Click to recommend this page:
The New York Times recently published an article titled,
As Rights Clash on YouTube, Some Music Vanishes.
It highlights Juliet Weybret, a 16-year old who recorded a video of herself
playing the piano and singing Winter Wonderland, and then posted it on
YouTube. Weeks later, the video was removed "as a result of a third-party
notification by the Warner Music Group, which owns the copyright to the
According to the Times article, this happens all the time, and
thousands of videos disappear for the same reason. Even common
family home videos that happen to include a portion of a song playing
in the background are removed with prejudice, along with
videos that use music in goofy ways, from montages to mash-ups.
What I find particularly interesting are these facts:
- No human is involved in the infringement decision. It's an automated
system that merely identifies songs; it does not distinguish whether
the given use of the song actually constitutes an infringement.
- No human is involved in policing or overseeing this process,
questioning it, objecting to it, challenging it, or even calling
anyone's attention to it. (Well, except for the New York Times.)
- No entity at allhuman or corporatehas any financial
incentive to defend these claims: to intervene, question, object,
It's certainly the case that Juliet's video is not an infringement (it's
an unambiguous example of the Fair Use provisions of copyright law).
But that doesn't imply that all cases are not infringements. Indeed,
it certainly may be that some uses actually infringe. The important
point is that no one is actually making that assessment and no one has
a financial incentive to challenge them.
The consequences of all of the above have a cascading effect that,
over time, has eroded the spirit and intent of copyright. One
of the first things that happens is that the common perception of
what constitutes a copyright infringement is being obfuscated. This
is the exacerbated by additional factors. The complexity of copyright
law is such that not a lot of people are aware of it or can interpret
it easily without extensive experience. Even copyright lawyers
often disagree over what kinds of uses are infringements and what are
permitted under Fair Use. In fact, it is this very reason that "claims"
of infringement need to be balanced by informed counter parties. The fact
that there is an automated mechanism that forces content to be removed
without "challenge" runs counter to the spirit of copyright.
Also, the reason why there is no challenge to these claims is that the
size and financial wherewithal of the litigants have a disproportionate
influence on their business partners. As the Times article sums it up,
"Users have the right to dispute a take-down. But few have. People are
somewhat intimidated by the possibility of being sued by one of the
music companies, even if they have a free lawyer."
No matter how you look at it, it's pretty clear that the copyright
system isn't working as it was intended for cases like this. But despite
my characterization above, this isn't entirely due to large companies
using their financial muscle to exploit the little guy, or the massive
amount of copyright infringements that users do. The source is much more
basic than that: the internet itself, where massive amounts of content is
being created and used in quantities and ways that the original copyright
infrastructure was never designed to handle.
Note: I said infrastructure; I did not say "laws." Copyright law
is perfectly capable of addressing these cases, of discerning which
of these uses actually are infringements, and which aren't, despite
the hyperbolic claims of copyright holders. Hypothetically, if
both sides of a dispute were properly represented by legal counsel,
the court system would make use of existing law and find appropriate
interpretations to maintain legal continuity. Granted, not every case
is actually judged "fairly", as anyone who's ever lost a case knows. I'm
not being naive. However, it's the aggregate net effect of rulings that
has maintained consistency and stability in the general perception of
the integrity of the courts.
But the reality is that both sides don't have balanced representation.
Some have said that free market conditions will correct for such
imbalances. The argument goes that the economic quandary for Google
is that, on one hand, user-generated videos are important to YouTube's
efforts to increase the flow of advertising dollars. On the other hand,
their licensing deal with Warner (to display the professionally-made
videos) generates significant traffic as well, which also contributes to
advertising revenue. Is it the case that Google's competing objectives
here would give them incentive to keep Warner Music from going too far?
For Google to take a position on the matter would imply a delicate
balance of economic interests. Is this a true balance of power?
The real answer is no, because unlike Warner Music, whose sole financial
incentive is to represent their own interests, even (or especially)
at the expense of Juliet's interests, Juliet has no one representing
solely her interests. Google is not a fair arbiter here; it is
only trying to determine what's in its own financial interests.
Its decision is not based on what's "right", but who has the larger
It would seem like Google would realize that Warner Music isn't more
valuable than the massive amount of user-generated videos like Juliet's.
But the economic reality is that 1) end-users won't go away, 2) they won't
stop producing and uploading content, and 3) they have no legal recourse.
Hence, they have no economic influence, so Google and other social
networks don't need to listen to them. By contrast, Warner Music is
a single entity that represents a large volume of works, it does have
legal recourse, and they could pull all their content. So Google
listens to them.
