September 16, 2020 Feature

What They Do for a Living: The Right of Publicity in Video Games and Movies

Robert C. Cumbow

©2020. Published in Landslide, Vol. 13, No. 1, September/October 2020, by the American Bar Association. Reproduced with permission. All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or stored in an electronic database or retrieval system without the express written consent of the American Bar Association or the copyright holder.

Why can a movie freely replicate the actions of an army bomb disposal expert while a video game cannot replicate the actions of a professional football player without getting permission?

In Brown v. Entertainment Merchants Ass’n, the U.S. Supreme Court found unconstitutional a California state law restricting the sale of certain video games to children without parental supervision. In announcing the Court’s decision, Justice Antonin Scalia wrote:

Like the protected books, plays, and movies that preceded them, video games communicate ideas—and even social messages—through many familiar literary devices (such as characters, dialogue, plot, and music) and through features distinctive to the medium (such as the player’s interaction with the virtual world). That suffices to confer First Amendment protection.

It has long been understood that the right of publicity protects a person’s right against the unauthorized commercial exploitation of that person’s name, likeness, and recognizable personality traits. It arose as an action against unpermitted use of an individual’s persona in advertising and traditionally did not purport to restrict the mention and depiction of an individual in legitimate news, entertainment, or expressive art forms.

The only right of publicity case ever addressed by the U.S. Supreme Court was not about whether the news media could depict an individual but only about how much depiction is too much. In Zacchini v. Scripps-Howard Broadcasting Co., the Court held that Ohio’s right of publicity law could prevail over the First Amendment defense of a television station that had broadcast the entirety of the plaintiff’s performance—a 20-second act in which he was shot from a cannon into a safety net. Even the news media’s combined First Amendment freedoms of expression and news reporting did not allow them to harm a man’s livelihood by enabling viewers to watch on TV what they would have to pay for to see live.

When the Ninth Circuit Court of Appeals held in 1993 that Vanna White had stated a prima facie case for violation of her publicity right by Samsung’s use of a blond-wigged robot turning letter tiles on the “game show of the future” in a television commercial, Judge Alex Kozinski dissented vociferously that his colleagues had taken the right of publicity too far: “The panel is giving White an exclusive right not in what she looks like or who she is, but in what she does for a living.” But of course the Supreme Court had already done that in Zacchini.

Kozinski was not really objecting to balancing First Amendment freedom of expression with a person’s right to earn a living, but to the majority’s view that an image that merely called Vanna White to mind was sufficient to implicate her right of publicity. The fact that the robot turned tiles on a game show went to the identifiability of Vanna White as the object of the Samsung commercial’s joke. Identifiability has been a critical part of right of publicity jurisprudence since at least as early as 1974 in Motschenbacher v. R. J. Reynolds Tobacco Co., where an ad’s depiction of a specific racing car was sufficient to identify the indistinguishable driver in the photo as Lothar Motschenbacher.

In Sarver v. Chartier, U.S. Army Sergeant Jeffrey Sarver maintained that he was identifiable as Staff Sergeant William James, the bomb-disposal expert at the center of the award-winning film The Hurt Locker. The defendant film producers acknowledged that they had imported certain aspects of Sarver’s character and episodes of his military service career—along with those of other soldiers—into the Sergeant James character. The court found that Sarver was identifiable, by those who knew him, but that this did not give him a right of publicity claim against the filmmakers because movies are an expressive medium that both entertains and educates by drawing from life, and the public interest in preserving such expression outweighs the private interests of individuals in the commercial value of their publicity rights.

Yet, during the five years preceding the Sarver holding, in a string of cases brought against Electronic Arts (EA) for its football video games’ use of identifiable amateur and retired professional athletes—Hart v. Electronic Arts, Inc.In re NCAA Student-Athlete Name & Likeness Licensing Litigation (Keller); and Davis v. Electronic Arts, Inc.—courts consistently held that the First Amendment did not protect EA from the former athletes’ claims. The plaintiffs argued that even without physical similarity in the avatars or matching jersey numbers, they were identifiable purely by statistics and by position, team, and year. Of course, identifiability alone is not dispositive but only a threshold step to making a prima facie case for right of publicity misappropriation. Increasingly, courts have held that individuals who are readily identifiable may still not be able to overcome defendants’ First Amendment defense if the images of the plaintiffs have been sufficiently transformed.

