Moderator: Brian Love, Santa Clara Law

Jeanne Fromer (trademark), New York University Law School

Search and examination on relative grounds (Europe doesn’t
do that)—has critiques but generally doing a decent job. Ironic b/c we think of
US as “free market” and Europe as paternalistic but registration works the
opposite way. Smaller businesses may not have registrations but can get
benefits from opting in; use is still the core of US TM. This is a way to give
them some protection/different pathways for TM rights, and having the two paths
(registered/unregistered) is generally good. Courts also helped systematize
distinctiveness—again with some things that aren’t working great, but the
systematization is a good thing. Semantic connection b/t mark and category of
goods/services must be evaluated: shouldn’t be protected as a mark when
connection is too strong, or protection shouldn’t be easily granted if it’s
pretty strong. Focus on consumer perception is also a good one from the
perspective of TM’s goals: to keep consumers from confusion in the marketplace
and being responsive to how they behave.

Rob Merges (patent), UC Berkeley Law

Volume/velocity of transactions based on IP rights has grown
amazingly since 70s/80s, a little bit invisibly in how many business models and
transactions enabled. Textualism/literal infringement doctrine gave rise to the
practice of claim charts, which made for transactional efficiency: a
formalistic procedure. Structures claim interpretation process into fairly
narrow channel. This allows a boom in patent licensing. Allows fast development
of vaccines—patent licensing is behind the scenes. ROP is also good b/c there
are things you can do with property rights that you can’t do with contract
alone in terms of transactions.

Rebecca Tushnet (copyright), Harvard Law School

Unlike my predecessors, I’m not going to do an overview, but
talk about a specific provision of copyright law. Just as democracy is the
worst form of government except for all the others that have been tried, so too
with section 512, the safe harbor provisions for internet service providers,
which has proven remarkably robust despite multiple efforts to destroy it over
the past nearly three decades. I have my own litany of complaints about
practical problems with 512, but in terms of dispute management it is a
resounding success. I will compare 512 to the recently enacted Take It Down Act
and discuss the evolution of caselaw by comparison to Carol Rose’s account of
property titling systems.

512 creates a safe harbor against monetary liability and
sweeping injunctive relief that would require changes in practices for online
service providers that follow certain rules about how to handle complaints of
infringement. It divides service providers into four categories, one of which
is essentially defunct; service providers that provide connections for content
like email or private messaging are supposed to have policies that terminate
repeat infringers, and that’s definitely created some problems now, but for
decades the key provisions were those for service providers who store
content—like YouTube—or provide links or search engines, like Google—who won’t
be held liable for direct or secondary infringement if they promptly take down
instances of infringement when properly notified about them. 512 explicitly
does not require services to monitor their services for infringing content. It
does provide for liability without notice if a host or linker has “red flag”
knowledge of infringement, but the courts have generally been pretty robust
about making sure that general knowledge that infringement is taking place on a
platform, or even general knowledge that there are multiple copies of an
infringing work on a platform, don’t count as red flag knowledge. So unless a
site is something like “top 50 movies to download,” it probably won’t have red
flag knowledge.

What went right with 512? Well, one way to measure it is how
many disputes it has resolved.  Caselaw
v. number of disputes—the number of disputes is in the billions (even if 1/3 of
the notices sent to Google are invalid, which seems to be the case, still
billions of correctly targeted notices). There are big 512 cases, but not that
many of them. The caselaw tells you only what was significant enough to fight
in court about for unusual reasons—either reasons of the defendant’s deep
pockets and structural significance in the entertainment ecosystem, or reasons
of the plaintiff’s specific interests, usually a moral sense of offense. The
everyday functioning of the system, though, is that lots of infringing stuff
gets quickly taken down, often—if you believe Google—even before anyone has
even seen it. And uploaders who disagree with the takedown can file
counternotices; a service that honors the counternotice is immune from
liability for reinstating the content and the copyright owner has to sue the
uploader. Very few counternotices are filed.

Another way to measure success: 512 immediately became the
default rule around the world, at least in practice—even in Europe, 512
compliance was for a long time sufficient to avoid being sued
successfully—suggests 512’s utility as a workable compromise between interests
of IP owners and accused infringers (compare the fate of the similar section
230, which definitely spurred US dominance in tech but was not routinely
accepted as the final word on intermediary liability in non-IP situations).

Europe eventually diverged, at least formally, by requiring
intermediaries to engage in licensing attempts and screening. But I say
formally because even today it doesn’t seem to me that this has worked except
for music and popular video; major forms of copyrightable works like
photographs and texts are just not amenable to that kind of licensing
requirement because ownership is not concentrated enough for comprehensive
licensing regimes to form. European regulators have the benefit of not actually
needing to require exact compliance with what looks like the plain meaning of
the law; being a “good guy” is generally enough in the EU, something that is
often surprising to US lawyers, who expect a law that doesn’t explicitly have a
good-faith standard to not have a good-faith standard for compliance. The DMCA,
that is, is still shaping behavior on the ground around the world.

There was a major attempt in the last 10 years to gut DMCA
and put in concepts like notice-and-staydown in the US, which would require
services to search for and remove similar copies after receiving notice about
the location of one infringing copy. This was a brilliant branding move by 512’s
haters—staydown sounds almost like takedown, so how hard could it be? But it is
actually a huge technical challenge, especially for smaller services, and would
have been a huge gift to YouTube in maintaining its dominance. Fortunately,
this attempt fizzled, which is one reason that sites like Wikipedia, Reddit,
Ravelry, and the Archive of Our Own can operate in relative confidence without
the resources of a Google or Microsoft.

