The Copyright Act at 50: Evolution and Impact
Shira Perlmutter
Copyright Act took a long time, with input from lots of
interest groups and attention to detail—hundreds of contending and overlapping
interests were involved. Hard to imagine this process today. Desire to avoid
need for constant amendment/future-proofing. But did they do enough? Didn’t
create a general right to exploit the work publicly, which would have obviated
the need for continued parsing the scope of each right, like public
performance.
Some changes over time, most prominently the DMCA. More than
codifying common law principles; tech-specific obligations. Less durable as
business models evolved; might have unexpected consequences in Cox.
[Hunh? Cox is not a DMCA case.] 1201: some provisions are highly detailed and
technical, and outmoded. But the rulemaking process is flexible and fair-use
based and has produced new exceptions. Allows © owners to rely more securely on
TPMs, enabling the celestial jukebox. Fair use has also played a critical role
as a flexible judicial tool. The bones are solid, even with AI.
Q: registration requirement is tough on creators. Can’t get
protection [statutory damages] before infringement.
A: You can register—the issue is remedies.
Chris Sprigman: why do you think that it was ill-considered
to add to fair use that unpublished status isn’t dispositive?
A: b/c courts had already walked back their overreading of
unpublished status. Worried about accretion of more specific language in a
statute that’s supposed to deal with rapid change. [Seems like a levels of
generality issue; unpublished seems general enough to be robust.]
Sprigman: it’s always good when Congress talks back to the
Supreme Court.
Laura Heymann: say more about moral rights?
A: A patchwork in the US; would love to see Dastar
reversed and some additional protections provided.
Q: how to design AI training licensing framework?
A: Doesn’t have proposal but thinks it would be possible;
easier in areas w/high value works. Small low value works w/authors who aren’t
organized are harder. At some point there may be a statutory solution building
on experience in the private sector with making a licensing system work.
Tyler Ochoa: Cox v. Sony?
A: Personal views! Shocked at how short the decision was and
how little thought there seemed to be about the implications. Threw out decades
of © law quickly w/o analysis. Repercussions well beyond the facts. Congress
clearly intended to continue the separate treatment of © contributory liability
from patent and aiding and abetting liability. [Don’t you know that only
Supreme Court cases count? My line on this: “This is easy and you are all
stupid” is a poor way to think in drafting most Supreme Court decisions.]
Do We Need a New One? William Fisher
Statute has grown by accretion, not revision, and only when
there can be agreement by major stakeholders. Hypothesis: useful to start
fresh. Draft from Oren Bracha, William Fisher, Ruth Okediji, and Talha Syed. A
couple of points: Limit scope of adaptation right. Reproduction
right/substantial similarity is almost overlapping with it, but matters when
there’s no reproduction. Independent of exclusive rights of © owner, wants to have
rights attached to, at least initially, authors rather than owners—right to
attribution, generously defined, and to integrity, narrowly defined. Shorter
duration. Compulsory licenses not just for music covers but for educational
uses.
Sprigman: Why is remuneration for authors the first
principle? The Court has said that’s a means to an end. Why not “vibrant
creative environment”?
A: order isn’t meant to connote hierarchy, but worth
thinking about. Utilitarianism isn’t the only goal; fair treatment of artists
is also a goal.
Q: like the use of lessons from laws around the world. Was
that a reason to delete statutory damages, which aren’t available in many places
around the world?
A: there are well-known specific defects in the US system of
statutory damages. The substantial range for willful infringement per work
becomes bizarre & punitive. There are workable models that would function more
like liquidated damages in contracts. The functions of augmented damages,
including incentives to bring suit, could be adequately performed by enhanced
damages for abusive positions (doubling) and attorneys’ fees. Fees should be more
likely for small creators and less likely for deep pocketed plaintiffs.
About Face: Deepfakes and the Misuse of Copyright Madhavi
Sunder
Denmark is granting © in a person’s face to combat
deepfakes. Incentives/progress/access aren’t just buzzwords but the raison d’etre
of our law. But roughly a decade ago, things began changing, not just b/c of
AI: using © as a tool to redress noneconomic social harm: safety, protection,
dignity, reputation.