The Saderup Transformativeness Test

The applicability of transformativeness arose in 2001 when the California Supreme Court in Comedy III Productions, Inc. v. Gary Saderup, Inc., borrowed a page from copyright fair use analysis. The complaint before the court involved not the familiar “use in advertising” that had characterized most publicity cases up to that time but rather an artist’s offering to the public his depiction of the Three Stooges on prints, shirts, coffee mugs, and other commercial merchandise. The artist, Gary Saderup, argued that the items embodying his drawing of the Three Stooges were works of expressive art, entitling him to First Amendment protection against the claims of Comedy III, guardian of the Stooges’ postmortem publicity rights.

Though a case of first impression to the court, the merchandise issue was hardly new to right of publicity jurisprudence. Haelan Laboratories, Inc. v. Topps Chewing Gum, Inc., the 1953 federal court ruling that officially recognized and legitimized the right of publicity, was not about the unauthorized use of an image in advertising but the licensed use of an athlete’s likeness for merchandise, namely baseball cards. But in Haelan, there was no claim that the card or the image on it was protected expression. It was an exclusively licensed commercial use that the court protected against use by a competing card company.

In Saderup, the issue was trickier. Courts have always resisted being called upon to be art critics, but the Saderup court felt that in transformativeness they had something they could apply without the need for professional expertise. To enjoy First Amendment protection, a work had to be expressive, and the presence of artistic expression should in some way transform what would otherwise be nothing more than a rote depiction of its subject, like Saderup’s charcoal drawing of the Three Stooges. Two years later, the same court applied the same test in Winter v. DC Comics to determine that the images of blues rockers Johnny and Edgar Winter were sufficiently transformed to preclude a publicity action against a fantasy-fiction comic book depiction of them as grotesque subterranean creatures called the Autumn brothers.

This made sense to judges everywhere. Despite the fact that it was devised only to decide whether and when a court should consider a First Amendment defense, the Saderup transformativeness test became in the ensuing decade the tool of choice for deciding the actual merits of a right of publicity claim. Courts called upon it in a diverse range of claims and cases, not just those involving images on merchandise, for which the test was originally tailored. In No Doubt v. Activision Publishing, Inc., the court held that a pop music group could prevail on its publicity right claim against a video game maker that had, under license, used images of the band in a game that enabled players to manipulate them into playing other styles of music and the songs of entirely different performers. A crucial part of the court’s analysis in dismissing Activision’s First Amendment defense was that the game depicted the band “doing exactly what they do as celebrities” and therefore was not transformative. The EA courts quickly picked up on this, writing that because the games depict the plaintiffs as football players—“doing exactly what they do as celebrities”—the use is not transformative and the First Amendment defense is not allowed. In a surprisingly short time, the Saderup rule that a showing of transformativeness is sufficient to invoke First Amendment protection has become a rule that failure to show transformativeness is conclusive of right of publicity misappropriation.

Equally Protected?

So why did the makers of The Hurt Locker enjoy First Amendment protection while EA and Activision did not, when both depicted their respective plaintiffs “doing exactly what they did”? No case has held that video games are commercial speech or merchandise or anything but constitutionally protected expressive works. So how can courts hold that movies have a First Amendment right to depict persons doing what they do for a living and an equally protected video game does not? The U.S. Supreme Court’s lone right of publicity outing in Zacchini, which was about what Zacchini did for a living, provides no guidance. The point in Zacchini was not that the news media unfairly benefited from portraying Zacchini doing what he did for a living but that in so doing they preempted him from earning his living that way. Though Samsung’s commercial made a joke about what Vanna White does for a living, it did nothing to impede her ability to do so. And neither did Activision or EA harm the rockers’ and footballers’ ability to make a living.

Zacchini does however hold that First Amendment balancing may be necessary, even where the highest First Amendment standards are invoked. Rogers v. Grimaldi, the 1989 case and test most frequently used to weigh Lanham Act section 43(a) claims against uses of names, likenesses, and trademarks in expressive works, acknowledged that an injunction against an expressive work might be appropriate in some cases. But the Lanham Act seeks to protect the public from confusion or deception and is not concerned with the personal economic rights of the individual. Only the right of publicity seeks to compensate individuals for the unconsented exploitation of the inherent economic value of their names, likenesses, or personas.

The Sarver court recognized the apparent contradiction between its ruling and those in the video game cases, and sought to distinguish Sarver from Keller in this way:

[U]nlike the plaintiffs in Zacchini . . . and Keller, Sarver did not “make the investment required to produce a performance of interest to the public” [citing Zacchini] or invest time and money to build up economic value in a marketable performance or identity [citing Keller]. Rather, Sarver is a private person who lived his life and worked his job. Indeed, while Sarver’s life and story may have proven to be of public interest, Sarver has expressly disavowed the notion that he sought to attract public attention to himself. Neither the journalist who initially told Sarver’s story nor the movie that brought the story to life stole Sarver’s “entire act” or otherwise exploited the economic value of any performance or persona he had worked to develop. The state has no interest in giving Sarver an economic incentive to live his life as he otherwise would.

