Souza v. Exotic Island Enterprises, Inc., No. 18-CV-9448
(KMK), 2021 WL 3501162 (S.D.N.Y. Aug. 9, 2021)
Another in the burgeoning genre of models suing “adult”
clubs for using unauthorized images in online ads for the clubs. As this case
shows, the Second Circuit is not as favorable a jurisdiction for these claims
as some others, given the short ROP limitations period and the skepticism about
non-ROP claims. Lexmark has crept into §43(a)(1)(A) via false
endorsement; it will be interesting to see whether courts recognize that other
trademark claims are likewise subject to a proximate cause requirement by that
logic.
Facts in the light most favorable to the plaintiffs: Each of
the plaintiffs has a significant number of followers on various social media
platforms, ranging from greater than ten thousand to several million, and most
are “considered social media influencers.” “Because they rely on their
reputation to get work, Plaintiffs are selective about the jobs they take, and
exercise ‘complete control’ over the use of their images and likenesses,”
especially given the persistence of images online. Mansion runs a club where
nude or semi-nude women offer entertainment; “[p]romotions containing
Plaintiffs’ images were without Plaintiffs’ permission posted to Mansion’s
social media pages.” Plaintiffs sued for false advertising and false
endorsement under the Lanham Act, violation of their right to publicity,
deceptive trade practices under New York GBL Section 349, and defamation.
The court was guided by Electra v. 59 Murray Enterprises,
Inc., 987 F.3d 233 (2d Cir. 2021), which considered all these claims except for
false advertising. Electra found that, even where models had transferred
all rights in the photos to a third party, they could still bring ROP claims:
the clubs didn’t claim to be beneficiaries of those agreements and the releases
didn’t constitute the necessary “written consent” for defendants’ uses, though “the
releases could provide a defense in an action against the releasees or those who
could assert lawful use by reason of assignment or license.” However, false
endorsement claims failed because they didn’t prove that they had the kind of
fame that meant that their appearance in an ad was an endorsement, as opposed
to an appearance by a model. There was no deceptive practices claim under
Section 349 because the conduct wasn’t consumer-oriented; this was “a private
dispute over a private injury visited on the individuals portrayed in the
photographs.” Defamation claims also failed: First, the ads didn’t
unambiguously indicate that the models would be appearing at the clubs and
might just indicate that they were in the ads, and second (and relatedly), plaintiffs
failed to show actual malice; at most, they failed to investigate whether
third-party contractors had the rights to the images, but that’s not actual
malice.
Given Electra, plaintiffs withdrew their GBL Section
349 and defamation claims.
False endorsement: Failed for similar reasons to the claims
in Electra. You can be really popular on social media without being
recognizable enough for the “strength” factor to favor likely confusion. Electra
quoted with approval a district court’s reasoning that “[t]he misappropriation
of a completely anonymous face could not form the basis for a false endorsement
claim, because consumers would not infer that an unknown model was ‘endorsing’
a product, as opposed to lending her image to a company for a fee.”
Plaintiffs didn’t show sufficient evidence of recognition.
Their affidavits that “[o]n any given day, regardless of where [they are] at,
[they are] recognized by complete strangers and [their] fans who follow [them]
on social media” were vague and conclusory. And their expert report was not good.
Martin Buncher’s putative testimony was based on a survey of 812 people who
were at least 21 years old living in the metropolitan area around Mansion, and
who had patronized a “Bikini Bar/Gentlemen’s Club/Strip Club” in the two years
prior to taking the survey. This survey showed that “almost half of the
respondents felt they recognized ... Plaintiff[s’] images in the ads in some
manner having seen them prior to this research.” The court excluded this as
unreliable.
