My presentation, galloping across a bunch of developments. Slides here.
Begin with legislative action: Trademark Modernization Act, Which
introduces three of the big themes of the year: (1) trademark use, (2) the harm
of infringement, and (3) the role of the First Amendment in limiting the scope
of trademark rights. As harm showings become less important, both trademark use
and the First Amendment become relatively more important as the remaining
limits on liability once a theory of how consumers might possibly be confused
has been introduced.
Briefly, as to trademark use, under the TMA, Expedited
cancellation and nonuse proceedings are on the way, with the PTO finalizing
rules—still remains to be seen how and how often they will be used to clear up
deadwood on the register.
Right now the expedited cancellation and use proceedings
represent just one aspect of how attention to trademark use has become more
important as perhaps the biggest remaining limit on the scope of trademark
rights other than the First Amendment—Mark Lemley has called the question of
whether the defendant is making trademark use “step zero” in the infringement
inquiry, and whether the plaintiff is making trademark use is the flip side of
that question.
Indeed, the PTO has increased its focus on whether the use
an applicant is making is trademark use, as opposed to ornamental or
informational use, in its registration decisions. [slide] Annual number of TTAB
decisions under failure to function and related categories: 2000-2020—you can
see an increase with a fall in 2020 due to the fact of 2020; 2021 looks to have
regained the momentum of failure to function.
Professor Alexandra Roberts has written an excellent recent
article on this, Trademark Failure to Function.
The PTO’s attention to failure to function as a mark is likely to
intensify, in part because some of the former grounds for denying a
registration are now unconstitutional. But even if the n-word isn’t
unregistrable because it’s scandalous or disparaging, it may still be
unregistrable because it already has so much expressive meaning that it’s
simply incapable of adding a trademark function.
failure to function can be significant in ordinary cases
with uncontroversial subject matter. In this recent case, Epic sought to
register this design for "downloadable video game software"—the Loot
Llama in Fortnite, which serves as a resource cache for players. The examiner
reasoned that the llama appears only as a character in the game, not as a
source identifier—the player clicks on it to get resources. The Board agreed: “The
repetition of the llama on the specimen in varying sizes, portions and vividity
detracts from Applicant’s claim that the single llama depicted in its drawing
would be recognized as its mark. In addition, consumers considering the source
of the webpage can look to the term FORTNITE which appears at the top left
portion of the specimen.” Thus, even though it would be possible to have a
registrable llama logo that functioned as a trademark, that’s not what Epic
had.
The past year also saw some media attention to another
puzzle of use as a trademark: registrations that are for unpronounceable,
unmemorable series of letters, not translated from another language, whose
registration apparently serves the lucrative purpose of allowing the registrant
to do better in Amazon search results because of the benefits Amazon gives to
registered trademark owners. My former student Grace McLaughlin has written an
excellent note about the fact that these putative trademarks don’t serve human
trademark functions—it’s very hard to remember them or distinguish one random
string from another random string in terms of knowing what you’ve seen
before—and has proposed some possible responses from the PTO.
At a more theoretical level, these algorithmically optimized
registrations help show why applications have spiked so much—other important
institutions, like Amazon and the Chinese government, are using the PTO for
their own purposes. Amazon gives registrants special powers over product pages,
while there reportedly have been various Chinese subsidies for entities that
secure US registrations as evidence of their productivity. But these other
insitutions don’t bear all the costs of effectuating their own purposes by
using US registrations as a marker, and that contributes to the load on the
PTO. At this point, as Barton Beebe and
Jeanne Fromer have demonstrated with empirical research, crowding on the TM
register is a serious problem, both in the US and Europe, and nonuse
proceedings may not be able to do very much on the back end to deal with that.