And therein lies the real problem. That's why it's an infrastructure
problem, and not about the law. What I mean by "infrastructure" is the
set of tools and conditions set up by the government to provide means
and mechanisms that give strong incentives for companies to be "good
citizens" in the copyright realm. When there are no incentives for
anyone to assume the counter-balancing position, the infrastructure is
out of balance.
One could say that this provides a great business opportunity: to create
a company like Warner Music, but who represents all the individuals who
upload their content to social networks. Imagine a hypothetical site
called "your-content.com" with a marketing statement that said, "if you
upload your content in exchange for letting us represent you legally,
we will pay you commissions on anything that is licensed from us,
or any damages collected if someone steals your work. You can place your
content on any site anywhere on the internet, and if someone licenses
it, we'll pay you a commission." While that's well and good, the problem
is that none of this is actually possible unless each work is actually
registered with the copyright office. And therein lies another major
problem with the copyright infrastructure.
Because of how copyright law is set upand this is for very good
reason beyond the scope of this articleonly those works registered
with the copyright office enjoy an extra level of legal protection,
which itself leads to financial incentive for others to comply with
their copyright. Warner Music has registered its massive media
library with the copyright office, which is what gives them the legal
and financial strength to assert their will upon Google or anyone else
that wishes to distribute their works. Juliet can't assert her rights
in any practical manner because she never registered her video with the
The hypothetical your-content.com could never exist today because
end-users like Juliet don't register their works. And, at a minimum
registration fee of $35 per application, it isn't likely that
your-content.com will pay to register all those works either. As a
consequence, your-content.com would never able to practically enforce
its legal protections, which would prevent them from deriving revenue
from potential licensees. User-generated content isn't that valuable
on a per-unit basis. It's only valuable in massive quantities, since
only a small proportion of it actually sells, and it's impossible to
predict which units will perform.
There's another hurdle to the problem: even if the registration
fees were zero, the copyright office has no infrastructure at all to
handle all those applications. Even today, with only tiny fraction of
even professional artists registering their works, applications for
copyright still takes 5-7 months to process.
Once again, this illustrates that the current copyright infrastructure did
not anticipate the volume or business models that the internet has created.
Price, processing capacity, and financial incentives built into the system
to encourage the private sector to give representation to individuals. These
are what's missing.
To address these failings, I submitted a plan (in 2007) to revise the
copyright infrastructure. In my article,
Proposal for Privatizing the Copyright Registration Process,
I proposed that the copyright office offload only its simple administrative
task of processing copyright registrations to the private sector.
These new companies would be called "Copyright Registrars," of which
there could be an infinite number. Their role is two-fold: First, to
process copyright registration applications using a more effective and
efficient automated system than what the copyright office does today.
Second, they would "legally represent" works that were processed within
a defined set of parameters (with oversight) to assure integrity. This
is the financial incentive as I just described. By delegating the
administrative tasks to a large number of private sector companies who
can automate the process en masse, the cost of registration quickly
approaches zero, and the capacity to process infinite number of
works rises. These remove two major barriers that allow companies
to build and justify business models that ultimately lead to equal
representation of people's interests.
Note, the proposal is not a transfer of copyright ownership, nor is it
consent or permission for the registrar to use the work. It is simply
a transfer of administrative tasks. Once a work is registered, there
are three events that would involve the Copyright Registrar:
- If a third party wanted to license the work, the copyright registrar
would then negotiate and collect the license fee. (This would largely
be an automated process for most common uses.) Because the work could
exist in any number of places around the internet, it doesn't matter where
the work is found, it only matters that someone wants to use it in
a manner that requires licensing. Here, the Copyright Registrar that
processed that particular work would have the right to price it and
collect fees. These fees would be shared with the copyright holder.
- If someone has infringed upon the work, the Registrar would
pursue those infringers, possibly collecting statutory damage awards
or license fees, either of which would be shared with the copyright holder.
- If a third party claimed the work infringed on another work, then
the Registrar would assess the claim and either comply with the demand
or challenge it.
This now lays the very incentive models for businesses like the
hypothetical your-content.com to exist. So now let's consider a more
realistic example: if Google were a Copyright Registrar. They would
have YouTube be a site that automated the processing of the end-user's
copyright registration application the moment the person uploaded one of
their own videos of original content (assuming the work hasn't already
Now, if Warner Music were to claim that Juliet Weybret infringed on its
copyright, then Google would challenge it. Why? Because if they didn't,
people would stop using YouTube, and move to another video site (that
was also a Copyright Registrar) who did a better job of helping to
monetize and protect users and their content.