First, Buncher “used copies of the images annexed to the
Complaint with ... [P]laintiffs’ names removed from the top, which resulted in
large parts of their faces and heads being removed.” The court found that “[s]urvey
respondents are highly unlikely to be able to accurately identify [plaintiffs]
based on photographs that do not show their face above their nose.” Yet, oddly,
his study “shows relatively uniform levels of recognition across the images of
all eight Plaintiffs, including those that show a Plaintiff’s face and those
that do not,” from 55-43% recognition. Perhaps this flaw was related to the
lack of a control group. Electra likewise rejected a Buncher survey and
his explanation that his survey was “a communications study, not a consumer
confusion study” as “insufficient to set aside the district court’s conclusion
that the Buncher Report was fatally flawed.
That the results bunch around 50%
recognition for each Plaintiff, regardless of whether her whole face is shown,
supports the view that many respondents were guessing. Another possibility is
that respondents—generally agreeable people who agreed to participate in the
survey—were yea-saying. Because Buncher made no effort to control for these
possibilities, he lacks good grounds for his conclusion that Plaintiffs were
recognizable, and the Court will not permit him to testify to this point based
on these survey questions.
Separately, the recognition questions were independently
defective because they “provided no opportunity for respondents either to
express uncertainty or to provide the identity of the [p]laintiff.” “As a
result, the Court has no way to verify whether respondents truly recognized any
of the Plaintiffs.” Yeah, that seems bad. Indeed, the survey responses didn’t
identify any plaintiff by name. That wasn’t absolutely required; a number of the
plaintiffs had modeled for Playboy, and there were at least 11 references to
Playboy in the responses, but “Playboy itself is a strong brand. No reasonable
jury could find that these references suggest that respondents recognized
Plaintiffs from their work with Playboy.”
Buncher has been allowed to testify with similar surveys in
other district courts, but those courts are in circuits that like to admit
surveys and then discount their probative value, almost whatever their flaws.
The Second Circuit is more discerning.
It was not enough for each plaintiff to be a “successful
model” with “substantial followers on their social media accounts.” In Electra,
while the successful Carmen Electra earned over $5 million modeling between
2009 and 2012, the unsuccessful plaintiffs made annual modeling incomes ranging
“from $400 ... to $92,000,” and those amounts weren’t significant enough to
favor a finding of recognizability. Electra
had “not just appeared in popular movies and television shows, but had regular and
starring roles in them.” While the other plaintiffs had “participated in
promotional campaigns for a wide variety of brands and appeared in magazines,
TV shows, and movies, their resumes [were] devoid of evidence that they
actually garnered recognition for any of their appearances.”
Here, plaintiffs didn’t establish their income, which was
their burden to do if they wanted it to weigh in their favor. In discovery, they
produced evidence that ranged from $107,000 in one year, when the plaintiff
earned $100,000 as Playmate of the Year, to “up to” $7000 for two roles. Nor
was the other evidence of prominence strong; though they provided “an extensive
list of the magazines and ad campaigns in which they were featured,” “[s]imply
listing brands or magazine titles is insufficient.” They were also seen in
additional roles, but their “resumes are devoid of evidence that they actually
garnered recognition for any of their appearances.” Only two showed a “starring
role[ ]” in something, but neither made any showing that these productions were
“popular” or that they had “regular” starring roles. No reasonable jury could
find that these facts supported strength. The court reasoned similarly with
respect to social media followings at the time the images at issue were
published.
Given their relatively weak marks, the absence of actual
confusion evidence was significant.
Plaintiffs relied on the Buncher report, which concluded
that “62% of survey respondents believed each Plaintiff had some affiliation,
connection[,] or association with Mansion; 75% believed Plaintiffs agreed to
sponsor, promote[,] or endorse Mansion; and 76% of respondents believed
Plaintiffs approved Mansion’s use of their images.” But that was excluded.
Buncher showed photos from the social media posts at issue,
and asked:
Considering that these are actually
real women shown in these ads and not just fictitious drawings, please indicate
using your strangest (sic) impression for each pair of opposing statements the
one you think is true based on your personal feelings. Remember, we want your
response based only on these ads you are seeing, and nothing else you might
have seen or heard previously.