Moving to the defense side, Lack of use as a trademark
doesn’t just limit attempts to claim TM rights. It can also limit attempts to
impose liability on defendants who aren’t using a symbol as a mark. In the
Redbubble case, the Ninth Circuit affirmed the rejection of LTTB’s trademark
claims against Redbubble for allowing artists to sell shirts and other items
that used the phrase LTTB as ornamentation. The Ninth Circuit framed this as a
matter of aesthetic functionality: giving exclusive use of the phrase as a
design for the front of shirts would put other competitors at a “significant non-reputation-related
disadvantage.” This analysis does make some sense, but raises the question of
what the how to identify what counts as a “significant” disadvantage. Another
way to put it is that aesthetic functionality requires you to have an
understanding of the definition of the market in which other clothing makers
should be free to compete. But the reason the phrase is aesthetically
functional as a prominent design on clothing and bags is that it is attractive
because of its message, not because of its source indicating significance, and
that may be more satisfyingly framed as a matter of ornamentality.
Interestingly, the court didn’t hold that the registered LTTB
marks were per se invalid—they could still be protectable if used on clothing
tags. That sounds like defense-side functionality: the defendant can win if its
use is functional, even if the plaintiff’s use is not functional.
Similarly, Sulzer Mixpac AG v. A&N Trading Co., No.
19-2951 (2d Cir. Feb. 18, 2021), held functional the defendant’s informational
use of colored mixing tips for dental equipment to show which mixing tip should
be used with which materials. The colors were arbitrary when initially chosen,
but the market had learned that colors corresponded to specific diameters and
lengths as required for specific mixing ratios. Informational use of color is
not trademark use for purposes of gaining trademark rights, and likewise
informational use by competitors is functional, meaning that claimed trademarks
over the informational matter are invalid.
Trademark use can show up in some surprising places. The
Fourth Circuit recently endorsed the idea that putting something in the
trademark “spot” on a product doesn’t mean it’s a trademark—in the course of
finding that certain uses didn’t have a significant effect on the market.
Perhaps ironically, the result was to give the plaintiff
claimant more rights in VAGI- as a formative for products for use in the
vagina, because products with noticeable market presence were disregarded
because their name was generic.
One lesson from these examples: use as a trademark is a
concept so useful that even after courts of appeals rejected a decade ago it as
a separate limit in cases involving keyword advertising, it is still constantly
being reinvented and applied to solve otherwise difficult problems: trademark
use is a simple way to explain why we shouldn’t do a complicated and expensive
and error-prone likely confusion analysis when defendants are engaging in
behavior that on its face seems pro-competitive and not likely to result in
material harm to consumers, despite the theoretical possibility of confusion.
Second big theme is harm and its relationship to the
scope of TM rights: Likely success on the merits, or success on the merits, now
justifies a presumption of irreparable harm for preliminary or permanent
injunctive relief, restoring the rule that many circuits used before eBay v
Mercexchange, though the knowledge of this change is percolating through
the lower courts.
Things we don’t know yet: will things like voluntary
cessation under conditions where resumption of the challenged conduct would be
difficult be able to rebut the presumption? Before eBay, lots of courts
held that delay rebutted the presumption, and I expect that to resurrect, but
we will have to see—indeed, we don’t even know if the presumption has any
weight or simply shifts the burden of production. One commentator has written: Based
on Congress’s failure to address the issue, it may well be that Federal Rule of
Evidence 301 provides the default rule and that the shift is merely one of the
burden of production. However, how you produce evidence of lack of harm is an
open question even if the presumption is a very thin one.
This change in the ease of getting relief interacts with the
substantive scope of TM law in ways that may prove challenging.
Infringement keeps getting bigger and bigger. In fact, one
reason that this year in review says nothing about trademark dilution is that
trademark dilution protection for famous marks doesn’t do much to expand
trademark owners’ rights, since all conduct that plausibly causes dilution
plausibly causes confusion as confusion has been expansively defined.
