Hart, a former Rutgers football player who saw significant success on the field, appealed EA’s district court win against his right of publicity claim based on EA’s use of his likeness and biographical information (size, weight, jersey number, typical attire of visor and armband) in its NCAA Football videogames. The court of appeals reversed.
Unsurprisingly, the court began with reference to the NCAA’s amateurism rules, which precluded Hart from making any money from his status, but didn’t stop the NCAA from monetizing his popularity.
NCAA Football comes out every year, and allows users “to experience the excitement and challenge of college football” by interacting with “over 100 virtual teams and thousands of virtual players.” Users typically play by choosing two teams to compete against each other. The game “assigns a stadium for the match-up and populates it with players, coaches, referees, mascots, cheerleaders and fans.” There are also multi-game play modes. Dynasty Mode allows users to control a college program for up to thirty seasons, allowing players to recruit high school players from a randomly generated pool. Race for the Heisman/Campus Legend mode allows users to control a single user-made virtual player from high school through college, making choices about practices, academics and social activities. EA licenses rights to team-related content from the NCAA, but does not license likeness and identity rights for collegiate players—unlike some of its other franchises. The game’s success depends on its focus on realism and detail, “from realistic sounds, to game mechanics, to team mascots. This focus on realism also ensures that the ‘over 100 virtual teams’ in the game are populated by digital avatars that resemble their real-life counterparts and share their vital and biographical information.” This included a Rutgers avatar that resembled Hart. While users can change avatars’ appearances and vital statistics, certain details were immutable: the player’s home state, home town, team, and class year.
Hart sued based on the use of his likeness/biographical and career statistics in the game, and also over the use of an actual photo of him in a photo montage visible only when players select Rutgers as their team. Preview: This last was held transformative and fair because it’s a montage, while the rest is not transformative because it’s a realistic videogame. If those becauses don’t make any sense to you, welcome to the club!
As you know, Bob, videogames are fully protected by the First Amendment. Brown v. Entm’t Merchs. Ass’n, 131 S. Ct. 2729 (2011). However, such protection can be trumped by other protected rights. EA conceded for purposes of the appeal that it misappropriated his identity for commercial exploitation; this required balancing the interests underlying the right to free expression with the interests in protecting the right of publicity. Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562 (1977) (also a cite to Eldred and to defamation law).
The court then engaged in a review of existing case law. The basic theory of the right of publicity is that people have the right to enjoy the fruits of their own industry; it’s unfair to allow others to commercialize, exploit, or capitalize on another’s name or reputation. People’s names and likenesses belong to them as property because they are things of value. New Jersey follows the Restatement (Second) of Torts approach that “[o]ne who appropriates to his own use or benefit the name or likeness of another is subject to liability to the other for invasion of privacy,” and that this is a property right. The Third Restatement adopted a freestanding right of publicity claim, but New Jersey has yet to adopt it.
In New Jersey, name, likeness, and endorsement therefore carry value, and unauthorized use harms the person by “diluting the value of the name and depriving that individual of compensation.” The goal of the right of publicity is therefore to protect an individual’s property interest in their identity gained through labor and effort, and to encourage further investment in this property interest. See Zacchini (“[T]he State’s interest in permitting a right of publicity. . . . is closely analogous to the goals of patent and copyright law, focusing on the right of the individual to reap the reward of his endeavors . . . .”). Relatedly, the court rejected the idea that no right of publicity was necessary because players are financially rewarded through endorsement, sponsorship, and the like—this doesn’t apply to college athletes, and the NCAA most recently estimated that “[l]ess than [two] in 100, or 1.6 percent, of NCAA senior football players will get drafted by a National Football League (NFL) team.” Ryan Hart was in the nearly 99%. (So does that mean that in our future case-by-case balancing we do get to count those other economic opportunities against granting control to celebrities who are getting paid?)
Zacchini called for balancing of First Amendment interests, but didn’t set forth a systematic analytic framework for doing so. Lower courts have tried several approaches. The court identified three as of interest: the execrable Doe v. TCI “predominant use” test, the trademark-based Rogers test, and the copyright-based transformative use test. It picked the transformative use test as the most appropriate.