What about my complaints? Well, there’s definitely
complexity, as the Supreme Court’s recent attempt to tackle intermediary
liability for internet service connection providers makes clear; we’ll see what
happens in the Cox case and that will tell us a lot about what 512 means for
connection providers in the US. And counternotice law is a mess: when abusive
notices are sent, it’s very hard to hold the sender accountable. But I want to
make the case that the crud that’s accrued around various aspects of 512 is the
standard fate of almost any law allocating rights, no matter how clear.

In her justly famous 1988 article Crystals and Mud
in Property Law, Carol Rose explained that all clear legal rules,
particularly rules created by legislatures, face pressure from two sources: the
ignorant and the conniving, the fools and the scoundrels. The ignorant don’t
know about the law, no matter how clear and crystalline it is, and they are
subject to mistakes that make them lose out despite them not having done
anything morally wrong. They didn’t know that land can only be transferred by a
writing, so they rely on an oral agreement and hand over their money and are
out of luck. The scoundrels exploit the clarity and hard edges of the legal
rule to get unfair outcomes: the law says a mortgage has to be fully paid by a
date certain or the borrower defaults and loses the land, so a conniving lender
can hide out and make it impossible to find them until the time has passed and
they get the land and most of the money. These abuses and unfairnesses pile up
and courts for completely understandable reasons will make up special
exceptions to allow equity back in, muddying the clarity of the written rule.
Equity of course has its own costs: it makes disputes more unpredictable,
therefore expensive, disadvantaging poorer actors, and equity is also more
reliant on the biases of the factfinder who may have its own predispositions
about the characteristics of good guys.

I believe that much of the uncertainty that has accreted in
the corners of 512 law is the inevitable effect of this crystals and mud
dynamic; it is not unique to 512 and therefore it’s extremely unlikely that
changes to 512 could do anything more than restart this process: replacement
rules would either be mud all the way down, which I think is worse, or be a
different and probably worse crystal that would not make either creators or
intermediaries better off. Rose concludes that, when the mud gets too bad, the
legislature often intervenes to put a new crystalline rule in place—the fact
that 512 has survived some reasonably well-resourced legislative assaults to me
suggests that this isn’t one of those situations where the mud has fully gummed
up the works.

I want to end by comparing 512 to the recently passed Take
It Down Act, aimed at sexually explicit images “indistinguishable from an
authentic visual depiction” published without the consent of the person shown
in the picture. TIDA is not a safe harbor regime. Instead, it requires two
things that aren’t in 512 to avoid liability: (1) services must remove accused
images within 48 hours of receiving notice (with no clarity on what qualifies
as receipt: I give it about 6 months before someone uses a mailing address and
sues based on the time the mailed notice arrived at a building); (2) services
must make reasonable efforts to remove known identical copies.

There are no counternotice provisions, even if the content
was fully protected by the First Amendment; no safe harbor against liability;
no guidance on what counts as reasonable efforts or what qualifies as
knowledge. 512 does better on all these counts. Harbinger of attempts to do
even more in proposals like Take It Down which cover any use of a digitally
altered likeness, even as the White House is posting AI-altered images on
official accounts.

Graeme Dinwoodie (International IP), Chicago-Kent College of
Law

US jurisdiction over foreign © claims was done right. US
wanted to become a leader in int’l © law to enable more effective enforcement.
Led to NAFTA, TRIPS. Public law side interventions not always received
enthusiastically outside the US—intrusion on sovereign choices. But private
litigants used them successfully to argue that US law shouldn’t be applied
extraterritorially. Predicate act: foreign profits from US infringement can be
secured in US courts, often w/o regard to law of other country, but there are
limits to that; would have liked more comity-facing analysis.

Historical reluctance globally to adjudicate claims of foreign
© infringement. SDNY, a few years after 1978 Act, allowed claims for
infringement in various South American countries to proceed. Impulse to provide
relief under multiple foreign © laws were spot on. NY was where D was amenable
to jurisdiction; hope that British courts would do the same thing in similar
circumstances. For some time, this case was alone, in part b/c of fear of
foreign “bramble bush” of law. Eventually, this approach got appellate
endorsement in Boosey & Hawkes v. Disney—simply having to apply foreign © law
of 18 countries was not a reason to decline jurisdiction. Not a torrent of
cases, but the availability of such relief has structured private behavior. Has
also been embraced by courts in Europe.

Especially valuable for small authors w/no resources to
litigate cross-continent. Streamlining duplicative litigation is not inherently
pro-© owner: Computer Associates v. Altai didn’t just occur in NY, but also in
France. After they won in the US, Altai unsuccessfully sought an anti-suit
injunction from relitigating in France, but it would have been more efficient
for a small defendant to secure global clearance in a single case. (Altai won
in France, but would have been much faster/cheaper in one court.) Better than
seeking enforcement of US law to entire dispute. Enhances legitimacy/embraces
sovereignty. But it does create complexity in applying foreign law; the good
news is that public law has reduced divergence; there weren’t 18 different
rules in the Disney case.

Different constraints in TM/patent. In patent, distaste for
litigating foreign patents in Fed Cir and ECJ. ALI principles endorse doing
patent/TM, with patent invalidation being given only inter partes effect. And
in last few years, courts signaled more willingness, though Albright in Texas
wants to grant anti-suit injunctions. That’s where this is going, along with
jurisdictional issues. Will be a dialogue b/t patent judges in different
countries; likely to be more respectful of sovereignty and thoughtful than the
public law debates we’re going to see.

from Blogger http://tushnet.blogspot.com/2026/01/santa-clara-school-of-law-intellectual.html