But this implies that EA’s use of identifiable player avatars in football video games somehow harms the players’ ability to do what they do for a living. If the court means that such use harms the players’ rights to their living by depriving them of a license fee, that begs the question by assuming that EA’s use of the players’ images in a video game is licensable in trying to decide whether EA’s use should require a license.

Only Celebrities Need Apply?

Simplified, the Sarver court’s analysis suggests that the right of publicity is limited to self-made celebrities who have actively exploited the commercial value of their images or likenesses. Many commentators, some courts, and even some state statutes have begun to take the position that only celebrities should be entitled to recover for right of publicity claims—a shocking turn of events for a cause of action that grew not from a celebrity’s right to be compensated for the inherent value of the celebrity’s personality but from an ordinary person’s right not to have the person’s likeness used to profit a business by advertising its products. In Roberson v. Rochester Folding Box Co., a noncelebrity sought remedies for the emotional damage she suffered when a flour company used a photograph of her without her knowledge or permission to advertise its product. At that time, even claims for violation of privacy did not exist, so the court was unable to provide her a remedy. But a remarkably visionary dissent argued that any person has an economic interest, and hence a property right, in whatever value the person’s likeness may have to others, and such value must exist—otherwise, why would a company want to use that likeness in promoting its goods?

Surely no one today believes that a person whose image is used in advertising without permission or compensation should be without a legal remedy merely because that person is not a celebrity. Yet the Sarver court seemed happy to distinguish Sarver from Keller on the basis that the soldier was not known to the public for what he did while the football player was.

Enter the legendary actress Olivia de Havilland, who is a celebrity. She sued the makers of a television miniseries, The Feud, over her portrayal in the fictionalized account of the turbulent rivalry between iconic actresses Joan Crawford and Bette Davis, in which Catherine Zeta-Jones portrays de Havilland in a supporting role. The result was the same as in Sarver: Movies are expressive works protected by the First Amendment. They are neither advertising nor merchandise, so their protection must not yield to right of publicity claims.

In both de Havilland and Sarver, as in the Winter case, the courts dealt with situations in which the plaintiffs were transformed despite remaining identifiable, and they had little trouble in finding that the public interest in First Amendment–protected, freely expressive arts and entertainment media outweighed the narrow commercial interest of the plaintiffs in keeping their own images as they wanted them to be, or in exacting a license fee for their use.

Sarver was rightly decided but incorrectly distinguished by its court, which dared not recognize the emperor’s new clothes worn by the EA and Activision cases. The de Havilland court had the courage of its convictions and noted that the Saderup test is limited to cases involving “merchandise,” which does not include movies (or, for that matter, video games). Moreover, the EA courts’ overly broad and simplistic treatment of transformativeness creates a catch-22 situation: if de Havilland is not “transformed” in The Feud, then her right of publicity has been violated, and if she is transformed, then she has been portrayed in a false light—either of which outcomes would eviscerate the movies, and expressive art in general.

Commentators have noticed a similar anomaly that has developed over the last decade of right of publicity jurisprudence: under the Saderup test, a plaintiff whose image is not transformed or artistically repurposed will overcome a First Amendment defense and prevail on a right of publicity claim; but under the Rogers test, that same plaintiff would usually fail to overcome the First Amendment defense because use of the plaintiff’s image was reasonably related to the work as a whole and unlikely to confuse or deceive the public. For example, in 2013 in Brown v. Electronic Arts, Inc., Jim Brown’s complaint that his image was used to sell EA’s Madden NFL game lost to a First Amendment defense, because the court found that showing a football player playing football was reasonably related to the purposes of a football video game, and the public was unlikely to believe that Brown sponsored or endorsed a game in which his image appeared along with those of many other players.

This is not quite a contradiction, since the Saderup transformativeness test seeks to determine whether the economic value of an individual’s likeness has been exploited, while the Rogers test asks whether the public is likely to be confused. Nevertheless, the fact that the same facts can both defeat a First Amendment defense in one instance and support it in the other encourages right of publicity claimants to accompany their state and common-law claims with a Lanham Act false endorsement claim, knowing that the courts’ increasing reliance on transformativeness pretty much guarantees them a win on one claim or the other. The bottom line, to paraphrase the de Havilland court, is that showing people doing “exactly what they do as celebrities” cannot mean there is no transformativeness. If it does, then every book, play, documentary, or fictionalized drama that portrays or is inspired by a real person can be liable for violating the right of publicity, and we will have a world in which celebrities and noncelebrities alike will dictate how movies and other media can depict them. So either the Saderup test has been misunderstood and misapplied, or it was not the right test to begin with.

Commercial or Not?

But is it possible that, despite the U.S. Supreme Court’s declaration that video games are just as entitled to First Amendment protection as are movies, there is still some difference between video games and movies that explains the disparate results between the movie cases and the video game cases? Could it be that the courts that have encouraged the athletes’ lawsuits and kept them alive do not believe that the Supreme Court meant what it said in Brown v. Entertainment Merchants Ass’n and persist in regarding computer games as a kind of commercial merchandise justifying a lower level of First Amendment protection?

One argument that video games should not be treated the same way as movies is the practice of in-game purchasing. Various performers—both celebrities and noncelebrities—sued the makers of the Fortnite series of video games for “emotes” enabling game players to cause their avatars to break out into dance moves that the plaintiffs had originated and made popular, usually through viral videos. These dances are not the subject of copyright registrations; the argument is that they invoke the personalities of their creators, thus calling them to mind in violation of their rights of publicity. This would seem at first glance to be something that Fortnite had every right to do—except they charge players extra for the emotes. You have to make an in-game payment to be able to make your avatar do the Carlton, the Floss, the Pumpkinhead, and various other quirky dance moves. This commercial purpose certainly improves the right of publicity plaintiffs’ position. However, no such in-game commercialization of any athlete took place in the EA games.

Some argue that the use of avatars based on actual football players works a commercial benefit to EA. Certainly, EA’s games would not sell as well if all of the players in the games were wholly fictitious; and games like Madden NFL and NCAA Football could not exist at all. They might also not exist if EA had to pay license fees to every athlete arguably represented in the games. Does the fact that EA made a deal with the NFL that covered its use of active players invalidate EA’s argument that the law does not require it to obtain licenses from former or retired players? Does the fact that EA made such a deal mean that it legally had to do so? Movie producers regularly enter into deals that are not legally required to gain certain advantages: protection against frivolous or nuisance lawsuits, gaining the cooperation of certain organizations or individuals, or acquiring exclusive rights to materials not available to competitors. EA was entitled to do the same thing with the NFL in order to gain the advantage of cobranding its game series with the NFL and thus obtaining a competitive advantage. The fact that consent to use the names and statistics of active players came with the deal does not necessarily mean that it was a legal requirement or that EA could not use other players’ information without making similar deals.

So did using the likenesses of real football players lend a commercial value to EA? Certainly. Far fewer consumers would buy a computer football game populated solely by fictitious players, without the romance of mixing and matching real teams and players, replete with their statistical performance history. But didn’t the persona of Sergeant Sarver lend a commercial value to the writing and making of The Hurt Locker? Yes, to the extent that Sarver’s experiences and attitudes provided the groundwork for the filmmakers’ vision of the effect of war on the human personality—though other soldiers’ experience could and did also contribute to that vision. Would The Hurt Locker have been as good a film—or, more properly, as commercially successful a film—without Sarver? Did the inclusion of Olivia de Havilland as an incidental character lend The Feud commercial value? Would The Feud have been less popular if it had excluded her character, or substituted a fictitiously named character based on de Havilland?

If these questions can be asked about both movies and video games, why should they matter in the case of video games and not in the case of movies? Should these questions even be asked when we are talking about two entertainment media that the U.S. Supreme Court has held to be entitled to the same level of First Amendment protection?


Today, the law of publicity is more uncertain than ever about whether the right protects only against commercial uses or against expressive uses as well, and whether the right belongs to all individuals, just to celebrities, or to both but with different analysis for celebrities than for what we might call “ordinary folks.” These issues have put right of publicity jurisprudence on course for corrective action, but the U.S. Supreme Court has shown no interest and has denied cert in the publicity cases presented to it since Zacchini. The last of the EA football cases recently settled, but there is no settlement in sight for the lingering contradictions these cases have caused for right of publicity law.


    Robert C. Cumbow is a nonpatent IP partner in the Seattle office of Miller Nash Graham & Dunn. He teaches right of publicity law at the University of Washington School of Law.


    The author thanks Leila Javanshir for research and technical editing assistance.