As for Doe v. TCI, asking whether the predominant use of a celebrity’s identity is commercial is “subjective at best, arbitrary at worst, and in either case calls upon judges to act as both impartial jurists and discerning art critics. These two roles cannot co-exist.” The court declined to get into whether using Hart’s identity added anything to the “First Amendment expressiveness” of the game:
Such reasoning, however, leads down a dangerous and rightly-shunned road: adopting Appellant’s suggested analysis would be tantamount to admitting that it is proper for courts to analyze select elements of a work to determine how much they contribute to the entire work’s expressiveness. Moreover, as a necessary (and insidious) consequence, the Appellant’s approach would suppose that there exists a broad range of seemingly expressive speech that has no First Amendment value.”
(This criticism is also true of transformative use, especially as applied here, but hey!)
The court then rejected Rogers because it would “potentially immunize a broad swath of tortious activity.” (You know, like New York Times v. Sullivan.) The Restatement (Third) of Unfair Competition restricts the right of publicity to “use for purposes of trade,” which does not “ordinarily include the use of a person’s identity in news reporting, commentary, entertainment, works of fiction or nonfiction, or in advertising that is incidental to such uses” unless a name or likeness is used solely to attract attention to an unrelated work. The Restatement specifically mentions uses in articles in fan magazines, feature stories on entertainment programs, unauthorized biographies, novels, plays, and motion pictures.
The court didn’t find the cases (allegedly) adopting Rogers persuasive. Parks v. LaFace, for example, was in tension with the transformative use test applied in ETW v. Jireh, decided by another Sixth Circuit panel a month later. And the Third Circuit had already, in dicta, expressed doubt that Rogers could apply beyond a title. Facenda v. N.F.L. Films, Inc., 542 F.3d 1007 (3d Cir. 2008). Though Rogers may be useful in “trademark-like” right of publicity cases, whatever that means, it wasn’t appropriate here (since it would mean that EA won). Rogers is a “blunt instrument” that doesn’t carefully calibrate the balancing of “two fundamental protections: the right of free expression and the right to control, manage, and profit from one’s own identity.” (Although the last time I checked, only one of those was in the Bill of Rights. Also, categorical balancing is often used in First Amendment cases, to avoid chilling effects—case by case determination of “transformativeness” has significant costs that have to be balanced while we’re balancing.)
The problem the court had with Rogers was demonstrated by EA’s argument, which was that Hart’s likeness wasn’t wholly unrelated to NCAA football, and the game wasn’t an ad for an unrelated product, therefore EA won. EA argued that Hart “should be unable to assert a claim for appropriating his likeness as a football player precisely because his likeness was used for a game about football. Adopting this line of reasoning threatens to turn the right of publicity on its head.” (That presupposes exactly the question to be answered, which is how far the right of publicity may constitutionally extend. Also, imagine this statement made about a conventional print biography. Or about a news photograph accompanying a story about the Rutgers football team. Or Joyce Carol Oates’ next roman a clef about a college football player who sues a video game company. Etc.) You see, Hart’s career means that the target audience for his “merchandise and performances” would be sports fans, so products “appropriating and exploiting his identity” would do best with them. But under Rogers, all products (and here we mean “expressive works protected by the First Amendment”) targeting the sports fan market would, as a matter of course, relate to Hart and fall outside his right of publicity. “It cannot be that the very activity by which Appellant achieved his renown now prevents him from protecting his hard-won celebrity.” (No, but it can be that engaging in speech about the very activity by which Hart achieved his renown prevents him from maintaining a successful claim against the speakers; what we have here is a failure to define the scope of the right, or explain why it should trump the interest in engaging in nondefamatory noncommercial speech.)