Of particular note, in the past year, IIC returned from the
dead. For about a decade, courts had realized that IIC had gone way too far,
and had expanded liability in ways that didn’t protect consumers and
facilitated anticompetitive claims about false advertising. But we’ve seen a
swing back towards requiring a jury to determine each case, which increases the
risks for comparative advertising a lot
In one case, The Eighth Circuit had previously held that IIC
didn’t make sense where any confusion to sophisticated purchasers of expensive
goods would be quickly resolved. However, in Select Comfort, a case about
keyword and other online advertising, it held that because there was a question
of fact about how sophisticated mattress consumers are, the initial interest
confusion theory should be submitted to the jury. The court or appeals also
commented that relying on IIC might change the available damages and relief—but
how exactly?
Relatedly, the 2d Circuit reversed the FTC’s victory on
antitrust grounds against 1-800-Contacts’ anticompetitive settlements with
competitors that precluded them from running ads triggered by searches for
1-800-contacts, whether or not the ads themselves were confusing; because
1-800-contacts articulated its interest as one in protecting its trademark, the
Second Circuit held that its settlements were legitimate, even though they
deprived consumers of useful and nonconfusing comparative advertising, and even
though these settlements even precluded competitors from buying the word
“contacts” generally without including an exclusion for 1-800-contacts. This
ruling helped resurrect trademark owners’ interests in threatening others
engaged in competitive keyword advertising, despite no persuasive evidence that
keyword-triggered confuse consumers unless they have confusing ad text.
The Bayer v. Belmora case also continues to influence
courts, which are increasingly likely to say that 43(a) confusion claims don’t
require having a valid mark—but current courts have lost sight of whether there
are limits on that principle. There are different reasons one might not
have a valid mark. In particular, there are cases where it would be possible to
have a registered trademark in the claimed subject matter, but the plaintiff doesn’t—say,
a descriptive term without secondary meaning. There are other cases where it
would not be possible to have a registered trademark in the claimed subject
matter, and that can be meaningfully different depending on why it’s not
possible. For example, one might have de facto secondary meaning in a generic
term, which as then-Judge Ginsburg held several decades ago could justify
limited relief requiring the defendant to distinguish its use from that of the
plaintiff, but could not justify enjoining the defendant from using the generic
term, in sharp contrast to the usual remedies available in a trademark case. Or
one might have US secondary meaning in a trademark valid in another country but
not in use in the US, so the claimant could not register a trademark in the US.
That imported secondary meaning—usually imported via immigrants who remember a
mark from back home—might justify either restrictions on how a US defendant
with US rights could use the mark in the US or invalidation of the US rights as
a whole and a total prohibition on the defendant’s use—this is the question in
the Bayer v. Belmora case; the Fourth Circuit allowed Bayer to seek the
complete elimination of Belmora’s ability to use the challenged mark, though
Belmora has sought cert on this question.
The Zamfir case is then an example of how 43(a)’s detachment
from “trademark” as such is diffusing across the country. Zamfir is a
researcher in the field of cryptoeconomics and distributed systems; he was allegedly
one of two lead researchers of the proof-of-stake blockchain protocols known as
Casper. A blockchain company using these protocols adopted the name CasperLabs
and secured a registration; the court allowed Zamfir to bring a §43(a) claim
without proving that he had a valid trademark interest, because it reasoned
that ownership of a valid mark isn’t a requirement under §43(a)—skipping
straight over whether he had a commercial interest at least analogous to a
valid trademark! However, the court then reasoned that he would face a higher
barrier to prevail in his claims if CasperLabs had a valid registration, which
would at least weaken the strength of the mark as it related to Zamfir.
Beasley v. Howard, (3d Cir. Sept. 17, 2021)), is another
case about the breadth of §43(a). It began with trademark cancellation
proceedings before the TTAB, where the plaintiff failed to secure cancellation
of the defendant’s mark. The Third Circuit ruled that TTAB proceedings don’t
have claim preclusive effect against trademark infringement lawsuits in federal
district courts, because more remedies are available in federal court. What
this means is that you can bring a 43(a) claim against a defendant who has a
federal registration and still claim that your unregistered rights trump the
registration. To be clear, I don’t think this is a wrong result, but I invoke
it here to make the point that trademark rights can be quite amorphous whether
or not a registration is involved.
Even cases that deny a finding of infringement are cases
that might not have been worth bringing a few decades ago b/c doctrine is now
so broad:
For example, For over a century, PTRA has hosted the Rose
Parade and Rose Bowl Game. Under its agreements with the City of Pasadena, PTRA
is the exclusive owner of the Rose Bowl Game trademark. They worked together
for many years but Covid disrupted the relationship as PTRA decided to move the
Rose Bowl to a state that wasn’t as worried about Covid. It then sued the City
for this Instagram post, alleging that it would cause confusion about
affiliation or sponsorship.
The court found that this was nominative fair use, which
provides a defense in the Ninth Circuit when a trademarked product or service
isn’t readily identifable without use of the mark, the use is no greater than
necessary, and the defendant does nothing else to suggest confusion. The court
found those requirements were satisfied. Pasadena posted from its Instagram
account, @rosebowlstadium. “It did not use any express or implied language of
sponsorship or endorsement or ‘tag’ Plaintiff’s accounts.” Conclusory
allegations of possible confusion were “insufficient to plead there was a
suggestion of association or sponsorship.”
Some other notable edge cases: Big Ligas v. Yoo: Lanham Act claim
against a lawyer negotiating on behalf of her client, allegedly using her
client’s name in commerce to promote his songwriting and recording services in
violation of the client’s exclusivity agreement with the plaintiff. The claim
failed because, the court said, negotiating on behalf of a client isn’t use in
commerce. This is obviously wrong in one sense—Congress meant to define
commerce broadly and engaging in economic transactions crossing state lines is
commerce by any ordinary measure—but it’s right in the larger sense that of
course it shouldn’t violate the Lanham Act to negotiate on behalf of one’s
client even if the client is in fact breaching its contractual obligations.
What’s notable is the paucity of doctrinal tools available to the court to find
no Lanham Act coverage, leading it to just pretend that this wasn’t use in
commerce.
Wakefern Food Corp. v. Marchese, 2021 WL 3783259, No.
2:20-cv-15949-WJM-MF (D.N.J. Aug. 26, 2021) Wakefern, the largest retailer-owned
supermarket coop in the US, sued Marchese for attempting “to lease commercial
real estate in violation of the Lanham Act … and New Jersey common law.” This
case was also resolved by holding that this wasn’t a use in commerce. The
plaintiff didn’t allege that there was ever any public infringing use, and the
allegations “suggest that Marchese made false representations to the broker in
order to take advantage of the broker’s services rather than to sell or promote
his own.”
Proactive Environmental Products Int’l, LLC v. Pine
Environmental Servs., LLC, No. 8:21-cv-250-CEH-CPT, 2021 WL 3025481 (M.D. Fla.
May 20, 2021) (R&R) renting existing equipment; Pine was formerly one of
Proactive’s licensed distributors for the pumps. The principal issue was whether
Pine can continue to rent Proactive’s pumps to its customers despite the
termination of that license, particularly where Pine has made or will make
repairs to the rented pumps.” In a system with a working first sale doctrine,
this would be an easy question, but TM’s definition of confusion has expanded
so far that it takes the judge work to conclude that Proactive probably
shouldn’t get an injunction. Ultimately, the court reasoned that Pine disclosed
that it wasn’t a licensed distributor, and that people expect rental equipment
to suffer wear and tear so they wouldn’t attribute any flaws in the rental
equipment to Proactive. If the result were otherwise, then Uber drivers could
be vulnerable to infringement suits by carmakers who alleged that the drivers
were driving around in dilapidated vehicles, and you wouldn’t be able to resell
a used car that had been repaired after an accident without infringing the
original manufacturer’s trademark.
quick excursion on first sale: The doctrine has long allowed
resale of legitimately produced goods, but courts have cut back on it
substantially where there have been alterations or allegations that defendants
held themselves out as authorized distributors—this year many of those cases
involved claims about N95 masks allegedly produced by 3M. The cause of first
sale was not advanced by those cases, which often involved allegations of
counterfeiting alongside the allegations that the defendants held themselves
out as authorized distributors. Where they were merely reselling legitimate
goods without making explicit claims to be authorized distributors, however,
first sale should protect that conduct—but first sale is not guaranteed to do
so under current doctrine.
The Second Circuit, in Hamilton International Ltd. v. Vortic
LLC, No. 20-3369-cv (2d Cir. Sept. 14, 2021), recently affirmed a finding that
vintage Hamilton pocketwatch movements converted to modern wristwatches were
not likely to cause confusion under the major first sale precedents. At most,
the Second Circuit refused to take further bites out of the doctrine as applied
to refurbished goods, and its emphasis on consumer sophistication may even make
things harder for Etsy jewelry makers. But the case’s larger significance may be
that the court apparently joined the Ninth Circuit in requiring likely
confusion before counterfeiting can be found.
Another interesting limit case:
D posting as an ASU student on Instagram. “Although it is
not uncommon for universities to attempt to appeal to students by imitating
their vernacular, no university would drop the f-bomb in an official party
invitation,” and a reasonable consumer would not have thought ASU was inviting
them to get drunk. Interesting to consider the language of the court, which
relies on simple assertion to explain why liability can’t be as broad as ASU
wanted, because there aren’t really doctrinal tools available:
The court stated: “[m]ore broadly, it cannot be the case
that every social media post written by a college student that happens to use
the school’s colors and/or logo in the post, and identifies the school’s
location as the location of the poster, creates initial interest confusion and
qualifies as an actionable trademark violation.” ASU is appealing to the Ninth
Circuit.
Final comment to make before I transition to part three is
about Art. III standing: In the recent TransUnion case, the Supreme Court held
that Congress can’t create standing just by enacting a law allowing a plaintiff
to sue without a real injury. Big question that recent cases are starting to
surface: Is nonharmful confusion that just amounts to free riding--unjust
enrichment to D without damage to P—sufficiently similar to existing common law
causes of action to be promoted to Article III-qualifying injury after
TransUnion? If courts notice this issue, which is a big if, it could create potential
problems for dilution, IIC and post-sale confusion, as well as false
endorsement theories, all of which are fundamentally about unjust enrichment
rather than any real risk of harm to the plaintiff. Example from recent case:
Abrahams v. Simplify Compliance, LLC, 2021 WL 1197732, No.
19-3009 (RDM) (D.D.C. Mar. 30, 2021)
Plaintiff Daniel Abrahams
formerly contracted with a publisher to author a series related to the Fair
Labor Standards Act. The publisher sold the publication rights and they ended
up in the hands of an entity that allegedly refused to pay him any fees or
royalties but continued sell his publications. He alleged that holding him out
to the public as the editor violated the Lanham Act. The court applied the
Lexmark zone of interests and proximate cause requirements, reasoning that “a
plaintiff may not prevail on a false association claim without alleging a
commercial injury.” But Abrahams failed to do so. It was insufficient to allege
that Simplify “has deprived [him] of the fundamental value of his name and
abilities with regard to editing the publication,” or that his inability to
control the quality of the work “depriv[ed] him of the ability to maintain his
reputation and standing in the marketplace.”
This case is interesting to me because
in standard trademark claims courts routinely accept similar allegations as
sufficient, and under the TMA lost control over reputation is now considered
irreparable harm—without ever having to be proved to be harm in the first
place! The court noted that Abrahams didn’t alleged that the publications were
low quality or that he disagreed with their contents. Nor did he allege any
lost business opportunities as the result of their presence on the market.
Although he alleged that confusion was likely, he did not allege that the
confusion would cause him a specific commercial injury.
Relatedly, Ice Cube’s Lanham Act
and right of publicity claim against Robinhood was dismissed because Robinhood
used an image from a movie in which he appears and a version of his
catchphrase, but only in a finance newsletter, which was noncommercial speech
rather than conventional advertising. The complaint was then refiled to allege
only a Lanham Act §43(a) claim. The renewed motion to dismiss points out that
the complaint alleges that Robinhood was unjustly enriched, but does not allege
that Ice Cube suffered any actual harm, bringing us back to the questions of standing.
Another thing that is likely to
ramp up the potential conflict between trademark law and Article III injury is
the continuing fallout from the recent Romag case, in which the Supreme Court
found that willfulness was not required to award disgorgement to successful
trademark plaintiffs. A few cases this year seem to suggest that disgorgement
should routinely be available to trademark plaintiffs—but disgorgement of
course is about the benefit to the defendant, not the harm to the plaintiff. A
plaintiff seeking disgorgement need not under current rules develop any
evidence that it was actually harmed by any confusion—which makes harm even
less relevant to TM cases in ways that seem to conflict with the Supreme
Court’s statements about constitutional standing.
Third and finally in my thematic overview, Congress
expressed its endorsement of Rogers v. Grimaldi as a test protecting
free speech against trademark liability. Under Rogers, uses of
trademarks in noncommercial speech are not subject to trademark liability
unless they are not artistically relevant to the speech or are explicitly
misleading about the source, sponsorship or endorsement of the speech. That
means that uses that merely imply that the trademark owner has endorsed the
speech are not actionable, even if there is consumer survey evidence suggesting
a relatively high level of confusion. This move is necessary because if you ask
consumers whether a trademark owner depicted in a fictional or factual work
consented to or approved of its appearance, a shockingly high number are likely
to say yes, and courts are unwilling to give trademark owners that much control
over public discourse.
The legislative history of the TMA represents Congress’s
first endorsement of Rogers, though only in legislative history and not in
statutory text the way that nominative fair use is now identified in the statutory
text as a defense to dilution.
It’s not clear whether there are enough judges left who care
about legislative history for this to matter; I would be surprised if the
Supreme Court gave any weight to this kind of statement if the issue reached
it.
But the Court turned down one opportunity to articulate a
rule for the trademark infringement/First Amendment interface this year—it
denied cert in the VIP v. Jack Daniel’s case, in which INTA and other large
trademark owners had invited the Court to adopt Rogers, but not for dog
toys. Denials of cert are not necessarily very meaningful, but in my view
taking up a Rogers case would require the Court to engage with extremely
difficult commercial speech questions that it would likely prefer to avoid.
A few Examples of Rogers’ breadth in protecting
noncommercial speech:
Miller v. Easy Day Studios Pty Ltd, 2021 WL 4209205, No.
20cv02187-LAB-DEB (S.D. Cal. Sept. 16, 2021): found that Rogers protected the
use of a skateboarder character in the game who strongly resembled a pro
skateboarder, who had actually allowed the game to do motion capture of him
allegedly on the representation that he wouldn’t appear in the game. Despite
the alleged contractual agreement, he didn’t have a Lanham Act false
endorsement claim because his appearance was artistically relevant to a
realistic skateboard game and there was no explicit claim of endorsement
Punchbowl, Inc. v. AJ Press LLC, --- F.Supp.3d ----, 2021 WL
3356848, No. 21-cv-03010-SVW-MAR (C.D. Cal. Jul. 16, 2021). AJ operates
“Punchbowl News,” “an online news publication that provides newsletters,
podcasts, and videos in the fields of government, politics, public policy, and
current events.” It uses “Punchbowl” because that is the Secret Service
nickname for the U.S. Capitol (a fact that Punchbowl didn’t contest), and its
logo also invokes the Capitol building. Plaintiff Punchbowl “is a technology
company that develops online communications solutions for consumers, including
online event invitations and greetings cards, with a focus on celebrations,
holidays, and events.” Punchbowl sued AJ for trademark infringement and related
claims.
Rogers precluded liability: There was artistic
relevance because the term “Punchbowl” had political and geographic relevance
to the content of Punchbowl News. Question: is a political newsletter really
artistic? [One of the issues running underneath the disputes over Rogers is
what courts means when they say there’s a special test for ‘expressive’ or
‘artistic’ works. Of course, there are lots of expressive works that are purely
commercial, like standard advertising. What Rogers really is, is a test for
noncommercial speech: when the product itself being sold is speech, like the
New York Times or a painting or sculpture, as opposed to commercial speech,
which is speech that is used to sell a separate product or service distinct
from the speech itself.
Although the Ninth Circuit has complicated Rogers analysis
for cases in which the defendant competes directly with the plaintiff, no
reasonable juror could find that defendant failed to add its own expressive
content to the work beyond the mark itself.
Indeed, the recent Dr. Seuss case in the Ninth Circuit found
that the Rogers test applied and precluded trademark liability even when the
defendant’s work was not a copyright fair use, and the work used Seussian font,
Seussian illustration style, and a similar title. Both parties used marks for
graduation-themed books but the defendant added expressive content to the work
beyond the mark itself, so Rogers protected this book from trademark infringement
claims.
People do keep pushing the boundaries, including for
artistic reasons. This a Tom Sachs painting from his recent series of brand
paintings. Note that the only attribution on the painting is a TM claim
asserting Reeses’ TM rights. Is this explicitly misleading? Even if it is
explicitly misleading, could it be actionably confusing?
One limit on Rogers: it’s really not for ordinary
consumer goods that are quite detachable from their labels, like alcohol. Here,
the defendant’s Rogers defense failed as inapplicable to product labels,
although the plaintiff, which owned rights in the TV show Peaky Blinders,
didn’t show that it was entitled to a preliminary injunction, in part because
of its delay
Finally, to contrast with Rogers, here’s a case
involving political speech where the Eleventh Circuit found infringement
liability over a vigorous dissent. The case is complicated by procedural
issues, but still quite striking in its use of law to suppress political speech.
The district court found that the defendant, a nonprofit that endorsed
political candidates, was liable to its counterpart AGG for infringement;
joined CBG’s principal Darleen Jacobs post-judgment; and awarded attorneys’
fees to AGG. The court of appeals affirms over a dissent that would have held
that the First Amendment precluded application of the Lanham Act to political
speech.
The key passage in the majority opinion is: “even if
Coalition’s speech is rightly considered noncommercial speech, this Court has
not previously held that § 32(1) of the Lanham Act, the section at issue here,
applies only to commercial speech.” In a footnote, the majority acknowledged
that it has held that §43(a) applies only to commercial use, but it has not
extended that holding to §32. (I will note: There is no language in §32 that in
any way could be considered broader than §43(a) in its coverage of political
speech.) Also, the court of appeals pointed out that the Second Circuit has
found that §32 applies to “[a] political organization that adopts a platform
and endorses candidates under a trade name.” United We Stand Am., Inc. v.
United We Stand Am. N.Y., Inc., 128 F.3d 86 (2d Cir. 1997).
Judge Dennis dissented: he considered it a patent error to
hold that “the Lanham Act can be constitutionally applied to the noncommercial
political speech of a political organization, such as the political
endorsements made by Coalition in this case.” The result of this case law is
that political speech gets less First Amendment protection than entertainment,
which is fairly backwards. This is the
kind of question that, if the Supreme Court did take up a Rogers v. Grimaldi
case, might give it some difficulties.
ROP: A less eventful year, but some trends that may be of
interest.
NY added a postmortem right of publicity, requiring
registration with the state, for people domiciled in NY at death, lasting 40
years after death. It leaves in place the existing statutory right of publicity
for the living and provides rights when a “deceased personality’s name, voice,
signature, photograph, or likeness” is used “on or in products, merchandise, or
goods, or for purposes of advertising or selling, or soliciting purchases” of
such items, largely tracking California’s statutory ROP.
Also prohibits the use of a “deceased performer’s digital
replica in a scripted audiovisual work as a fictional character or for the live
performance of a musical work . . . if the use is likely to deceive the public
into thinking it was authorized by the person” or her heirs (as specified in
the bill). Notably, this provision
expressly states that a “conspicuous disclaimer in the credits” and in related
advertisements stating that the use is not authorized will insulate users from
liability.
Lots of model cases against strip clubs, with mixed results.
Apparently strip clubs often advertise by searching for images of sexy women
and putting them in their online ads, without regard to copyright or trademark.
States often have shorter statutes of limitation for ROP claims than for
trademark claims, which can leave the models relying on false endorsement
theories instead of ROP claims if they didn’t sue quickly, which is often the
case when the defendant is just one strip club that advertises on social media.
When the statute of limitations expires, the models can lose their false
endorsement claims on a variety of grounds: although courts in Arizona have
issued a series of plaintiff-favorable rulings, other courts have thought that
plaintiffs’ evidence isn’t good enough to show that their presence in ads is
likely to be perceived as an endorsement, as opposed to their images simply
being used as stock photos—after all, no one thinks that the average anonymous
actor in a conventional ad actually endorses the product they’re selling.
Interestingly, the FTC is quite explicit about this distinction for its own
endorsement guidelines: the guidelines do not apply to ordinary actors in
standard ads who would not be perceived as personally endorsing the products or
services in the ads. Only if there is some reason to think that the actor is
appearing in a personal capacity is there an endorsement according to the FTC,
which then means that the endorser must actually use the product, and the
endorsement must remain current so that if the endorser changes their mind the
ad has to be pulled. None of that has to happen with Flo the cheerful insurance
advocate or the Geico gecko. Anyway, with respect to the strip clubs, some
courts in Arizona have found confusion surveys persuasive, but other courts
especially in New York have excluded them because the plaintiffs’ survey expert
who does all these cases has a really bad, suggestive methodology.
Interestingly, other false endorsement Lanham Act claims
have been foundering on the Supreme Court’s statutory standing rules set out in
the Lexmark versus Static Controls case some years back. Courts increasingly
recognize that the Court was interpreting statutory standing for the entire
Lanham Act, not just for false advertising, meaning that a plaintiff who brings
confusion or false endorsement claims must come within the statute’s zone of
interest and plead a commercial, competitive injury proximately caused by the
defendant’s conduct. Some courts have held that models’ injury in not being
paid for their images is not a competitive injury for Lanham Act purposes, and
while it’s easy to assert that using a model’s image for a strip club ad hurts
her reputation, it turns out to be very hard to prove that at summary
judgment. This means that the ROP may be a person’s only avenue to recover for
unauthorized use of their images to promote products and services.
The recent OSU v. Redbubble case was mostly about trademark
claims, but it also made some potentially momentous statements about ROP as
well. Ohio’s ROP statute prohibits using a persona in connection with a product,
advertising a product, or soliciting the purchase of a product. “That broad
language expands liability beyond directly selling trademark-infringing goods.”
So even if Redbubble was passive, the ROP would apply if people who used
Redbubble to create goods infringed the ROP.
But TM remains broader than ROP in many ways: Walkowicz v.
American Girl Brands, LLC, 2021 WL 510729, No. 20-cv-374-jdp (W.D. Wis. Feb.
11, 2021): the case involved TM false endorsement claims and ROP claims by Lucianne
Walkowicz, who “has achieved a measure of celebrity as an astronomer,” based on
the claim that defendants misappropriated distinctive aspects of their personal
identity into a space-themed American Girl doll named Luciana Vega. Because
Wisconsin law covers a person’s “name, portrait, or picture” but not their
identity in any aspect, the court rejected the ROP claims. But false
endorsement is more flexible, so the court found it plausible that people would
believe that Walkowicz endorsed American Girl despite the differences in name.
No comments:
Post a Comment