Denmark goes beyond using © to serve non © ends/do an end
run around 230, as past proposals in the US have done (thanks
for the shout-out) to expanding the scope of © beyond what it should cover.
Denmark’s amendment covers life of the author+50 years and
protects all natural persons against digitally generated images of personal
characteristics. Limitations for caricature, satire, parody, criticism of
power, social criticism, etc. But this would cover foreign nationals as well. Includes
a takedown right. Drafters suggest that the new right is not really copyright
but personality right, and the law should be changed to be officially called
the “Copyright, etc. Act.” [It’s ©, Jim, but not as we know it!] Attribution
and integrity for authors is not the goal; broad dignity harms to individuals,
society, and democracy.
EU is considering whether to adopt a similar proposal. US
may be heading in a similar direction—Jennifer Rothman identifies convergence
between ROP and ©. Digital replica report by Copyright Office suggests new laws
are needed.
Faces and voices aren’t authored in the way © has
traditionally required; we allow soundalikes. However, some people (Balganesh,
Gilden) suggest that © has always had concerns with dignity. Likewise, the
Court allowed photographers to own © in depictions of faces. This tension
raises charges of unfairness, as in Moore v. University of California. Descendants
of enslaved people can’t claim ownership of daguerrotypes of their ancestors;
Prince, who decried ownership of his name and music, becomes the subject of a
photographer’s © claim at the Supreme Court. Surveillance: your face belongs to
us. The issue about face is not whether property, but whose property.
IP and blackface: Jim Crow was a minstrel character—“love
and theft” of black dances and bodies—loved and despised, coveted and expropriated.
Elvis painstakingly listened to recordings of Black artists on repeat so he could
copy them, and Tennessee then called its voice ROP law the ELVIS Act—irony! Digital
replicas are the next frontier. Abba has created a concert featuring digital
replicas of their younger selves; they sang and danced in motion capture suits
with monitors and cameras everywhere. This show will last as long as people
will pay to see it.
Sunder’s about-face: She criticized the goal of efficiency
in © and argued for considering other interests like semiotic democracy. Is
this the same thing? No. © can’t be everything everywhere all at once.
Doctrinal coherence matters. Doctrinal collapse b/t © and privacy has
structural harms including threats to the rule of law. © is too consequential
and long-lasting and easy-to-get to be careless about; statutory damages and
notice and takedown are big deals.
© is about authors, whether you’re a high protectionist or
low protectionist. In an age where we’re all curated online, we should have a low
threshold for protection, but not create mutant copyrights far from the real thing.
Cathy Gellis: Implications for national treatment?
A: will think about it—interested in whether we’re
replicating it for ourselves.
Lemley: is the right alienable in Denmark? © as a regime is
usually about being able to sell rights.
A: all premised on consent.
Quasi-copyright and the Copyright Act, Rebecca Tushnet
My focus here is on 1201 and 1202. My argument is that their
evolution in the courts shows something about the workings of the legal system
and the incentives of both plaintiffs and judges.
As most of you know, 1201 prohibited circumvention of access
controls and trafficking in technology that circumvented rights controls or
access controls, with a variety of statutory exceptions that are essentially
too complicated to be used, and a provision for allowing additional temporary
exceptions after Copyright Office rulemaking, but only for the direct access control
circumvention provisions not for the trafficking provisions, so you have to
both have an exemption and somehow get the technical capacity to use the exemption
which is illegal for someone to give you.
Tony Reese wrote a great article explaining the benefit to
the copyright owner of characterizing a technological protection as an access
control rather than a rights control – no individual circumvention is allowed
in the absence of an exception– thus in every case, copyright owners plead that
a TPM is an access control, and courts have uniformly accepted this
characterization—so this supposed four part scheme of access and rights
controls, direct circumvention and trafficking, quickly became a two part
scheme involving only access controls. Rights controls immediately lapsed into
desuetude.
Because of how broad 1201’s access control provision was it
initially seemed to offer copyright owners broad new rights. This was
especially important for manufacturers of machines that happened to have
software in them—providing compatible products, for example, could be reframed
as violating access controls. However, in two prominent decisions courts—using interpretive
methodologies that would probably not be adopted today—interpreted 1201 to try
to prevent its use to control markets that aren’t really based on the value of
the copyrighted works; the major cases are perhaps tenuously based in the
statutory language but they probably do track what Congress thought it was
doing.