• All of the women shown in these
ads have some affiliation, connection or association with those clubs in whose
ad they appear
• All of the women shown in these ads
do not have any affiliation, connection or association with those clubs in
whose ad they appear
[similar binary "all of the women" sponsorship/endorsement/approval/participation in the club events/women were paid to be in the ad questions]
It’s like a list of what not to do in surveys! There was no
anti-guessing instruction or “not sure” or “no opinion” options. These omissions
made the survey leading. Each question forced the respondent into binary
answers about “all” of the women. But there were three relevant possibilities:
(1) the ads suggest affiliation; (2) the ads suggest lack of affiliation (etc);
(3) the ads don’t suggest anything one way or another about affiliation. “By
failing to provide the third option, Buncher’s survey led respondents to
answers favoring Plaintiffs.” The court pointed out that (3) was the most
obvious answer for respondents who weren’t confused. “It is logically difficult
to see a person in an ad and draw the affirmative conclusion that she has no
affiliation whatsoever with the advertiser.” So a respondent who wasn’t confused
would especially need option (3).
At most, the evidence showed a possibility, not a
probability, of confusion.
Bad faith: In Electra, the Second Circuit held that
the plaintiffs “failed to establish ... bad faith” where “the record merely
show[ed] that [the defendants] failed to investigate whether the third-party
contractor responsible for the advertisements secured legal rights to use [the
plaintiffs’] pictures in the promotional images—not that [the defendants]
intended to use the pictures without legal right to do so.” So too here.
Even if a reasonable jury could find that the remaining Polaroid
factors favored plaintiffs, they’d still lose, as in Electra.
Plaintiffs argued that they could still win on affiliation
confusion even if they failed to show endorsement confusion. First, Electra
was controlling. Second, “Plaintiffs’ distinction is immaterial.” It’s all Polaroid.
“The Court’s analysis applies with equal force to the claim that consumers were
likely confused about Plaintiffs’ association with Mansion as it does to the
claim that consumers were likely confused about Plaintiffs’ endorsement of
Mansion.”
False advertising: Plaintiffs didn’t come within the zone of
interests protected by §43(a)(1)(B). Their alleged harm, lost licensing fees,
was not the requisite type of harm—lost business to the defendants.The plaintiffs
argued that they directly competed with defendants, because “both seek to
attract customers and vie for the same dollar via the use of an image of a
beautiful woman.” But while they share a marketing strategy, Plaintiffs and
Defendants “each perform different functions within the marketplace.” “That two
products are sufficiently related that consumers could be confused about the
association between them does not suggest that these two products are direct
competitors.”
They didn’t show cognizable injury by asserting a right to
compensation from use of their images. “This assertion misunderstands the
nature of a false advertising claim, which is focused on how false assertions
in the market harm a plaintiff’s present and future prospects.” Lost wages “are
not within the zone of interests that the Lanham Act protects.”
Plaintiffs also alleged that use of their images hurts their
reputation and business. But the burden was on them to show that this was true,
and they did not. Evidence of lost opportunities wouldn’t be required if they
could show “other evidence of reputational or competitive harm,” but they didn’t.
Plaintiffs relied on Lexmark for the proposition that
“when a party claims reputational injury from disparagement, competition is not
required for proximate cause” and that “a defendant who seeks to promote his
own interests by telling a known falsehood to or about the plaintiff or his
product may be said to have proximately caused the plaintiff’s harm.” But that
was about proximate cause, not the zone of interests.
Finally, they argued that “a court may award a defendant’s
profits solely upon a finding that the defendant fraudulently used the plaintiff’s
mark.” But this rule requires first that a Lanham Act violation has been
established.
Finally, many but not all of the state ROP claims were
barred by a one-year statue of limitations (as established in Electra).
The court declined to exercise jurisdiction over the two remaining plaintiffs’
claims, which is an extra yikes. First, the court found that it lacked diversity
jurisdiction since the amount in controversy for each plaintiff didn’t exceed
$75,000. And because it was kicking out the federal claims, it declined to exercise
supplemental jurisdiction despite how far the litigation had progressed.
No comments:
Post a Comment