For these reasons, Rogers was inapplicable, though the Rogers court was right to note that the right of publicity doesn’t have consumer confusion as an element and was therefore “potentially broader” than trademark. Of course, this is a reason why the right of publicity does less well in any balancing test than trademark—deceptive speech is easier to regulate than nondeceptive speech, at least when it does harm (Alvarez)—not a reason why a right of publicity claim trumps freedom of speech when an equivalent trademark claim, supposedly protecting both the public and the trademark owner, wouldn’t be sufficiently weighty to do so. But that’s not what the Hart court says, since we are in topsy-turvy land: “the right of publicity is broader and, by extension, protects a greater swath of property interests. Thus, it would be unwise for us to adopt a test that hews so closely to traditional trademark principles.” (Traditional pre-Sullivan defamation law protected a “greater swath” of personality interests than current defamation law; that made it unconstitutional. And one could say, as Justice Rehnquist once did, that the US flag is the “property” of the American people, and yet flagburning statutes are still unconstitutional. Calling an interest “property” doesn’t make it First Amendment-immune. Even Eldred and Golan don’t go that far—not that any legislature adopting a right of publicity has engaged in even the token factfinding backing up the CTEA.)
So, the court turned to a “broader, more nuanced test,” transformative use. Copyright was a better analogy than trademark, because the right of publicity is an economic right against misappropriation. It applies to noncommercial speech too, offering “protection to a form of intellectual property that society deems to have social utility.” Therefore, taking transformativeness from copyright’s fair use test was appropriate, especially since transformative uses are less likely to interfere with the economic interests implicated by the right of publicity. Of course, transformativeness may include factual reporting and fictionalized portrayal along with parody and satire. The inquiry, as Comedy III v. Saderup said, was “in a sense more quantitative than qualitative, asking whether the literal and imitative or the creative elements predominate in the work.” (Wasn’t that the test the court just rightly rejected from Doe v. TCI?)
If the marketability of the challenged work doesn’t derive primarily from the fame of the celebrity depicted, there’s generally no violation of the right of publicity. However, even if the work’s marketability does derive from the celebrity’s fame, it can still be transformative. The question is whether the celebrity likeness is one of the “raw materials” from which an original work is synthesized, or whether the depiction is the “very sum and substance” of the work. (But of course biographies and news reporting are fine! The reason why is … left as an exercise for the reader.) So the question is whether the celebrity’s likeness “has become primarily the defendant’s own expression rather than the celebrity’s likeness,” and “expression” has to be “expression of something other than the likeness of the celebrity.”
(Not that word portraits will never be in jeopardy. The thing about medium discrimination is that it’s invisible to the people engaging in it. The work of portraying a celebrity in a textual biography is so self-evidently transformative creative “work” of the writer's "own" that the court feels no need to explain why it’s qualitatively different from the work of a portraitist. The work required to create a realistic video game, which involves extensive programming and creative design, by contrast appears to the court mere reproductive labor. I would still think this was dumb, but I wouldn’t be as aggravated if courts would outright say “the right of publicity, though in theory having something to do with ‘name,’ is only a right in images, at least as applied outside the context of false endorsement.”)
Winter v. DC Comics, 69 P.3d 473 (Cal. 2003) expanded on the California rule: “[a]n artist depicting a celebrity must contribute something more than a ‘merely trivial’ variation, [but must create] something recognizably ‘his own’ in order to qualify for legal protection.” ETW v. Jireh decided that a painting was transformative because it was a collage of Tiger Woods with other golf greats.
This leads to Kirby v. Sega of America, Inc., 50 Cal. Rptr. 3d 607 (Cal. Ct. App. 2006), finding transformative use of a musician’s image to create a video game character who was “a reporter in the far flung future.” Sega seemed to have copied Kirby’s signature phrases, hair style, and clothing choice, but differences in appearances and movement also existed. This additional new expression was enough to be transformative, even without a particular meaning or message.
In No Doubt v. Activision Publishing, Inc., 122 Cal. Rptr. 3d 397 (Cal. Ct. App. 2011), by contrast, the court found a violation of the right of publicity when the video game at issue used “immutable” images of the “real” musicians and didn’t transform their context, instead having them perform rock songs as they did in life. Even though the avatars could be manipulated to perform at fanciful venues or to sing songs the real band wouldn’t ever sing, and even though the game had many other creative elements, they were still “exact depictions” of the musicians doing their thing. The use of highly realistic digital depictions “was motivated by a desire to capitalize on the band’s fan-base,” so there was a violation of the right of publicity. (But fan magazines and entertainment programs are fine! Because … they’re not video games?)
The Hart court concluded that transformative use struck the best balance between the interests at issue “because it provides courts with a flexible — yet uniformly applicable — analytical framework.” Unlike Rogers, transformativeness allows courts “to account for the fact that misappropriation can occur in any market segment, including those related to the celebrity.” (And to ignore the difference between commercial and noncommercial speech, and deceptive and nondeceptive speech. No big deal.) But, unlike Doe v. TCI, transformativeness is “more circumscribed,” focusing on specific aspects of a work “that speak to whether it was merely created to exploit a celebrity’s likeness.” Transformativeness “recognizes that if First Amendment protections are to mean anything in right of publicity claims, courts must begin by considering the extent to which a work is the creator’s own expression.” The marketing power of the celebrity identity is a secondary question, since the balancing suggested by Zacchini can’t begin and end with commercial purpose or value. Transformative use “effectively restricts right of publicity claims to a very narrow universe of expressive works.” (Categorical balancing, with a focus on avoiding chilling effects on protected speech, might ask whether this game, no pun intended, is worth the candle.)
And the test was consistent with the outcomes of other cases, including Estate of Presley v. Russen, 513 F. Supp. 1339 (D.N.J. 1981), which held that an Elvis impersonator’s act was subject to right of publicity claims because “entertainment that is merely a copy or imitation, even if skillfully and accurately carried out, does not really have its own creative component and does not have a significant value as pure entertainment.” (Which totally explains why people go to see impersonators instead of watching video of the original! Oh, wait, it doesn’t. Also, Tina Fey would like a word.) As in Zacchini, the Elvis impersonator “appropriated the ‘very activity [live stage show] by which [Presley initially] acquired his reputation.” Where a work contains “‘merely a copy or imitation’ of the celebrity’s identity … there can be no First Amendment impediment to a right of publicity claim.” (Except for biographies and news reports/photography.)
Transformativeness was also, in the court’s view, consistent with the Restatement test, because the examples listed as protected in the comments to the Restatement “all exemplify the sort of transformative uses that would generally pass the analysis set forth in Comedy III.” Thus, “already-existing First Amendment protections in right of publicity cases apply to video games with the same force as to ‘biographies, documentaries, docudramas, and other expressive works depicting real-life figures.’” Except that they win and EA loses; other than that, totally the same.
Here, Hart’s protected “identity” included not only his likeness but also his biographical information; it was the combination that identified the digital avatar at issue as an in-game recreation of Hart:
This joint focus on both likeness and identifying information avoids a conflict with C.B.C. Distribution & Mktg., Inc. v. Major League Baseball Advanced Media, L.P., 505 F.3d 818 (8th Cir. 2007), which held that use of major league baseball players’ records in a fantasy baseball game was protected by the First Amendment even against right of publicity claims because such information was publicly available. The presence of a digital avatar that recreates Appellant in a digital medium differentiates this matter from C.B.C.
Okay, I lied: this is still aggravating. Why does the presence of his image make a difference to the First Amendment issues? Why do celebrities have more rights in pictures of them than in facts about them? A case could, I imagine, be made having to do with “if value then right” and the greater value of the image in present-day society, but it’s being assumed rather than articulated.
The court’s focus on visual representation causes it some uncertainty about the scope of its holding: “It remains an open question ... whether right of publicity claims can extend into the bygone days of 8-bit graphics and pixilated representations.” Here, though, the digital avatar closely resembled the “genuine article” in hair color, hair style, skin tone, accessories, vital information, and biographical details. “And while the inexorable march of technological progress may make some of the graphics in earlier editions of NCAA Football look dated or overly-computerized, we do not believe that video game graphics must reach (let alone cross) the uncanny valley to support a right of publicity claim.”
Digitization was not itself transformative. Recreating a likeness in a medium other than photographs or video can’t in itself satisfy the test, or the test wouldn’t make enough uses liable. (A couple of things going on here—on top of the initial disregard of the work involved in creating photos and video as creative, contra copyright, there’s a different erasure of creative labor in video game design/programming, portraiture, and other non-reprographic methods of depicting people. Except words, of course. Missing here and throughout is any explanation of why biographies etc. are transformative; this is what makes the opinion so clearly non-medium neutral.)