These two decisions, Chamberlain and Lexmark, dampened the
appetite among many non-copyright-reliant manufacturers to use 1201 to try to
control repair and resale. There’s a real case to be made that 1201 has importance
for phones and apps, but it’s no longer a big part of copyright litigation.
In addition, the rulemaking process proved so exhausting
that the Copyright Office decided to streamline it for existing exemptions. And
because the trafficking provisions only cover traffickers, not customers of
traffickers, people with exemptions use circumvention software they got from
elsewhere and we all just generally ignore the issue in the exemption process. I
would suggest that, at least for the time being, we’re no longer in a legal
innovation phase with 1201.
Meanwhile, 1202 litigation has exploded. 1202 covers knowing
removal of copyright management information that facilitates infringement or
provision of false CMI, and although there were always a few cases about it, it
has been discovered in the last decade—as causes of action sometimes are
because lawyers are innovative—and gained new prominence in cases like the AI training
cases. 1202 doesn’t require registration in order to get statutory damages and
so questions about what constitutes removal of CMI or the relevant intent are
actively being litigated. Pam Samuelson and her coauthors have written a good
article about the arguments, but I just want to point out that lawyers have
done exactly what they’re good at: pushing the boundaries of the law in order
to achieve interests for their clients even when the more obvious claim—like copyright
infringement—won’t work for copyright-specific reasons.
Given this increased use, it’s not surprising that we see countervailing
theories attempting to limit the growth of 1202 cases. One court even recently
dismissed a lawsuit brought under 1202(b)(1) against ChatGPT on Article III
standing grounds—under the TransUnion case, 1202 can’t constitutionally
authorize a private cause of action for internal CMI removal that goes no
further—plaintiffs didn’t allege any actual harm beyond the removal of CMI in
the training dataset, so they didn’t have standing to seek damages, and they
didn’t plausibly allege that a substantial amount of their creative expression
would appear in future results, so they didn’t have standing to seek injunctive
relief.
I have some broader thoughts about this incredibly abbreviated
account, based on Carol Rose’s classic article, Crystals and Mud in Property
Law: Fools and scoundrels are the bane of the law because they make it
unpalatable to follow the most natural understanding of a clear rule. Hard
edged rules written into law—like the prohibition on circumventing access
controls—predictably lead scoundrels to abuse their fellow citizens, as in Lexmark
and Chamberlain, and subjects fools to disproportionate liability, especially where
statutory damages are involved. Courts then understandably push back, inventing
equitable limits and turning a clear rule into something more muddy. But muddy
rules are expensive to navigate and create their own set of problems.
In Carol Rose’s story about real property law, legislatures eventually
intervene to create a new and different clear rule designed to solve the problems
created by existing fools and scoundrels under the previous regime. This works
for a while and then the infinite creativity of humans, both good and bad, produces
new fools and new scoundrels.
I think Rose’s story has key lessons for copyright. (1) Future
proofing is something of a myth. It’s worth trying, because immediate obsolescence
when a few facts about the market change is not good—I’m looking at you, vessel
boat hull and mask works protection and 512(b)—but the idea that you can set
and forget a law ignores the fact that lawyers and judges are human beings—at least
for now—and human beings are collectively really good at finding ambiguity or opportunities
for arbitrage.
(2) If we face a situation where we don’t trust that the
legislature will intervene, or can intervene productively, then things get a
lot harder. When that’s combined with a judicial approach to statutes that
focuses on the dictionary meaning of specific words rather than an appreciation
for the structure of the legislation and the context in which the legislature
was operating, scoundrels are likely to prosper and fools are likely to be
abandoned to their fates. I don’t have solutions but I am predicting a long roll
in the mud.
Lots of interesting comments; I think both the legislative
process (actual deliberation) and judicial concepts of the role (neither
entirely free to disregard the statute in favor of the common law/equity nor laser
focused on individual words in isolation from the structure and purpose of the
law as a whole) need change from where they are.
Posts