And let's face it, if such conditions were to exist, then Warner Music
wouldn't even bring the case in the first place, nor would Google permit
the automated robot that sent out the take-down notice. So, already there
would be a stabilizing effect by the very nature of there being a balanced
weighting of legal representation. There would be less Copyright obfuscation,
less exploitation, and a better, more efficient economic model for growth.
The other thing to consider is that Warner Music itself could also be
a Copyright Registrar, and Juliet could very well have registered her
little video with them, and still put it on YouTube. In this case,
the music company would have more incentive to keep it there, since
they would be receiving a portion of its ad revenue. With that kind
of financial incentive spread across massive amounts of user-generated
content, this translates to real money. Now there would be no incentive
in falsely demanding that non-infringing works be removed from YouTube
or anywhere else.
Predictably, everything stated above needs to stand to scrutiny. So,
let's get into that.
Is there really economic value in Juliet Weybret's video? Perhaps not
hers, but one needs to consider the aggregate value of massive amounts
of user-generated content, especially when its collective weight adds
up to substantial negotiating power. This is all part of the latest
modern-day financial model called "long-tail economics." This is where
money is made not by selling "the best" content available, but to amass as
much of it as possibleirrespective of qualityand letting the natural
sorting and searching mechanisms of the internet allow things to be
"found" and subsequently sold. Given the massive amount of user-generated
content on sites like YouTube, Flickr, MySpace, Facebook, and thousands
of others, the aggregate value of licensible content is enormous.
This has been the focus of my research for close to ten years,
which I have written about on my blog
The most recent articles that pertains to this can be found
Content is king, and everyone from social-networks to search engines
knows it. The holy grail has been figuring out a business model that can
capitalize on it. Facebook itself tried to change its end-user agreement
that gives them rights to use content that people uploaded to their site,
but they had to back-off because their "deal" wasn't fairpeople weren't
going to be compensated for their contributions.
The whole idea of the Copyright Registrar is for this very objective
to be more open, transparent, and competitive. As more Copyright
Registrars come online in their various forms, critical masses of people
gravitate and congregate around those sites that are known to perform
well in licensing and protection. There's no risk of monopolies or
price/performance shenanigans because the competition would be diverse
enough to keep anyone from cornering the market.
As higher quality content (along with higher quantities of it)
add more and more to the global media library, all of which now
have equal opportunities for licensing, it would be extremely rare
for individuals to represent themselves anymore; almost all content
online would be represented by someone with far more power and
capability than individuals have on their own. Furthermore, licensing
terms and disputes regarding infringement would be exchanged between
more equally-sized parties, who have more balanced legal weight and
economic power. Gone would be the days of individuals stressing over how
much to price a photo that someone wanted to license, or how to deal
with a company that stole their photo from Flickr for use in a brochure,
or having to pout because Warner Music forced her song off of YouTube.
As for licensing fees, again automation comes into play. There's already
models built in other industries for auction-based license-pricing
systems, very similar to those used by Google to set advertising
rates. Applying them to the lion's share of user-generated content would
not only be simple, but lucrative. Distribution of commissions to the
creators would be part of this business segment.
The last part, which is really where the potential cost and complexity
comes into play, is the legal entanglements: assertion and protection of
rights. Determining whether the use of a given work is an infringement
is the elephant in the middle of the room. Copyright law and its
various provisions for Fair Use and issues concerning the First
Amendment make for a never-ending set of challenges. But I don't want
to overstate the concern eitherit's certainly manageable once you're
dealing with informed parties of equal weight and economic interests.
Deals are struck, middle ground is found, and balance is achieved.
This, combined with the more ubiquitous and nationwide use of
the copyright infrastructure, would actually reduce the number of
infringements, since it would be too easy to be caught, and you'd be
caught by very large and powerful interests. This would increase
the rate of legitimately licensed material, which has a multiplier
effect on economic activity. Best of all, the people who create the
content would be recipients in the system for a change. This itself
has a stimulative effect on participation: people have more incentive to
participate and to produce better, more competitive content when there
is a merit-based compensation model.
The lesson that history teaches is that "economic incentives" is the great
This article is by no means a detailed description of my copyright
infrastructure proposal. For that see the link to the original proposal
I cited above, which can be found
In it, I also address questions and criticisms that have come up, as well
as necessary and detailed policy issues that would have to be considered.
Click to recommend this page: