” (Id. ¶ 380.) Kobalt, at its election, received royalties and royalty reports from HFA on a quarterly basis. (Id. ¶ 282.) Kobalt passed much of the royalty information that it received along to Eight Mile Style in the form of itemized royalty summaries that, among other things, identified specific payments as related to Spotify. (Id. ¶ 285.)

HFA was a recipient of Kobalt's 2011 LOD, in which it purported to have an “exclusive administration agreement” with Eight Mile Style. (Doc. No. 437 ¶ 184.) Because neither Bridgeport nor Eight Mile Style sent the Bridgeport LOD that had been drafted at around the same time, however, it received no notice, at the time, that U.S. mechanical licensing authority had actually been assigned to Bridgeport. Pursuant to HFA's policies, if it had received the Bridgeport LOD, it would have updated its records accordingly. (Id. ¶ 111.)

D. HFA's NOIs to Kobalt on Behalf of Spotify

From 2011 to 2017, HFA sent Kobalt thousands of NOIs concerning the EMS Compositions, despite the fact that mechanical licensing authority had been reassigned to Bridgeport. (Doc. No. 436-1 ¶ 261.) That number is higher than the number of compositions themselves, because, as this court explained in its Memorandum of July 8, 2024, pre-MLC Copyright Office regulations and music industry practice generally required NOIs to be directed at particular “phonorecords”-meaning that, when a new recording of a song was released or an old recording was included on a new album, it was typically necessary to file a new NOI.⁵ (See Doc. No. 688 at 27-28.)

5 For example, one of the EMS Compositions is a song entitled “Superman.” HFA sent its first NOI in connection with that song in 2011, based on the song's inclusion on the Eminem album “The Eminem Show.” In the ensuing years, however, HFA, sent dozens of other NOIs in connection with versions of the song released by artists including the Piano Tribute Conservatory, the Mantovani Orchestra, and the Lounge Brigade, as well as Eminem's own release of a deluxe version of the original album. (See Doc. No. 347-11 (native file on thumb drive).)

Eight Mile Style concedes that “Kobalt never told senders of NOIs for Eight Mile Works that Kobalt was not an authorized recipient of those NOIs.” (Doc. No. 436-1 ¶ 267.) HFA's reliance on the wrong recipient, however, was only the start of the problem with HFA's NOIs. Most strikingly, nearly all of the NOIs were obviously, facially untimely-often by a great deal of time. By Spotify's own telling, HFA's NOIs were timely with regard to only 18 of the 242 EMS Compositions at issue.⁶ (Id. ¶ 382.). When HFA sent an untimely NOI, it would include an expected date of distribution, or “EDD,” reflecting a date in the past-meaning that the NOI was faulty on its face, as long as one realized when it was transmitted. HFA says that it engaged in the practice of sending NOIs, even when clearly untimely, because those were the instructions it received from Spotify. (Doc. No. 438-1 ¶ 127.)

6 Those compositions are the songs “6 in the Morning,” “Census Bureau,” “Drips,” “Git Up,” “Hailie's Song,” “How Come,” “Moment of Clarity,” “My Band,” “Out on Bail,” “Puke,” “Rap Game,” “Renegade,” “Soldier” “Spit Shine,” “Square Dance,” “Superman,” “The Kiss,” and “White America.” (Doc. No. 325-10 at 25.)

Spotify concedes that “certain NOIs issued by HFA on Spotify's behalf contained an expected date of distribution that preceded the date of issuance (or ‘printed on' date) also reflected on such NOIs.” (Doc. No. 427 ¶ 55.) The record does not definitively explain why Spotify would engage in such a practice. Eight Mile Style suggests that the NOIs were intended to deceive rights holders into believing that licenses had been obtained. It is, however, not clear to the court that a facially untimely NOI would serve that purpose. Indeed, if anything, it would seem to run the risk of alerting the recipient that something was amiss. It is also possible that Spotify was trying to build a paper trail in the hopes of some eventual mechanism through which the untimeliness of the NOIs might be waived-either as part of a settlement or through remedial legislation-but that is purely speculation. In any event, the one thing that is clear about the untimely NOIs is that they were, under actually existing copyright law, legally ineffective and served no legitimate purpose.

E. Kobalt's Entering into the 2016 NMPA Settlement

The March 17, 2016 settlement agreement between Spotify and the NMPA permitted publishers to opt into the agreement in connection with compositions for which the publisher had the authority to grant the applicable release and/or covenant not to sue. (Doc. No. 437 ¶ 312.) Shortly after the settlement was finalized, Kobalt executed a Participating Publisher Pending and Unmatched Usage Agreement (“PPPUUA”) with Spotify, pursuant to which it opted into the settlement with regard to all compositions for which it had “legal and contractual authority to do so.” (Doc. No. 606-10 at 15.) Kobalt states, however, that that opting in did not encompass the EMS Compositions, because, by the time it opted in, it was no longer the U.S. mechanical licensing administrator for those compositions and did not otherwise possess any authority to enter into the NMPA settlement based on them. (Doc. No. 437 ¶ 316.) Kobalt did not provide Spotify a list of the compositions in connection with which it was opting in, and it concedes that, until 2019, it did not expressly exclude any particular client's works from its decision. (Id. ¶¶ 317-18.)

On September 21, 2016, moreover, Kobalt sent Spotify a data file for the purpose of identifying works covered by the settlement for which Kobalt was entitled to payment. That file included 168 of the EMS Compositions, although Kobalt maintains that it was never its intent to give the impression that it was claiming a right to enter into the settlement with regard to those compositions. (Id. ¶¶ 322-24.) Kobalt sent a similar transmission in connections with another claims period in July 2017. (Id. ¶ 331-333.)

Eight Mile Style and its affiliates did apparently receive some money associated with the settlement, albeit not a particularly large amount. Specifically, Kobalt credited a bit over $20,000 of the settlement funds it received to Bridgeport in connection with the EMS Compositions. (Id. ¶¶ 336-340.) Kobalt eventually discovered that it had inadvertently distributed settlement payments to some clients that it should not have, and it applied a debit against the accounts of those clients. (Id. ¶ 338.) Kobalt, however, did not apply a debit to Eight Mile Style's account. (Id. ¶ 337.) Kobalt says that it did not do so, because the payment to Eight Mile Style was proper, owing to the fact that Spotify itself had made a corresponding payment to Kobalt, despite the fact that there had been no release in connection with the EMS Compositions. (Id. ¶ 339.) The parties agree that Eight Mile Style did not contemporaneously realize that its statements from Kobalt included the settlement payment. (Id. ¶ 341.)

E. The 2016 Kobalt-Spotify Blanket Mechanical Licensing Agreement

On December 15, 2016, Spotify entered into a blanket mechanical licensing agreement with Kobalt (“2016 BMLA”), through which Kobalt granted Spotify “a non-exclusive, . . . irrevocable license throughout the [United States] to . . . reproduce and distribute” certain compositions identified as the “Publisher Compositions,” “as embodied in sound recordings on servers owned or controlled by Spotify.” (Doc. No. 265-1 at 2.) The “Publisher Compositions” were defined as “the Compositions or portions thereof that [Kobalt], whether now or during the Term [of the Agreement], owns, controls, or administers.” (Id.) Neither “administer” nor “control” is defined by the Agreement. (Id.)

The 2016 BMLA does not expressly list the compositions contemplated by the agreement, and it does not carve any compositions out of the license granted. Spotify had suggested, in an earlier draft, that Spotify provide a full list of its catalog as part of the agreement, but Kobalt rejected that suggestion, and the parties agree that it is not unusual for a large blanket license like the 2016 BMLA to omit such a listing. (Doc. No. 437 ¶¶ 346-48.) Kobalt's primary stated reason for that approach is that the catalog of a large administrator like Kobalt would be routinely changing, rendering any list almost immediately out of date regardless. (See Doc. No. 349 at 3-4.)

However, while it might have been impractical to list every work that was included, there would have been precedent if Kobalt had sought to identify certain specific works for which it performed administration functions, but which it was not licensing to Spotify. Specifically, Kobalt expressly excluded some works-including the EMS Compositions-from a 2013 blanket mechanical licensing agreement with YouTube. (Id. ¶¶ 249-50.)

As part of the 2016 BMLA, Spotify and Kobalt each made certain representations and warranties. Kobalt's representations and warranties included the following:

Publisher further represents and warrants to Spotify that (i) it has the right and authority to grant the rights to Spotify in this Agreement; (ii) Spotify's Use of the Publisher Compositions in accordance with the terms and conditions of this Agreement will not infringe any third party's proprietary or intellectual property rights and; (iii) except for payments made to third parties administering the public performance right and the applicable sound recording rights holders associated with the Publisher Compositions, no other payments, licenses, rights, authorities or permissions are necessary for Spotify to Use the Publisher Compositions.

(Doc. No. 265-1 at 7-8.) The representations and warranties were followed by an indemnification provision stating, in relevant part:

Each Party (the ‘Indemnifying Party') will indemnify and hold the other Party (the ‘Indemnified Party') harmless from any and all third party claims, damages, liabilities, costs and expenses (including reasonable legal expenses and counsel fees) (‘Claims') relating to any allegation, that, if true, would constitute a breach of the Indemnifying Party's representations, warranties, or covenants in [the preceding section of] this Agreement.

(Id. at 8.)

F. Other Key Actions and Communications Involving the Parties

It does not appear that high-level executives of the key companies in this litigation ever communicated with each other with sufficient clarity or detail to ensure that everyone involved knew what rights the companies, respectively, claimed and/or possessed. However, employees of the companies did interact with each other, and these interactions sometimes involved discussion of the parties' rights.

For example, on August 6, 2013, a data management agent for HFA, Collen Calabro, emailed Catlett with the following:

We recently received a lead that Bridgeport Music Inc may administer publishing for many Marshall Mathers titles in our system. On the attached, can you please provide a catalog breakdown and share for each title should you claim?

If you have any questions, please do not hesitate to contact me.

(Doc. No. 335-9 at 2.) Catlett forwarded HFA's inquiry to Martin and asked whether she should “tell [HFA] that Kobalt administers for our titles,” but Martin did not respond, and Catlett ultimately did not reply to Calabro's email. (Doc. No. 514-1 ¶ 221.)

Calabro followed up with another email on August 29, 2013, and Catlett again emailed Martin, asking if she should “say that no Eminem titles are administered by Bridgeport,” to which Martin replied, “Yes.” (Id. ¶ 222.) Ultimately, however, neither Catlett nor anyone else at Bridgeport or Eight Mile Style ever responded to this set of inquiries. (Doc. No. 437 ¶¶ 117-21; Doc. No. 514-1 ¶ 223.)

On November 14, 2016, Kobalt employee Bob Bruderman sent HFA “a list of [recordings] that we control the underlying composition to,” and that list included recordings embodying EMS Compositions. (Doc. No. 437 ¶ 188.) In this litigation, Bruderman testified that by “control” he meant “control the right to receive money for.” (Id.; see also Id. ¶¶ 327-29.)

Kobalt's and Eight Mile Style's statements suggesting that Kobalt administered or controlled the compositions, however, were not wholly uncontradicted. In at least three other email exchanges, Kobalt employees specifically informed HFA employees that Kobalt did not possess licensing authority for the EMS Compositions at the relevant time. In August 2013, for example, a Kobalt employee told an HFA employee that “[t]he 8 Mile Style catalog is no longer administered by Kobalt for licensing to my knowledge” and that, as she understood it, “Bridgeport is handling the licensing but should be directing licensee payments directly to Kobalt.” (Doc. No. 596-18.) Similar exchanges about either the catalog or “Lose Yourself” happened in March 2015, February 2016, and August 2018. (Doc. Nos. 596-19 to -21.)

G. Procedural History

The full procedural history of this case is far too lengthy to detail here, but the court will provide the basics. Eight Mile Style filed its initial Complaint on August 21, 2019. (Doc. No. 1.) That Complaint named Spotify as the sole defendant, and Eight Mile Style stated only one count, for direct copyright infringement-although that one count potentially encompassed billions of discrete instances of infringement. (Id. ¶¶ 76-86.) On October 21, 2019, Spotify provided notice of this litigation to Kobalt. (Doc. No. 437 ¶ 367.) On October 29, 2019,⁷ Kobalt denied the indemnification request. (Id. ¶ 368; Doc. No. 342-8.)

7 Kobalt's admission regarding this letter lists the date as October 29, 2021, but the text of the letter confirms that it was sent in 2019 and that “2021” is a typo. (See Doc. No. 342-8.)

On May 29, 2020, Spotify filed a Third-Party Complaint against Kobalt. The Third-Party Complaint states five causes of action. Spotify's Count I is for breach of contract in connection with the 2016 BMLA. Its Count II is for contractual indemnity, and its Count III is for anticipatory repudiation in connection with those rights. Spotify's Count IV is for breach of contract in connection with Kobalt's opting into the NMPA settlement through the PPPUUA, and its Count V is for negligent and/or intentional misrepresentation. (Doc. No. 93 ¶¶ 74-102.)

On July 1, 2020, Eight Mile Style filed an Amended Complaint, in which it added HFA as a defendant. In addition to the direct infringement claim against Spotify-which Eight Mile Style now designated as Count I-Eight Mile Style stated claims against HFA for contributory infringement (Count II) and vicarious infringement (Count III). (Doc. No. 97 ¶¶ 125-34.) HFA moved to have the claims against it dismissed, and, on April 22, 2021, the court granted that request with regard to vicarious infringement but denied it with regard to contributory infringement. (Doc. No. 165 at 17.)

Eight Mile Style filed no other superseding or supplemental complaints, and each party now seeks summary judgment. The court heard oral arguments regarding the motions on July 12, 2024.

II. LEGAL STANDARD

Rule 56 requires the court to grant a motion for summary judgment if “the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed.R.Civ.P. 56(a). To win summary judgment as to the claim of an adverse party, a moving defendant must show that there is no genuine issue of material fact as to at least one essential element of the plaintiff's claim. Once the moving defendant makes its initial showing, the burden shifts to the plaintiff to provide evidence beyond the pleadings, “set[ting] forth specific facts showing that there is a genuine issue for trial.” Moldowan v. City of Warren, 578 F.3d 351, 374 (6th Cir. 2009); see also Celotex Corp. v. Catrett, 477 U.S. 317, 322-23 (1986). Conversely, to win summary judgment as to its own claims, a moving plaintiff must demonstrate that no genuine issue of material fact exists as to all essential elements of her claims. “In evaluating the evidence, the court must draw all inferences in the light most favorable to the non-moving party.” Moldowan, 578 F.3d at 374 (citing Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 587 (1986)).

At this stage, “the judge's function is not . . . to weigh the evidence and determine the truth of the matter, but to determine whether there is a genuine issue for trial.” Id. (quoting Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249 (1986)). But “[t]he mere existence of a scintilla of evidence in support of the [non-moving party's] position will be insufficient,” and the party's proof must be more than “merely colorable.” Anderson, 477 U.S. 242, at 252. An issue of fact is “genuine” only if a reasonable jury could find for the non-moving party. Moldowan, 578 F.3d at 374 (citing Anderson, 477 U.S. at 252).

III. ANALYSIS

A. Actual or Apparent Authority Under the 2016 BMLA

Spotify argues that nearly all of the questions raised by this litigation are ultimately beside the point, because, by the time that most of the streams for which timely infringement claims would still exist occurred, Spotify had already fully licensed the EMS Compositions through Kobalt pursuant to the 2016 BMLA. Neither Kobalt nor Eight Mile Style appears to dispute that, if Kobalt had had the power to grant such a license, the 2016 BMLA would have done so, because the EMS Compositions, as works within Kobalt's library available for U.S. mechanical licensing, would have been among the “Publisher Compositions.” Kobalt and Eight Mile Style argue, however, that Kobalt had no such authority and the EMS Compositions, therefore, were not “administered” or “controlled” by Kobalt. Spotify disputes that reading, which the court will consider, in greater detail, in connection with Spotify's claim for indemnification. For the purposes of Eight Mile Style's infringement claims, however, what matters is whether Kobalt had either actual or apparent authority to license the compositions to Spotify. Without such authority, the 2016 BMLA could not have conveyed a mechanical license to the EMS Compositions, regardless of whether that had been the parties' intent.

1. Actual Authority

“Actual authority granted to an agent to bind [its] principal is created by direct manifestations from the principal to the agent, and the extent of the agent's actual authority is interpreted in the light of all the circumstances attending these manifestations, including the customs of business, the subject matter, any formal agreement between the parties, and the facts of which both parties are aware.” Matter of Elizabeth T., 214 A.D.3d 815, 817, 186 N.Y.S.3d 264, 266 (2023) (quoting N.Y. Cmty. Bank v. Woodhaven Assocs., LLC, 137 A.D.3d 1231, 1233, 29 N.Y.S.3d 377, 379 (2016)).⁸ Spotify argues that Eight Mile Style's dealings with Kobalt resulted in actual authority, because, despite the formal granting of rights to Bridgeport, Kobalt remained the party that actually did most of the work associated with licensing the EMS Compositions.

8 The parties agree that the underlying agreements are governed by New York Law.

Spotify's argument, in other words, is that Kobalt possessed nearly every role and responsibility that a U.S. mechanical licensing administrator would ordinarily have-merely without any formal authorization to be the party granting such a license. That may be a fair characterization of the facts, but, as a legal argument for actual authority, it fails on its face. The key word in “actual authority,” in this instance, is “actual.” If Kobalt had everything short of actual authority, then that is just another way of saying that, precisely as Eight Mile Style contends, it did not have actual authority.

The formal agreements at issue are unambiguous. Actual authority to enter into mechanical licenses for the EMS Compositions was withdrawn from Kobalt and granted to Bridgeport in 2011. Although Kobalt appears to have exceeded its authority, at times, during the first two years of that new status quo, there is no evidence that music industry custom would treat any of the underlying dealings as sufficient to negate the written allocation of rights to which Kobalt and Eight Mile Style agreed. There is, therefore, no basis for finding actual authority.

2. Apparent Authority

“[T]he existence of apparent authority depends upon a factual showing that the third party relied upon the misrepresentation of the agent because of some misleading conduct on the part of the principal-not the agent. Moreover, a third party with whom the agent deals may rely on an appearance of authority only to the extent that such reliance is reasonable.” Lisi v. New York Ctr. for Rehab. & Nursing, 225 A.D.3d 590, 591, 206 N.Y.S.3d 688, 690 (2024) (quoting Hallock v. State, 64 N.Y.2d 224, 231, 474 N.E.2d 1178, 1181 (1984)). Accordingly, while actual authority can be established solely based on the actions of the purported agent and principal, a finding of apparent authority requires a showing of sufficient facts regarding three parties: the putative principal, the putative agent, and the party asserting apparent authority in litigation. Even if a putative principal said or did things that could have given rise to apparent authority, that apparent authority can only be asserted by someone who actually relied on it.

That principle is fatal to Spotify's argument for apparent authority. There is no evidence in the record that Spotify believed, in 2016, that Kobalt was licensing it the EMS Compositions. Indeed, despite the substantial resources that have been expended on this litigation, Spotify has failed to identify a single employee or ex-employee who would testify that he or she understood the 2016 BMLA to include the EMS Compositions when the agreement was signed. (See Doc. No. 413 ¶¶ 48-57.) Spotify was already streaming those compositions before the 2016 BMLA and continued its approach, unchanged, afterwards. Kobalt had a valuable library,⁹ even without the EMS Compositions, and the available evidence suggests that Spotify obtained a blanket license from Kobalt, not because it specifically wanted the rights to any particular songs, but because it was executing a strategy of trying to obtain as many sets of rights as possible- possibly a wise approach for a streaming service, but difficult to square with any showing of reliance with regard to any particular composition.

9 According to Kobalt, the number of compositions in its catalog, as of December 15, 2016, was higher than 450,000. (Doc. No. 509 ¶ 95.)

Internal communications between Spotify personnel suggest that Spotify did eventually come to believe that Kobalt administered the EMS Compositions-in 2018, well after the 2016 BMLA had been finalized. (See Doc. No. 344-9 at 2.) That later-arising misunderstanding, however, cannot retroactively grant Kobalt apparent authority in 2016. Without the necessary showing of reliance, Spotify cannot rely on a theory of apparent authority to overcome Kobalt's straightforward lack of actual authority to mechanically license the EMS Compositions within the United States.

B. Other Potential Sources of an Express License

Spotify has identified a handful of other ways that, it argues, it validly, expressly licensed some or all of the EMS Compositions: (1) in connection with the NMPA Settlement Agreement; (2) through dealings with Sony/ATV; (3) through licensing agreements with HFA; and (4) through NOIs that HFA sent in connection with the EMS Compositions.

None of these potential sources of rights provides a persuasive ground for ruling in Spotify's favor. It does not appear that Kobalt actually had the authority to enter into the NMPA Settlement on behalf of Eight Mile Style, and the agreement through which Kobalt opted into the settlement explicitly carved out works for which it had no such authority. Even if Eight Mile Style had entered into the settlement, moreover, it would, at most, have resulted in a release of claims that accrued prior to June 30, 2017-leaving much of this case intact. (See Doc. No. 321 at 42.)

Sony/ATV and HFA, like Kobalt, did not have mechanical licensing power on behalf of Eight Mile Style at the relevant times. (See Doc. No. 427 ¶ 6.) Spotify suggests that HFA had the right to license some of the compositions on behalf of co-owners other than Eight Mile Style, but it has not established that those rights were sufficient to result in full mechanical licensing authority. Finally, Spotify itself admits that nearly all of its NOIs were untimely and/or sent to the wrong party. Spotify, accordingly, had no express license to use at least most of the EMS Compositions.

C. Implied License

Based on the foregoing, there is no basis on which a reasonable finder of fact could conclude that Spotify had an express license to transmit recorded performances of the EMS Compositions to its users. It is undisputed, however, that that is what Spotify did-billions of times. Spotify's streaming of the compositions, therefore, appears, on its face, to have been infringing. Spotify argues, however, that there is one final way that its actions may have been non-infringing: because Eight Mile Style's acceptance of royalty payments for Spotify streams, without protest or complaint, gave rise to an implied license to continue.

The party claiming the existence of an implied license has the burden of proving that such a license arose. See Bourne v. Walt Disney Co., 68 F.3d 621, 631 (2nd Cir.1995). There is no precise formula for determining whether an implied license exists; rather, the court is to examine the “intentions” of the parties from the “totality of the circumstances.” Jeffrey A. Grusenmeyer & Associates, Inc. v. Davison, Smith & Certo Architects, Inc., 212 Fed.Appx. 510, 514 (6th Cir.2007). That is, the court should explore, based on the facts and the circumstances of the individual case, whether the evidence supports the notion that the parties, in essence, made an agreement permitting the defendant to use the work, consistent with certain understandings or terms. Johnson v. Jones, 149 F.3d 494, 500 (6th Cir. 1998). The key question is whether the facts and circumstances demonstrate that “the copyright owners intended that their copyrighted works be used in the manner in which they were eventually used.” Id. at 502. Spotify concedes, in its own briefing, that such a license could not arise unless Eight Mile Style intended for it to. (Doc. No. 321 at 47.)

There is, however, a considerable obstacle to any conclusion that Eight Mile Style intended to grant Spotify a mechanical license at the standard compulsory rate: the fact that doing so would have been extraordinarily foolish from a business perspective. Why would any party in Eight Mile Style's position grant a free license at the compulsory licensing rate, subject to no concessions, despite the fact that Spotify missed the window for locking that rate in without the consent of the copyright owner? There is evidence suggesting that it was not common for parties who negotiated licenses to receive royalty rates that were, in and of themselves, more generous than the statutory rate. (E.g., Doc. No. 422-5 at 49.) That does not, however, mean that there was nothing to be gained by negotiating. Eight Mile Style could have demanded a large, one-time payment for a voluntary license at the customary rate, or it could have negotiated for nonmonetary concessions, such as free promotion of Mathers' work through Spotify's platform-which would have led to more streams and more money for Eight Mile Style. Eight Mile Style could not have plausibly intended to grant a free, implied license at the standard rate unless it was either very foolish or wanted, for some reason, to do Spotify a favor.

Spotify protests that Eight Mile Style may have intended to grant an implied license because it wanted to continue to receive the ostensible royalty payments that Spotify was sending. At most, though, that would be an argument for estoppel, not an argument for finding an implied license. It may well be that Eight Mile Style was happy to receive the royalty checks, as opposed to no checks at all. But the same rationale that would favor accepting the payments- that is, basic economic self-interest-would also preclude any assumption that Eight Mile Style intended to relinquish its rights in so doing.

Spotify attempts to overcome this argument by analogizing these facts to other cases in which implied licenses were found. See, e.g., Parker v. Yahoo!, Inc., No. CIV.A. 07-2757, 2008 WL 4410095, at *4 (E.D. Pa. Sept. 25, 2008); Experexchange, Inc. v. Doculex, Inc., No. C-08-03875 JCS, 2009 WL 3837275, at *24 (N.D. Cal. Nov. 16, 2009); Keane Dealer Servs., Inc. v. Harts, 968 F.Supp. 944, 947 (S.D.N.Y. 1997). Those courts, however, considered the matters before them in the context of the specific business relationships at issue, just as the court has done here. Insofar as courts have, at times, suggested that certain contextual factors necessarily “give[] rise to an implied license as a matter of law,” UMG Recordings, Inc. v. Disco Azteca Distributors, Inc., 446 F.Supp.2d 1164, 1178 (E.D. Cal. 2006), such an approach would be inconsistent with the approach of this circuit.

Superficial similarities between cases cannot change the fact that, under Sixth Circuit law, Eight Mile Style could not have granted an implied license unless it actually meant to do so. There is no direct evidence of that intent, and every available piece of evidence regarding Eight Mile Style's approach to managing its assets weighs against assuming that the company would grant such a windfall to Spotify. The only argument available to Spotify, then, is that the court simply must infer the requisite intent from the fact that Eight Mile Style accepted Spotify's royalty payments. There is, however, no dispute that Spotify's streaming of the EMS Compositions resulted in its owing Eight Mile Style at least as much as it paid; the disagreement, in this case, is about whether Spotify owes more. Eight Mile Style's continuing receipt of “royalties” for the EMS Compositions, without alerting Spotify to the fact that no license was in place and Eight Mile Style would eventually come back for more, may reasonably be viewed as unfair or misleading. Arguments of that sort, however, are relevant to the issue of estoppel, which the court will discuss in the next section-not to whether an implied license arose. An unfair reality is still reality, and the court has no power to ascribe an intent to Eight Mile Style that all of the available evidence suggests it did not have. The court, therefore, has no basis for finding an implied license.

D. Estoppel

Infringement is fundamentally a strict liability cause of action, and it does not require the plaintiff to establish very much-only “‘(1) ownership of a valid copyright, and (2) copying of constituent elements of the work that are original'” by the defendant without a right to do so. ATC Distribution Grp., Inc. v. Whatever It Takes Transmissions & Parts, Inc., 402 F.3d 700, 705 (6th Cir. 2005) (quoting Feist Publ'ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 361 (1991)). Based on the foregoing analysis, Spotify lacked any valid license-compulsory or negotiated, express or implied-to include recordings of the Eight Mile Style Works in its streaming catalog, but it did so anyway. Eight Mile Style, therefore, appears to have established a strong prima facie case of copyright infringement and is entitled to summary judgment unless Spotify can avail itself of some valid affirmative defense.

As Spotify points out, however, there is a well-established copyright defense that is, on its face, an obvious candidate for potential application in this case: equitable estoppel. Equitable estoppel, broadly speaking, prevents a party from enforcing a right “when he has ‘by his representations or his conduct induced the other party . . . to give him an advantage which it would be against equity and good conscience for him to assert.'” Trustees of Mich. Laborers' Health Care Fund v. Gibbons, 209 F.3d 587, 591 (6th Cir. 2000) (quoting Union Mut. Ins. Co. v. Wilkinson, 80 U.S. 222, 233 (1871). The Supreme Court has expressly confirmed that, while the Copyright Act's statute of limitations precludes the application of the equitable doctrine of laches, it leaves intact the defense of estoppel, the “gravamen” of which is not delay alone, but the “consequent loss” to the defendant from the “misleading” actions of the plaintiff. Petrella, 572 U.S. at 685 (citing Wehrman v. Conklin, 155 U.S. 314, 327 (1894) 6 Patry on Copyright § 20:58 (2013).).

In the licensing context, equitable estoppel addresses considerations that are parallel to, but distinct from, those raised by the doctrine of implied license. An implied license arises when a rights holder consents to a particular use but does not make that consent explicit. Estoppel, on the other hand, arises when a rights holder goes out of its way not to consent to a use, but also not to stop it-choosing, instead, to patiently accumulate leverage pursuant to the Copyright Act's generous damages provisions and to file suit when most advantageous. See Soc. Sci. Hist. Ass'n v. Duke Univ., 31 F.Supp.3d 781, 788 n.1 (E.D. N.C. 2014). Accordingly, while Eight Mile Style's established intent is incompatible with a finding of an implied license, that same intent, applied to the same facts, fits easily within the framework of an equitable estoppel defense.

There are a number of different formulations of what a defendant must establish in order to succeed on a defense of equitable estoppel in the copyright context, and, while those formulations differ in their details, they largely look at the same fundamental set of issues. See 6 Patry on Copyright § 20:58 (2024). Generally speaking, a copyright infringement defendant who wishes to rely on equitable estoppel based on the plaintiff's inequitable conduct must establish “four conjunctive elements ....: (1) the plaintiff must know the facts of the defendant's infringing conduct; (2) the plaintiff must intend that its conduct shall be acted on or must so act that the defendant has a right to believe that it is so intended; (3) the defendant must be ignorant of the true facts; and (4) the defendant must rely on the plaintiff's conduct to its injury.” Carson v. Dynegy, Inc., 344 F.3d 446, 453 (5th Cir. 2003) (citing Nimmer on Copyright § 13.07 (2002)); accord Clever Factory, Inc. v. Kingsbridge Int'l, Inc., No. 3:11-1187, 2013 WL 5375258, at *5 (M.D. Tenn. Sept. 24, 2013) (Griffin, M.J.). Spotify argues that all four elements are present here.

1. Eight Mile Style's Knowledge of Spotify's Infringement

It is undisputed that Eight Mile Style was aware of Spotify's streaming of the EMS Compositions for several years before taking any affirmative steps to assert its rights. Eight Mile Style suggests that it was nevertheless ignorant of Spotify's lack of a license, at least for most of that time. Eight Mile Style's history of managing its rights, however, renders that explanation implausible. For nearly the entirety of Spotify's time as an active service in the United States, the mechanical licensing rights for the EMS Compositions were controlled, in the U.S., by Bridgeport, an entity over which Eight Mile Style had essentially complete day-to-day control. Spotify theoretically might have gotten the rights earlier, in anticipation of launching in the U.S., from either Kobalt or, for some compositions, Sony/ATV. Eight Mile Style's relationships with those companies, however, were marked by consistent communication and documentation. The idea that Eight Mile Style would simply be unaware of whether those old counterparties had licensed its valuable compositions to the world's largest streaming service-and then fail, for nearly a decade, to actually look into the matter-is beyond what can reasonably be believed without some solid explanation.

Eight Mile Style, however, has no such explanation. Martin has stated, in a Declaration, that, until 2018, he “believed that Spotify had obtained valid licenses via a pass-through license of Universal Music Group (‘UMG') at the statutory rate”-although the parties now agree that no such license existed. (Doc. No. 659-2 ¶ 10.) However, Martin's deposition testimony-which goes into the issue in significantly greater detail-cannot be reconciled with the cursory statement in his Declaration. Martin testified that, historically, UMG, as Eminem's record label, was “passing through licenses to third parties, and we sued them. There was a big public lawsuit. And there was a decision at some point that they were going to come directly to us if there was a license after 2012.” (Doc. No. 416-15 at 445.) A copy of the October 19, 2012 settlement agreement between EMS and UMG confirms that, if UMG ever had a right to issue licenses to the EMS Compositions without Eight Mile Style's participation, it relinquished that right as part of the settlement. (Doc. No. 428-9 at 3.) UMG, therefore, could not have granted a license to Spotify without Eight Mile Style's knowledge after that settlement agreement was finalized on October 19, 2012. That leaves open the possibility that a license could have been granted in 2011 or 2012, but Martin also testified that any such license would have been rendered ineffective relatively early in the period covered by the pending motions. According to Martin, his understanding was that, “at or around 2016[,] . . . Spotify would have the obligation to come to us for a license, because there would have been a termination” of any pass-through license that existed. (Doc. No. 329-9 at 508.) The settlement agreement confirms that the restrictions placed on new licenses also applied to renewals of previously issued licenses. (Doc. Nio. 428-9 at 3.)

The court finds it highly implausible that Martin was ever in the dark regarding whether UMG licensed the EMS Compositions to Spotify, unless he affirmatively chose to remain ignorant on the matter. By Martin's own account, there was a “big public lawsuit” over UMG's contested authority to issue such licenses. See, e.g., F.B.T. Prods., LLC v. Aftermath Recs., 621 F.3d 958, 962 (9th Cir. 2010); F.B.T. Prods., LLC v. Aftermath Recs., 827 F.Supp.2d 1092, 1095 (C.D. Cal. 2011). That litigation called for a searching, detailed dissection of what rights UMG had regarding the compositions at issue, and it is difficult to believe that those matters were litigated for as long as they were, and as aggressively as they were, without Eight Mile Style's ever learning the basic fact of whether UMG had granted mechanical licenses to Spotify. The settlement between UMG and Eight Mile Style, moreover, required UMG to send quarterly statements and supplemental reports regarding its use of the EMS Compositions, and that data was specifically required to identify the streaming service at issue. (Doc. No. 428-9 at 5-7.) It is not clear to the court, then, how Eight Mile Style could have failed to notice that UMG had not licensed the works to Spotify.

Even taking Martin at his word regarding his initial false belief that a passthrough license had been issued by UMG, however, he also testified that he believed that the relevant rights reverted to Eight Mile Style around 2016. He has tried to revise that date back by a year or two, but, even then, it would still be the case that Eight Mile Style has affirmatively admitted, in this litigation, that it “began to investigate the licensing status of the Compositions” no later than January 2018-over eighteen months before filing suit or taking any kind of direct action to rectify the situation. (Doc. No. 514-1 ¶ 240.) Eight Mile Style's own account, therefore, still supports the conclusion that there was a substantial period of time in which it knew that UMG had not licensed the compositions yet did not act to assert its rights either to Spotify or in court.

Eight Mile Style's other primary excuse for its long failure to act is that, “based on HFA's near century long reputation, Eight Mile Style reasonably believed that the receipt of royalties at the statutory rate meant that there must have been licenses in place.” (Doc. No. 436 at 23.) That argument, however, is impossible to reconcile with the fact that Eight Mile Style was fully aware that HFA's database incorrectly listed Kobalt, not Bridgeport, as the administrator of most of the works and had placed other of the works in Copyright Control. Eight Mile Style has explained that it kept the works' registrations under Kobalt's name in order to facilitate royalty collection, and maybe that is true. That decision, however, makes it impossible to believe that anyone at Eight Mile Style could have thought that HFA knew more about Eight Mile Style's rights than it did.

Licensing can be complicated, and a reasonable finder of fact might be persuaded that an isolated individual composer or a small music publishing house inadvertently allowed its compositions to be streamed unlawfully for eight years, because it mistakenly assumed that payments from HFA meant that the compositions were duly licensed. Eminem songs, though, are big business. The 2012 Martin Affiliated prospectus valued that company's publishing catalog, alone, at well into the millions of dollars. (See Doc. No. 375-4 at 5.) That same prospectus boasted that Eminem's status as an “iconic musical and cultural figure” made Martin Affiliated “well-positioned to maintain and even grow revenues” through the period of “changes taking place in the larger music industry.” (Id. at 4.) It further touted Martin's years of “experience in profitably managing music-related assets” and “well-earned reputation for his protection of his music assets, thereby safeguarding their value as well as generating revenue from them.” (Id. at 4-5.) The idea that a multimillion-dollar operation like Eight Mile Style would, for years, simply not bother to do the rudimentary legwork to find out whether its core assets were being wrongfully exploited on the scale at issue here is close to unthinkable. That implausible proposition is not salvaged by the suggestion that Eight Mile Style did not investigate further because it was trusting HFA-which Eight Mile Style specifically knew to have incorrect rights information in its database that Eight Mile Style had made a conscious choice not to correct.

The court concludes, therefore, that a reasonable finder of fact would have to find that Eight Mile Style was aware of Spotify's infringement long before it filed suit. Eight Mile Style knew that neither it nor Bridgeport had ever granted a license. It knew that it had assigned its U.S. mechanical licensing rights to Bridgeport in an unusual agreement that it had not meaningfully publicized to others in the music industry-making confusion likely. It knew that HFA, which was handling these matters for Spotify, did not acknowledge Bridgeport as the works' administrator in its database. It knew that other parties, as well, were confused about who controlled the compositions. And it knew that Spotify was streaming the compositions, despite Eight Mile Style's having no documentation whatsoever evidencing an effective license. The only reasonable explanation for this evidence is that Eight Mile Style either (1) affirmatively knew that Spotify was committing infringement or (2) was aware that Spotify was probably committing infringement and chose not to investigate further because it preferred for the infringement to continue. Either possibility is consistent with an application of equitable estoppel.

2. Eight Mile Style's Intent and Culpable Action

Even when a party knowingly and intentionally sleeps on its rights, the court cannot simply assume an improper intent for the purposes of estoppel. Music publishers, generally speaking, are in the business of making money, not abstractly vindicating the copyright regime, and there are often legitimate reasons not to pursue enforcement of rights every time that the opportunity comes around. See, e.g., Tolliver v. McCants, 684 F.Supp.2d 343, 349 (S.D.N.Y. 2010) (“Plaintiff's testimony indicates that the decision not to sue for royalties may simply be a business judgment that the cost of litigating outweighs any potential recovery, which does not implicate ownership rights.”). The fact that a copyright owner chose not to enforce its rights at a particular time, therefore, does not necessarily establish that it had an improper reason for doing so. Sometimes, sound business judgment suggests that combating infringement costs more than tolerating it. If that happened here, there would be no basis for applying estoppel.

Eight Mile Style, however, has not identified any valid business reason for tolerating Spotify's infringement, and the court can discern none. After all, Spotify was no minor or isolated infringer. The parties' evidence plainly establishes that the years in which the events giving rise to this litigation occurred saw streaming services like Spotify rise to a central place- probably the central place-in the business of selling music to individual consumers. Eminem was, moreover, a star with the kind of stature that would have given Eight Mile Style an enviable position in any license negotiations. From a business perspective, simply allowing Spotify to stream the compositions at the compulsory license rate, despite Spotify's not having been entitled to either the streaming or to that rate, would have likely been the worst thing Eight Mile Style could have possibly done. The court cannot conceive of a legitimate business purpose for such a decision.

When one considers the possibility of opportunistic litigation, however, Eight Mile Style's otherwise confounding behavior makes sense. If Eight Mile Style had simply sent a single, clear cease-and-desist letter-as countless other rights holders, many far less well-resourced than Eight Mile Style, do every year-that might have put an end to the ongoing infringement and forced Spotify into negotiations. Whatever payout those negotiations netted, however, would almost certainly have been lower than damages under the Copyright Act- which would entitle Eight Mile Style not only to compensation for its own loss but a chunk of Spotify's profits or statutory damages. There is, moreover, little doubt that Eight Mile Style was aware of the money to be made as a victim of infringement. Its 2012 prospectus specifically touted its likelihood of receiving millions of dollars in damages “either by award or settlement” in the UMG matter, and Eight Mile Style admits that it has “initiated copyright infringement actions approximately eight to ten times.” (Doc. No. 375-4 at 3; Doc. No. 438-1 ¶ 62.) Accordingly, while there is no direct evidence of why Eight Mile Style allowed this situation to persist, a desire to maximize its eventual remedies against Spotify appears to be the only option that makes sense.

The most potentially convincing argument that Eight Mile Style's delay in suing Spotify was not strategic is based on the fact that Eight Mile Style did not bring its claims until after the MMA liability limitation became available to Spotify. Why, one can reasonably ask, would Eight Mile Style choose not to file its claims before the MMA limitation went into effect? Eight Mile Style's actual prosecution of this litigation, however, sheds some possible light on why the strategic considerations for a litigant in Eight Mile Style's position might be more complicated than just avoiding a potential liability limitation-particularly with regard to the parties' respective hands in any settlement negotiations. If Eight Mile Style had sued Spotify prior to the MMA's applicability, it would have had a good chance of receiving potentially massive damages for infringement. By filing after the MMA's effective date, Eight Mile Style has a reduced chance of receiving those full damages, but it has gained something else: status as what appears likely to be the most well-funded, well-positioned party to challenge the constitutionality of one of the key aspects of the MMA, the liability limitation. The MMA framework was the culmination of what may have been one of the most high-stakes policymaking efforts in the history of copyright,¹⁰ and whether that framework survives has implications for the economy of music that go far beyond the rights of any individual artist, even a popular one like Eminem. A lawsuit that imperiled the MMA could cost Spotify a great deal more than any one artist could ever claim-and could, potentially, justify a more generous settlement. The timing of the lawsuit, therefore, in no way negates the possibility of a strategic delay, because there would be valid strategic arguments both for filing before the MMA liability limitation went into effect or for waiting and teeing up a constitutional showdown.

10 See U.S. Copyright Office, The Creation of the Music Modernization Act, available at https://www.copyright.gov/music-modernization/creation.htm (detailing process of drafting and adopting the Act, which “represent[ed] years of dedicated work by members of Congress, as well as negotiation and cooperation between a wide range of creators and business interests,” including “representatives from the creative communities, music industry, and digital music services”); see also Tanner J. Kramp, Rage Against the Machine: Why the Music Modernization Act Is but the First Step in Musicians' Battle to Reclaim the Value of Their Works, 64 B.C. L. Rev. 219, 233-37 & n.97-133 (2023).

All of the foregoing establishes that a reasonable finder of fact would have to conclude that Eight Mile Style chose not to resolve the problems surrounding the presence of its compositions on Spotify for strategic reasons. Even that, however, would not alone be sufficient to support estoppel, because estoppel depends not only on intent, but also action; that is, the defendant must show that the plaintiff intended the defendant to rely on the plaintiff's own statements and/or conduct. Watermark Publishers v. High Tech. Sys. Inc., No. 95-3839-IEG (CGA), 1997 WL 717677, at *8 (S.D. Cal. Aug. 12, 1997). There is, however, no absence of culpable conduct by Eight Mile Style here.

Eight Mile Style entered into an unusual arrangement whereby its valuable U.S. mechanical licensing rights were assigned to a third party that it controlled, after which it took none of the ordinary, customary steps to inform third parties of that change. Then, for a period of years, it continued to openly receive license requests through its old U.S. mechanical licensing administrator, Kobalt, which also publicly performed the most prominent recurring function that one would expect an administrator to perform-collecting royalties. Internal discussions confirm that Eight Mile Style was fully aware of the understandable confusion that its actions had fostered, but still it did nothing-did not update databases, did not send a belated LOD, and did not send Spotify a single cease-and-desist letter. The record provides no reason to doubt that, if Eight Mile Style had come forward to contest the status quo, it would have brought this situation to a much quicker end, but it did not. The only plausible reason for this course of action is that, when a rights holder has sufficient litigation resources-which Eight Mile Style did and does- allowing infringement to continue on a large scale is more economically beneficial to the purported victim than the licit streaming economy would be.

3. Spotify's Knowledge

Eight Mile Style, of course, was not the only sophisticated party that theoretically could have done more to rectify this situation. Spotify, as well, had every reason to try to get its rights in order, at least in the long term. Unlike Eight Mile Style, however, Spotify had no direct knowledge of the unique breakdown of rights in the EMS Compositions that had occurred behind the scenes. Music industry practice, moreover, makes it surprisingly plausible that Spotify might be genuinely confused, at times, regarding which rights it possessed and which it did not. By all accounts, it was in the practice of licensing catalogs without knowing, with any specificity, what was in them. (See, e.g., Doc. No. 420-8 at 25 (discussing blanket license practices); Doc. No. 439-20 at 8 (same); Doc. No. 439-19 at 14 (same).)

Eight Mile Style nevertheless argues that Spotify was on notice of Eight Mile Style's rights, and it points to the emails HFA received in which Kobalt informed it that it did not possess U.S. mechanical licensing authority for the EMS Compositions. Although HFA is distinct from Spotify, the “general rule is that notice to or knowledge of an agent may be imputed to the principal . . . where the agent is acting within the scope of his authority and the knowledge pertains to matters within the scope of the agent's authority.” Ford Motor Credit Co. v. Weaver, 680 F.2d 451, 457 (6th Cir. 1982) (citing Curtis, Collins & Holbrook Co. v. United States, 262 U.S. (1923); Anderson v. Gen. Am. Life Ins. Co., 141 F.2d 898 (6th Cir. 1944)).

The parties disagree regarding whether HFA was, as a matter of law, an agent of Spotify. Even if it was, however, the isolated communications that Eight Mile Style has identified would not be sufficient to establish that Spotify knew the “true facts” of the situation and would, therefore, be prevented from asserting estoppel. None of the cited communications actually informed HFA that Spotify lacked a license for the EMS Compositions. At most, the emails alerted HFA to the fact that, at those particular times, Kobalt did not have U.S. mechanical licensing authority. That information, however, did not necessarily establish that Spotify was without rights altogether. As Kobalt's expert Clark Miller testified, “in any catalog, in any day, in any moment, songs are coming and going.” (Doc. No. 329-5 at 120.) An email informing HFA or Spotify that Kobalt lacked U.S. mechanical licensing authority on any given date would not establish that its past licenses were necessarily ineffective, nor would it preclude the possibility that the relevant compositions were covered by some non-Kobalt license.

As numerous experts have now explained, moreover, music industry practice calls on publishers to inform key parties of changes in the allocation of rights through clear, formal LODs. (Doc. No. 442 ¶ 110.) The emails at issue, in contrast, were largely informal communications between ordinary employees related to discrete issues at particular times. The emails, accordingly, do not preclude a finding of estoppel.

It is undeniable that some of Spotify's actions suggest that it knew that its licensing of some works-including the EMS Compositions-may have been on shaky footing. For example, while Spotify's practice of sending facially ineffective NOIs is difficult to justify by any standard, the practice would make especially little sense if Spotify had been confident in its rights. The evidence now before the court, however, suggests that Spotify's uncertainty was a general, inevitable feature of the way that it had assembled licenses-not evidence of its awareness of any particular problem related to the EMS Compositions. Spotify's reliance on blanket mechanical licenses without binding lists of the works at issue meant that Spotify could never know, for sure, whether it had the rights to any work for which it had not either (1) obtained an express voluntary license specifically enumerating the work or (2) filed a timely NOI-an option that became impossible the moment that Spotify first streamed the track at issue. Spotify has stated, in this litigation, that this reality led it to adopt a “belt and suspenders” approach that favored doing everything possible to license a composition, even if it meant taking redundant steps. Eight Mile Style has scoffed at that characterization, but, whether or not Spotify's particular framing of the issue during litigation has been opportunistic, the reality is undeniable: for a great many works-of which the EMS Compositions were just a few-Spotify was relying on blanket licenses of uncertain scope. Whether or not that was a wise decision, it belies any assumption that Spotify's actions regarding the EMS Compositions establish that it knew that those particular compositions were unlicensed. Spotify's knowledge, accordingly, does not preclude the application of equitable estoppel.

4. Injury to Spotify

The injury to Spotify from Eight Mile Style's conduct is self-evident. The potential upward bound of liability in this case is strikingly high, and there is every reason to think that it would be a great deal lower if Eight Mile Style had simply taken the ordinary step that rights holders across the country take every day and simply told Spotify to stop. There is no doubt that Spotify wants to be able to stream the EMS Compositions, nor is there any doubt that Eight Mile Style wants to make money off of the compositions' use-including through streaming. Although Spotify missed the window for compulsory licensing, the parties still could have come together and worked out a licensing deal that would have benefited both. Eight Mile Style's wait-and-sue strategy, however, effectively took that option off the table. That injury is sufficient to support estoppel.

5. Application to this Case

“Where equitable estoppel is established, all relief on a claim may be barred.” A.C. Aukerman Co., 960 F.2d at 1041. The court finds that that is the outcome appropriate in this case. Rewarding Eight Mile Style's decision to allow the ambiguity around its rights to persist would not only be against conventional principles of equity, but would also frustrate the constitutionally legitimate objectives of the U.S. copyright regime, which, as a matter of constitutional principle, does not exist simply to maximize the recovery of individual rights holders, but, rather, to serve the public's interest in the production and availability of valuable work. See U.S. Const. art. I, § 8, cl. 8.

In furtherance of that objective, Congress adopted a system of compulsory licensing based on its recognition that a singular focus on “‘securing the right[s] of composers'” risked “the creation of a . . . ‘music monopoly'” that granted all of the power to copyright owners at the expense of the public. Duchess Music Corp. v. Stern, 458 F.2d 1305, 1309 (9th Cir. 1972) (quoting H. Rpts. 5 (1909)). The compulsory licensing regime, in other words, was intended to make licensing simpler-not more treacherous. The practical unwieldiness of the NOI system in the age of the digital music, however, resulted in a thicket of confusion that was ripe for abuse- by both the licensor and the licensee. Given the sheer volume of the rights at issue, industry participants often had little choice but to proceed on the “reasonable assumption” that licensing relationships that looked legitimate were legitimate, ABKCO Music, Inc. v. Sagan, 50 F.4th 309, 321 (2d Cir. 2022), which inevitably created a risk of abuse. While opportunities undoubtedly existed for digital streaming providers, including Spotify, to take advantage of such assumptions, there is no reason to doubt that sophisticated rights owners had such opportunities as well- particularly in light of the damages available for infringement.

Indeed, the tendency of the Copyright Act's generous damages provisions to reward strategic behavior is well-established. See Freeplay Music, LLC v. Dave Arbogast Buick-GMC, Inc., No. 3:17-CV-42, 2019 WL 4647305, at *13 (S.D. Ohio Sept. 24, 2019) (discussing plaintiffs who “are more focused on the business of litigation than . . . licensing their [copyrights] to third parties”). Congress granted victims of infringement the right to recover damages that go beyond merely compensating them for their losses, based on the rationale that those enhanced damages are necessary to “prevent the infringer from unfairly benefiting from a wrongful act” and to deter future infringement. Bridgeport Music, Inc. v. Justin Combs Pub., 507 F.3d 470, 489 (6th Cir. 2007). Whatever the merits of that approach, one side effect is that being the victim of infringement is, in many instances, more remunerative than being an above-the-board licensor. If boundaries are not placed on sophisticated rights holders' ability to exploit that disparity, the result, inevitably, will be a copyright regime that rewards rights holders in proportion to their strategic acumen and litigation budgets-not the value of their works.

It is ultimately up to Congress to strike the right balance regarding all of the competing interests and objectives at play in the digital music economy. Congress, however, chose to adopt a copyright regime against the preexisting background of common law equitable defenses like estoppel, meaning that, in a select few circumstances, the job of safeguarding the system from misuse falls to the courts. Spotify has established all of the substantive prerequisites for invoking estoppel here, and the underlying dynamics of the situation support such an application. The court will, accordingly, hold that all of Eight Mile Style's claims against Spotify are barred.

The court stresses that this holding is based on the specific facts of this case and that it would be a mistake to assume that other parties in Eight Mile Style's general position of having had its rights violated by Spotify will or should face a similar bar. An innocent rights holder who simply failed to recognize that Spotify lacked the right to stream its music might face an issue of timeliness with regard to its oldest claims, but it would face no barrier based on equitable estoppel. Here, however, Eight Mile Style, as a sophisticated actor with substantial resources, affirmatively fostered and failed to rectify the confusion surrounding its rights and then, rather than simply making use of the already-favorable position it found itself in, allowed Spotify to keep infringing, over and over, in a way that would make no sense other than as a business strategy by Eight Mile Style. That is why estoppel applies here-not because Spotify's general policies were excusable or because rights holders have any obligation to sue as soon as infringement becomes discoverable, but because Eight Mile Style improperly chose the cultivation of infringement damages over the proper functioning of the copyright system.

HFA is, of course, distinct from Spotify, and it is possible that estoppel could bar claims against one defendant but not another. Spotify and HFA, however, acted in concert for all relevant purposes, and the analysis that supported applying estoppel to bar Eight Mile Style's claims against Spotify similarly applies with regard to HFA. As with Spotify, Eight Mile Style could have brought matters to a head at any time, and it instead chose to wait. Like Spotify, HFA has been forced to defend itself against a potentially devastating damages award. If anything, the failure to keep HFA apprised of the actual allocation of licensing rights with regard to the EMS Compositions is particularly striking, given HFA's role in operating a central repository of rights information relied upon throughout the U.S. music industry. The court, accordingly, finds the claims against both defendants to be barred by the doctrine of equitable estoppel.

E. Indemnification

This court already addressed the issue of indemnification in connection with Spotify's Motion for Leave to File Early Motion for Summary Judgment Against Kobalt (Doc. No. 265), and, while the court found that resolving the issue at that juncture would have been premature, its preliminary analysis remains unchanged. See Eight Mile Style, LLC v. Spotify USA Inc., No. 3:19-CV-0736, 2022 WL 4097710, at *2 (M.D. Tenn. Sept. 7, 2022). On its face, the 2016 BMLA (1) purports to grant Spotify a mechanical license with regard to any composition that Kobalt “controls” or “administers,” (2) does not define the terms “control” or “administer,” and (3) requires Kobalt to indemnify Spotify for damages in the event that Kobalt misrepresented its authority to grant such a license. As the court has already held in this opinion, Kobalt did not actually have authority to license the EMS Compositions for streaming in the U.S. at the relevant times, and Spotify now faces substantial potential liability that it would not face if the EMS Compositions had been validly licensed in 2016. Kobalt therefore has an indemnification obligation if it “controlled” or “administered” the EMS Compositions, as those terms are used in the 2016 BMLA.

“The ultimate goal of interpreting a contract is to ascertain and give effect to the intent of the parties as reasonably manifested by the language of their written agreement.” LHR, Inc. v. T-Mobile USA, Inc., 112 A.D.3d 1293, 1295, 977 N.Y.S.2d 816, 819 (2013). “When the parties have used contract terms which are ‘in common use in a business or art[,]' . . . but which ‘convey no meaning to [t]hose who are not initiated into the mysteries of the craft,'” then the court typically must ascertain “the meaning of the language as generally understood in that business, in the light of the customs and practices of the business.” L. Debenture Tr. Co. of New York v. Maverick Tube Corp., 595 F.3d 458, 466 (2d Cir. 2010) (quoting Fox Film Corp. v. Springer, 273 N.Y. 434, 436, 8 N.E.2d 23, 24 (1937)). As the court noted earlier in this case, “all of the underlying transactions and events at issue [in this case] occurred against the unique backdrop of the music publishing industry, with its preexisting norms, customs, and terminology,” and “[t]he definitions of ‘control' and ‘administer,'” as used in the music publishing/licensing context, “implicate the need for that kind of industry-specific understanding.” Eight Mile Style, 2022 WL 4097710, at *2.

The parties have now had the opportunity to present evidence regarding industry practice, and that evidence both provides some clarity and highlights the degree to which rights administration, as a concept, may not lack a single clear definition at all. The parties and experts broadly agree that “[p]ublishing administration involves different administrative functions,” which may include “registering musical works with mechanical rights organizations” and “collecting and distributing royalties.” (Doc. No. 437 ¶¶ 166-73.) The parties also specifically agree that “[a]dministration can include authority to issue some types of licenses but not other types of licenses in a given territory” and that a party can be an “administrator” even if it does not perform all of the duties that administrators sometimes perform. (Id. ¶¶ 171, 177.) It is, therefore, clear that there is not a full, comprehensive list of responsibilities that every administrator possesses.

Kobalt, however, takes the position that, regardless of any ambiguity in the general definition of “administration,” “an administrator must have mechanical licensing authority in order to be an administrator for mechanical licensing purposes.” (Doc. No. 437 ¶ 166.) In support of this reading, Kobalt relies chiefly on an Expert Report authored by Trodd Brabec, an entertainment lawyer and former ASCAP executive. (See Doc. No. 596-14 at 17.) According to Brabec, “[t]he common understanding in the music industry of the term ‘Administrator' is simply a company or entity that manages and licenses the rights of another publishing company, whether it be a corporation, partnership, individual ownership (e.g. songwriter) or other legal entity.” (Id. at 7.) Another Kobalt expert, Clark Miller, offered a similar opinion. (Doc. No. 596-10 at 149.)

Brabec's definition, read literally, would seem to suggest that Kobalt was an administrator of the EMS Compositions, given the role it played in synchronization licenses. Brabec, however, rejects that reading on the basis that Kobalt's rights to “pitch” synchronization licenses was non-exclusive, and any licenses were subject to final approval by Eight Mile Style. Such a role, Brabec suggests, would be understood, in the music industry, to be that of a “synch agent,” not an administrator. Accordingly, he concludes, “Kobalt is not and has not been since at least February 4, 2011, the administrator of Plaintiffs' Compositions in the United States and Canada.” (Doc. No. 342-10 at 4-8.) Other experts' discussions of what it means to “administer” a copyright had a similar focus but showed no consensus about precise boundaries. (See Doc. No. 509 ¶ 44.)

Based on the experts' opinions, it appears that the term “administers” is ambiguous. See Williams v. Vill. of Endicott, 91 A.D.3d 1160, 1162, 936 N.Y.S.2d 759, 761 (2012) (“A contract is ambiguous if the language used lacks a definite and precise meaning, and there is a reasonable basis for a difference of opinion”) (quoting Pozament Corp. v. Aes Westover, LLC, 27 A.D.3d 1000, 1001 (2006)). New York follows the general rule that, “[i]f a court finds an ambiguity in a contract's terms, it then should be resolved by the trier of fact.” Pozament Corp., 27 A.D.3d at 1001. That rule, however, is subject to two important caveats. First, courts often distinguish a general susceptibility to multiple interpretations from “true ambiguity,” which arises only when a contractual term is genuinely susceptible to multiple reasonable interpretations. See BKD Twenty-One Mgmt. Co., Inc. v. Delsordo, 127 So.3d 527, 530 (Fla. Dist. Ct. App. 2012); see 731 W. Lake Rd., Inc. v. Boheen, 58 A.D.2d 1038, 1038, 397 N.Y.S.2d 507, 508 (1977). More fundamentally, the very nature of the court's authority-that is, its authority to decide cases and controversies under Article III of the Constitution-means that what matters is never whether a contractual term is ambiguous in some abstract sense, but whether it is ambiguous in the context of the specific dispute before the court. Many contractual terms are susceptible to ambiguity on the margins, but the margins are not the only place where litigation arises. The evidence before the court establishes that the terms “administer,” “administration,” and “administrator” are, in fact, genuinely subject to multiple reasonable interpretations, but that ambiguity means very little unless it actually raises a question about who should prevail under the agreement.

Based on the foregoing, the court finds that no reasonable interpretation would permit a finder of fact to reject the proposition that Kobalt administered the compositions in some sense. Kobalt granted licenses; it received license requests; it accepted NOIs; it collected royalties; and it was identified as the compositions' administrator time and time again. It was, by a considerable margin, the third party most involved in managing the licenses of the EMS Compositions, and, even if one accepts the most stringent definition of “administer” available- absolutely requiring that such a party have actual licensing authority-that definition would not exclude Kobalt, because Kobalt did have licensing authority for the EMS Compositions. It could license them for a variety of purposes worldwide, and it could license them for synchronization purposes in the U.S.

While it is true that synchronization licenses were subject to approval by Eight Mile Style, that is unsurprising, given that synchronization agreements, unlike compulsory mechanical licenses, would need to be negotiated, including in terms of price. (See Doc. No. 436-1 ¶ 239.) For example, in September 2011, Kobalt, as the “licensor,” granted a synchronization license to Walt Disney Studios Motion Pictures for a price that was more than double what Disney had originally proposed. (See Doc. No. 437 ¶ 236; Doc. No. 357-8.) The fact that Kobalt had to obtain Eight Mile Style's approval in connection with major financial decisions like that is unsurprising and does not undermine the fact that the work being done by Kobalt consisted of “administration,” as normally understood in the music publishing industry.¹¹

11 Synchronization licenses, moreover, were an important part of both the EMS Compositions' value and Kobalt's own business. Correspondence from 2013 suggests that, by that point, Kobalt had worked on as many as 214 “projects concerning synch licenses or potential synch licenses for works authored in whole or in part by” Mathers in advertisements, television shows, and movies. (See Doc. No. 344-20 at 2.) Eight Mile Style's efforts to minimize Kobalt's authority are, therefore, unconvincing. Moreover, even aside from the issue of synchronization licenses, there remains the fact that Kobalt was openly fielding licensing requests of all types for Eight Mile Style, with Bridgeport exercising its formal authority behind the scenes after Kobalt passed the requests along.

The idea that such actions did not amount to some form of administration appears to be a post hoc invention that cannot be squared with the actions of the parties themselves. Kobalt was openly discussed as an administrator by Eight Mile Style's own employees, including with regard to U.S. dealings. Although Bridgeport was the only party with the authority to actually grant a U.S. mechanical license, all of the available evidence suggests that the reorganization that resulted in that arrangement was much more of an allocation of formal rights than of responsibilities, with Kobalt continuing, largely, to oversee the works.

Even the formal paperwork confirmed that Kobalt remained an administrator. The letter agreement that granted Kobalt its collection and synch pitch rights was expressly drafted as an adjunct to an “Administration Agreement” pursuant to which Kobalt was “the Administrator” of the EMS Compositions. Kobalt's LOD that specifically coincided with the 2011 reorganization identified Kobalt as an administrator. And, finally, Kobalt has conceded, in this litigation, that it “is, and has been since 2004, the administrator of the EMS Compositions with respect to the territory of the world outside of the United States and Canada.” (Doc. No. 437 ¶ 273 (emphasis added).) Insofar as this case comes down to whether Kobalt was an administrator in some sense, then, the question presented is not difficult. Admittedly, at a few points, individual experts have offered bright lines that would exclude Kobalt from the category of “administrator” in the U.S.- for example, that an administrator simply must have formally exclusive licensing rights or a right to license a work without the approval of the owner-but all of the available documentary evidence confirms that such strict boundaries are not accepted or observed in actual music industry practice.

Ultimately, though, a singular focus on defining administration, in the abstract, fails to capture the full nature of the actual disagreement at hand. Certainly, if the evidence had shown that what Kobalt was doing clearly was not administration or control, then Kobalt would have no indemnification obligation. However, Kobalt also argues that, regardless of the potentially broad meaning of “administer” generally, the 2016 BMLA should be understood to adopt the definition that, Kobalt suggests, would make the most sense in context- that is, “administer for the purpose of U.S. mechanical licensing rights.” That argument does not actually call on the court to choose between reasonable interpretations of “administer” as a term of art-because, whatever “administer” means, it definitely means something broader than “administer U.S. mechanical licensing rights.” This version of Kobalt's argument, therefore, is not ultimately about the definition of the term, in and of itself. Rather, Kobalt asks the court to read an implicit limiting clause into the 2016 BMLA's language based on the overall structure and purpose of the contract. The court can determine, as a matter of law, whether such an implicit limitation exists.

Because Kobalt is suggesting that the court depart from the express terms of a written contract between two sophisticated parties, it bears a heavy burden. Kobalt's position does find some general support from the principle that, under New York law, “indemnification provisions must be narrowly construed.” Lebedev v. Blavatnik, 193 A.D.3d 175, 187, 142 N.Y.S.3d 511, 520 (2021). There are also policy reasons why one might prefer Kobalt's interpretation. In particular, a bright-line rule that the 2016 BMLA did not purport to license any composition for which Kobalt lacked the authority to grant such a license would arguably be simpler and more workable than a more literal reading of the agreement's language. A fuzzy definition of “administer” that required parties to ascertain the powers granted to Kobalt in connection with any given composition, and then discern which side of the “administered by Kobalt” line that composition fell on, could be a recipe for confusion and litigation. Certainly, it has created a great deal of expensive confusion here.

Broad principles of construction and policy arguments, however, are not enough to overcome the contract's plain language, as applied to the facts of this case. The 2016 BMLA unambiguously purports to grant mechanical licensing rights to all compositions administered by Kobalt, and it expressly warrants that Kobalt has the power to grant such a license. The EMS Compositions were, in both the eyes of Eight Mile Style and the broader music industry, administered by Kobalt. Kobalt's only argument that the compositions should nevertheless be treated as exempted from the agreement is that Kobalt did not possess the specific right to grant the necessary licenses-that is, by denying the exact warranty that Kobalt made in the agreement. If anything, this situation-in which Kobalt, as a matter of both practice and industry parlance, administered certain works but happened to have signed away the U.S. mechanical licensing rights, in particular, for those works-would seem to be exactly the type of situation that the warranties were intended to guard against.

The suggestion that Kobalt's alternative reading would be more consistent with the agreement as a whole ignores the fact that the warranties and indemnification obligations of the 2016 BMLA represent a voluntary allocation of risks and responsibilities between the parties that is as fundamental to the agreement's purpose as any other provision. Kobalt has conceded, in this litigation, that “[i]t is not customary for blanket digital licensees to conduct due diligence into the works covered by a blanket digital mechanical license agreement” and that “[a] publishing administrator that grants a blanket digital mechanical license is generally more aware than the blanket licensee [of] what the administrator has authority to do on behalf of each of its clients.” (Doc. No. 437 ¶¶ 360-361.) This transaction reflected both of those realities: Spotify licensed Kobalt's catalog, but only Kobalt was in a position to know the details of what it could and could not grant. Kobalt's warranties and indemnification responsibility, therefore, represented a rational, mutually agreed upon allocation of risk to the party with the far superior capacity to assess and avoid that risk. The reading of the 2016 BMLA that best honors that intention is also the reading most consistent with the agreement's plain language.

Kobalt's attempts to explain how the warranty and indemnification provisions would make sense under its reading of the agreement only highlight the untenability of its position. Kobalt apparently recognizes that it cannot plausibly deny that it agreed to indemnify Spotify with regard to at least some compositions. It suggests, however that its

indemnification obligation only extends to those compositions to which it was given the contractual right to license, where a third party - not the party which granted Kobalt the authority to issue licenses - successfully challenges the validity of the underlying grant of rights to Kobalt or Kobalt's counterparty (usually a songwriter) such that it is determined that Kobalt's client did not have the rights it granted to Kobalt.

(Doc. No. 441 at 38.) What Kobalt is describing, however, is just a more circuitous route to the exact same scenario at issue here, where, as in the hypothetical, Kobalt seemed to be the administrator of a composition but actually lacked mechanical licensing authority for reasons unknown to Spotify. Why, then, would the indemnification provision apply to the hypothetical scenario, but not this one? The provision itself contains no language drawing the kind of line that Kobalt would impose. Kobalt, then, would have the court do more than simply read an implicit limiting clause into the word “administer”; it would have the court essentially rewrite Kobalt's indemnification obligation to adopt a detailed substantive framework that has no support whatsoever in the text. The court has no authority to do so.

The court, accordingly, will grant Spotify summary judgment on its indemnification claim, as well as its breach of contract claim based on the assertion that Kobalt failed to license the compositions it purported to license through the 2016 BMLA. Kobalt's rejection of Spotify's indemnification demand further supports an award of summary judgment to Spotify with regard to Spotify's claim for anticipatory repudiation. “Anticipatory repudiation occurs when a party ‘attempt[s] to avoid its obligations by advancing an untenable interpretation of the contract, or . . . communicate[s] its intent to perform only upon the satisfaction of extracontractual conditions.” Fonda v. First Pioneer Farm Credit, ACA, 86 A.D.3d 693, 694-95, 927 N.Y.S.2d 417, 419 (2011) (quoting SPI Commc'ns, Inc. v. WTZA-TV Assocs. Ltd. P'ship, 229 A.D.2d 644, 645, 644 N.Y.S.2d 788, 790 (1996)). Kobalt's informing Spotify that it would not honor the indemnification obligation was sufficient to support an award of summary judgment on such a claim.

F. Remaining Spotify Claims against Kobalt

1. Intentional/Negligent Misrepresentation

“On a cause of action alleging negligent misrepresentation, the plaintiff is required to demonstrate ‘(1) the existence of a special or privity-like relationship imposing a duty on the defendant to impart correct information to the plaintiff; (2) that the information was incorrect; and (3) reasonable reliance on the information.'” Ginsburg Dev. Companies, LLC v. Carbone, 134 A.D.3d 890, 894, 22 N.Y.S.3d 485, 490 (2015) (quoting J.A.O. Acquisition Corp. v. Stavitsky, 8 N.Y.3d 144, 148, 863 N.E.2d 585, 587 (2007)). “To recover on that cause of action, plaintiff would be required to establish a misrepresentation or a material omission of fact which was false and known to be false by defendant, that the misrepresentation was made for the purpose of inducing plaintiff to rely upon it, justifiable reliance by plaintiff on the misrepresentation or material omission, and injury.” Horn v. Toback, 44 Misc.3d 42, 46, 989 N.Y.S.2d 779, 782 (App. Term 2014) (citing Shao v. 39 Coll. Point Corp., 309 A.D.2d 850, 851, 766 N.Y.S.2d 75, 76 (2003)). To prevail under either theory, then, Spotify would need to show that it was actually deceived by Kobalt and actually injured by that deception.

As the court has already discussed, that is not what happened here. Spotify was neither tricked, nor negligently misled, into believing that Kobalt had the U.S. mechanical licensing authority for the EMS Compositions in 2016. It cannot establish otherwise by retroactively cobbling together a set of facts that theoretically might have deceived it, if it had actually been aware of and relied on those facts. Kobalt is, therefore, entitled to summary judgment with regard to intentional and negligent misrepresentation.

2. Breach of Contract (NMPA Settlement)

As the court has already discussed, there is no reasonable interpretation of the PPPUUA that would have included the EMS Compositions. The PPPUUA's approach to works for which Kobalt lacked authority was essentially the opposite of the 2016 BMLA's: while the 2016 BMLA offered warranties and indemnification, the PPPUUA expressly provided that Kobalt was only releasing what it could release and granting what it could grant. There is, therefore, no basis for finding breach of contract by Kobalt in connection with the PPPUUA or the underlying NMPA settlement, and the court will grant Kobalt summary judgment on that count.

G. What Remains of this Case

The foregoing is sufficient to resolve the question of which party should prevail on each of the claims stated in this case. There is, however, still work to be done, because the fact that Kobalt has an indemnification duty to Spotify does not resolve the question of what it will owe Spotify based on its failure to honor its duty. If Eight Mile Style had prevailed on its claims, the full sum at issue under that indemnification obligation might still require substantial discovery and litigation. As it stands, however, it appears that Spotify's recovery will be limited to reasonable attorney's fees and expenses, which can typically be ascertained through a much more streamlined process. The court, accordingly, will require Spotify to support its damages within ten days of the court's resolution of these motions.

IV. CONCLUSION

For the foregoing reasons, HFA's Motion for Summary Judgment (Doc. No. 317) will be granted, Spotify's Motion for Summary Judgment (Doc. No. 320) will be granted in part and denied in part, Kobalt's Motion for Summary Judgment (Doc. No. 328) will be granted in part and denied in part, and Eight Mile Style's Motion for Summary Judgment (Doc. No. 348) will be denied.

An appropriate order will enter.