The next question was whether the context was transformative. Here, digital Hart did what actual Hart did at Rutgers: he played football. (*cough*Unauthorized biographies*cough*.) Though users could alter the avatar’s appearance, and though EA created a game in which users could make those kinds of decisions, that wasn’t enough. All games are interactive.
True, video games have First Amendment protection because they’re interactive. (Here I thought it was because they were expressive.) But the balancing test isn’t about whether a work loses First Amendment protection in the abstract. Rather, it’s about whether “the interests protected by the right of publicity are sufficient to surmount the already-existing First Amendment protections. As Zacchini demonstrated, the right of publicity can triumph even when an essential element for First Amendment protection is present.” And “[i]f the mere presence of [the ability to alter an avatar] were enough, video game companies could commit the most blatant acts of misappropriation only to absolve themselves by including a feature that allows users to modify the digital likenesses.” In a footnote, the court cautioned that interactivity wasn’t necessary for transformativeness, even for videogames. Placing avatars of the Dalai Lama and the Pope in a violent “shoot-em-up” game with a mechanism for manipulating their characteristics wouldn’t automatically satisfy the test, though the court didn’t “discount the possibility” that such a juxtaposition of spiritual leaders and hyperviolence could be transformative “on other grounds.”
Here, the interactivity was insufficient to transform Hart’s likeness into EA’s own expression. NCAA Football sought to create a realistic depiction of college football, including realistic representations of players, seeking to capitalize on teams’ and players’ fan bases. This heightened realism made the games more enjoyable and thus more valuable. Realistic depictions were the “sum and substance” of the digital players. Since Hart’s unaltered likeness was “central to the core of the game experience,” the ability to change the digital avatars wasn’t important. The ability to make minor alterations in the avatar, such as changes in his hair style, face, accessories, basic biographical information, or playing statistics, was insufficient, because an artist depicting a celebrity has to contribute more than a trivial variation.
“[T]he ability to modify the avatar counts for little where the appeal of the game lies in users’ ability to play ‘as, or alongside’ their preferred players or team.” Thus, even avatars with “superficial modifications” were suitable proxies or market subsitutes for the original. And once a user made major changes to the avatar, it wasn’t a “use” of his identity and was irrelevant to the question of how his likeness was used in the game.
Also, it didn’t matter that Hart was a tiny part of the overall game. Consideration of the other creative elements was improper, because transformativeness is about “how the celebrity’s identity is used in or is altered by other aspects of a work,” without consideration of “[w]holly unrelated elements.” (Forrest Gump is looking pretty problematic!) Consideration of the broader context of a work should be aimed at “whether this context acted upon the celebrity identity in a way that transformed it or imbued it with some added creativity beyond providing a ‘merely trivial variation.’” So, you’d better have a good creative reason for your celebrity reference in your book, movie, etc. Or just your video game.
An alternative holding, considering the creativity of the rest of the work, would allow “[a]cts of blatant misappropriation” as long as “the larger work, on balance, contained highly creative elements in great abundance. This concern is particularly acute in the case of media that lend themselves to easy partition such as video games.” (Calling Gawker and the Huffington Post—those short celebrity stories and pictures, so easily partitionable from the rest of the site, are now going to get you in trouble! Or would, if this were a medium-neutral decision.)
Just to make clear that video games are special, the court held that the photo of Hart that appeared in the 2009 edition of NCAA Football was not actionable. It was a fleeting component of a montage, and was transformative under ETW: putting the photo in the context of a Rutgers montage “imbue[d] the image with additional meaning beyond simply being a representation of the player.” (Emphasis added—note the slippage between the celebrity and a particular image of the celebrity, which contributes to the medium discrimination at work here. The photo puts Hart in the context of the Rutgers team, and that’s transformative. But the avatar, which plays on the digital Rutgers team, is not.) The use of the photo was protected by the First Amendment.
Judge Ambro dissented. He agreed that EA took Hart’s identity and used it for commercial success, and that realism was key to the franchise’s profitability. And he agreed that student athletes are exploited, and that this seemed unfair. Further, he agreed that transformativeness was the proper test. However, he interpreted and applied it differently. (Not for nothing, giving further credence to the objection that transformativeness too requires judges to be art critics.)
Transformativeness balanced “an individual’s right to benefit financially when others use his identifiable persona for their own commercial benefit [with] the First Amendment interest in insulating from liability a creator’s decision to interweave real-life figures into its expressive work.” But the inquiry couldn’t be limited to Hart’s identity alone, and EA shouldn’t be penalized for its commercial success or its realism, given “First Amendment protections traditionally afforded to true-to-life depictions of real figures and works produced for profit.”
First, the work should be evaluated in its entirety to figure out whether a work is “literal” or a celebrity likeness is simply one of the “raw materials” from which an original work is synthesized, as stated in Comedy III and reemphasized in Winter, which noted that the Autumn brothers were placed in a larger storyline with plenty of creativity of its own. (In a footnote, the dissent pointed out that the use here wasn’t harming Hart’s commercial value—if anything, ticket sales for college football games and the value of Hart’s endorsement would have been augmented by the popularity of EA’s games. This gets to the core of “misappropriation”—what is it we think counts as such? To say that a person is harmed by not getting payment for a “use” presupposes the proper scope of the right.)
The majority erred in disregarding the non-Hart creative elements of the video game, conflicting with other courts’ approaches, such as that taken by ETW. “My colleagues do not—and, in my view, cannot—explain how the photographic images of Woods were transformed if they limit their analysis to ‘how the celebrity’s identity is used.’” The use here, “creating digital avatars of football teams and placing them in an interactive medium designed for user interaction and manipulation” was “significantly more ‘transformative’ than the use in ETW, which simply combined several photographs into a photomontage.” (Who was reading the actual cases? Of course I don’t think it matters legally that Rick Rush is a painter and not a collage artist, but this is embarrassing.) To Judge Ambro, a narrow focus on an individual’s likeness, rather than on how it was incorporated into and transformed by the work as a whole, was flawed. The whole was the better baseline.
Judge Ambro reminded us that the First Amendment protects “biographies, documentaries, docudramas, and other expressive works depicting real-life figures, whether the accounts are factual or fictional.” And it protects them when they’re sold for profit. Thus, using a celebrity name or likeness to increase an expressive work’s value and marketability doesn’t diminish its First Amendment protection.
The majority nodded in the direction of medium neutrality, but wasn’t medium neutral in its application of the test. “EA’s use of real-life likenesses as ‘characters’ in its NCAA Football video game should be as protected as portrayals (fictional or nonfictional) of individuals in movies and books.” Not all digital portrayals should be entitled to First Amendment protection, but they should be when a likeness, as part of a creative composition, “has been transformed into something more or different than what it was before.” The majority further erred in focusing on the marketability of the game; the “effect on the market” factor of copyright fair use is not part of the right of publicity test as promulgated by Comedy III. And even if it were appropriately considered, there was no argument that EA’s use decreased the commercial value of Hart’s identity or persona. The majority wrongly equated the use of realism to increase profits with misappropriation. “This approach is at odds with the First Amendment protection afforded to expressive works incorporating real-life figures. That protection does not depend on whether the characters are depicted realistically or whether their inclusion increases profits.”
Judge Ambro therefore concluded that EA’s use of realistic avatars was entitled to First Amendment protection. Hart’s mere likeness had been transformed into an avatar that, “along with the rest of a digitally created college football team, users can direct and manipulate in fictional football games.” The many original graphics, videos, sound effects, and game scenarios allowed users to direct the play of a college football team with many different players. Users weren’t reenacting real games, but directing the avatars in invented games and seasons. Campus Legend/Dynasty Mode also allowed users to control players and teams across multiple seasons, allowing them to generate their own narratives. These were “imaginative transformations of the games played by real players.” The permutations allowed for each player image added to this transformative interactivity. These elements were not “[w]holly unrelated” to the use of Hart’s likeness, as the majority contended. Users who selected the 2005 Rutgers team and didn’t change the avatar’s characteristics would experience the numerous creative elements of the game as part of every fictional play. Thus, his likeness was one of the raw materials from which a new work was synthesized, rather than the sum and substance of the work in question.