Click to recommend this page:
In 2007, I submitted a proposal for how the Copyright Office
could positively benefit the public interest by allowing for the creation
of "authorized Copyright Registrars." These new entities would assume
the administrative tasks of copyright registration processing, the effect
being a reduction in processing fees and a massive increase in processing
capacity. The copyright office itself should focus on policy and
oversight. The economic incentives given to the private sector would
create a positive economic stimulus by setting the stage for more business
interests to engage in and encourage more people to register their works,
and for more users of those works to be more compliant with copyright law.
Since I wrote that proposal, I was asked to provide deeper analysis
and to substantiate it with background data. This article reintroduces
the proposal with that information.
- The amount of content online produced by consumers is enormous;
it's also growing by orders of magnitude, the result of social
networks and content-sharing sites of Web 2.0. As Web 3.0 emerges,
where search algorithms that semantically analyze content for
its true nature, a greater proportion of old and new content alike
is showing up in mainstream search results.
- Because businesses use the internet to find content for both
commercial and editorial uses, more user-generated content is being
used in ways that should require licensing, but isn't.
- Both creators and users of works are largely unaware of whether and
how to comply with copyrights, partly because copyright law
itself is not well understood, but mostly because people's behaviors
are established by "common usage" on the internet. This applies to
how they use others content, but also of how they view their own
creative works. Consumers are entirely unaware of "copyright
registration," and corporate policies do not assure compliance when
user-generated content is found and used. This lack of awareness on
both sides has been the basis for an incorrect perception that
user-generated online content is free, unmonitored, and "valueless."
- The lack of truly effective economic penalties for "stealing from
the consumer" further perpetuates the corporate culture of "passive
infringement," resulting in both wasted and lost economic opportunities.
- The reason there are no counter-balancing interests (companies
that would help people monetize their content and enforce effective
protection), is because it's impossible to build a viable business
model. For reasons discussed in this article, an entire industry (and
a shift in how online content is monetized and protected) hinges on
one thing: the cost and process associated with copyright registration,
which, if solved, opens the door for investment into business models
and technologies that solve these inefficiencies.
- Of the many proposals that have been proffered to alleviate this problem,
only one achieves the dual objective of not altering existing copyright
law, policies and structures, while also creating stimulative conditions
for economic growth. And that is for the copyright office to delegate the
copyright registration process to the private sector.
- This article proposes a "Copyright Infrastructure" modeled after the
same model as the ICANN infrastructure that manages domain names and
numbers. A new entity, a "Copyright Registrar," would be created to
assume the administrative responsibilities that the copyright office
currently does. These new entities would scale up the processing of
billions of copyright applications, thereby reducing the costs (and
fees) to negligible amounts, and making the database of all registered
works available online.
- The online database would be accessible by applications through
open, standardized protocols that allow third party applications and
services to access to build a new economy that facilitates transactions
that benefit both buyers and sellers.
- Registrars would then be able to act in the interests of the copyright
holders, by helping sell/license content to interested parties, and to
protect works from infringements. The creators of works would themselves
have financial incentives to participate in the global economic system
of content creation and licensing.
- The result would be a huge economic boom of licensing of micro-valued
Let's start with this simple quiz.
Which of the following is a form of copyright infringement?
- A product marketing manager copies photos from a photo-sharing
website to use for his company's website.
- A teenager goes to someone's personal website and downloads a photo
to use as a wallpaper for her computer monitor.
- A guy copies to his iPod a song he got from a local band's MySpace page
and uses it to mix in his own guitar work on top and share it with his friends.
If you said that the product marketing guy's use of the photo is
copyright infringement because he's using the photo for business
purposes, while the teenager using the photo for a computer wallpaper
is not an infringement because it's a personal use, and that you
aren't really sure about the song on the MySpace page, then you're as
confused as most people are.
Here's the twist: It was a trick question. We don't really know whether
any of these are infringements because I didn't say whether any of
the artists stipulated "usage terms" associated with their works.
Unless copyright owners indicate which uses (if any) are permitted,
then some uses of these works may be copyright infringements, while
other uses would not be. Trying to figure out which uses are permitted
and which aren't is actually besides the point for the moment. People are
aware of what "copyright" is, but that doesn't affect their understanding
of "compliance." Far more importantly, it doesn't affect their online
behaviors. The economic and social consequences to this reality of
21st century culture are profound in ways that most people are only now
beginning to appreciate. The question is what to do about it.
Copyright law lies at the core of two critical aspects of society: it
is the financial basis for many industries, which is responsible for an
enormous amount of economic activity that creates jobs and fuels growth;
and it's an important element in the expression of creative thoughts and
ideas, not to mention free speech. Copyright acts as a catalyst for
education, dissemination of information, and a stimulus for innovation,
all of which intertwines back into the economic engines that finance the
mechanisms that make it all possible. So, the mutual dependency between
"finance" and "freedoms" creates a paradoxical relationship between
"access vs. infringement".
Before the internet, managing copyright issues has been a manageable
process because the dissemination of copyrighted works was difficult
and more easily monitored. Access to and use of content was controlled
through publishers and others that represented the interests of the
copyright holders. But the internet has complicated things significantly
because works are now being created by everyonemostly consumersand
scattered throughout the internet without any of the structure of
traditional distribution models. This has been largely a good thing for
society and economics on the whole, but it has also introduced some
serious problems. Many of the uses of such content by certain types
of entities are copyright infringements, and the rapid growth of it
has ballooned to proportions that the original framers of copyright law
never anticipated. Because the copyright holders of such content are
consumers, they generally have no awareness of the value of their works,
nor would they have any practical or meaningful legal representation,
even if they were magically aware of such value.
Stock Photography, the Consumer, and the Future,
describes how corporate use of user-generated content is extremely
prolific, but that the licensing of it is so inefficient (even where it
exists), that the transaction costs alone (not including the actual
license fees, if any) account for billions of dollars of wasted or lost
economic activity. Among copyright infringements, these claims cost both
sides thousands of dollars, totaling billions more in wasted economic
activity (not to mention the indirect costs time and productivity that
such actions siphon from both sides of the legal dispute). Most of these
cases are settled out of court with confidentially clauses attached,
so you'll never hear and read about them in the media.
Web 3.0: Content has Value
The increased "use" of user-generated content (in ways the should require
licensing) reflects what is emerging as the next "new economy." The article,
The Economics of Migrating from Web 2.0 to Web 3.0,
describes how the quantity of this sort of user-generated content is
not just growing exponentially, but all of itold and newis
becoming better understood and "mined" more efficiently via semantic
search algorithms. Searches engines are now unearthing dormant data that's
been online for years, but were previously unknown. With all this new
content being found and better understood, the economic byproduct of
Web 3.0 is that content is king, no longer just traffic, as it was for
One measurement of this trend towards content acquisition is the rapid
rate of end-user agreements being rewritten by social networks and
content-sharing sites. The revised language in these agreements require
end-users (who never read them, and just click the "Accept" button)
to unwittingly grant the hosting site unrestricted and unlimited use
of whatever it is they contribute. Some of these agreements include
redistribution and relicensing rights, terms that would otherwise
cost thousands of dollars under more traditional licensing terms.
These companies are grabbing these rights because they know there's
future value; they just don't know how to monetize it yet.
But their ambitions also come with risk. If someone uploads content
that isn't their ownan extremely common occurrenceand the site
thinks they now have rights to use it, there's a lawsuit waiting to
happen. And these sites will never effectively get indemnification from
these end-users; they're consumers. So the company may have access to
content that is, in the end, too risky to use.
Risk aside, the basic presumption is correct about the value of that
content, that it has intrinsic value. That no one has figured out an
effective way to monetize it is due, in part, to the barrier just
outlinedthey can't necessarily trust the content or the person that
contributed it. Addressing that concern is therefore a pivotal lynch pin
in unlocking the future to the next "new economy."
People often look at record labels, movie studios, television networks
and stock photo agencies as potential trail-blazers to lead the way
for figuring out how to capitalize on user-generated content, since
they're already in the business of managing and licensing media.
But those business models only work (and marginally so on very thin
margins) because they limit the amount of content they represent.
And they do this becauseas is the old business model before the
internetthey have to invest in each piece (even at minimum levels)
in order to get a return on their money. Such business models are
impractical for dealing with an unlimited amount of user-generated
content. So, existing media companies are no more capable of finding
a solution to this problem than anyone else.
Of the many proposals proffered to address this array of problems, they
can be broken down into two distinct camps: One is to either rewrite or
dispense with many copyright laws, and the other is to develop "digital
rights management" (DRM) technologies to use as an oversight and policing
mechanism. Both approaches are premature at this juncture. It's not
that I dispute the potential value of each: there are deficiencies
in copyright law, and DRM instruments could be effective under certain
market conditions. But my position is that they don't solve the problem
Though it may sound odd to hear it, there is one and only one barrier for
solving this business problem: Copyright Registration. To understand
that, I'm going to start with basic mechanics and work my way back to
The Role of Copyright
Copyright is the mechanism that provides accounting, accountability, and
protection by way of copyright law. This is because only works that are
registered with the copyright office can be protected in a manner that
has any real, economic leverage and protection for a company to build a
business representing such works. Under copyright law, only registered
works are eligible for statutory damages. Here, if someone "steals" a
work, fines start at $750, and can grow to $30,000, per infringement. In
fact, they can rise to $150,000 if the infringement were deemed to be
willful. (This particular deterrent will be important to remember later.)
By contrast, the fines for infringing on a non-registered work is
whatever the prevailing rate would be if it were properly licensed in
the first place. (Ok, there are other values that can be calculated,
but that's besides the point in the bigger picture, since those are more
rare than common.) Since we're talking about micro-valued works such as
songs and photographs, a "prevailing rate" of $.99 (or less) per unit is
too minimal to justify the legal effort in pursuing infringements. It was
for this reason that statutory damages were created in the first place;
it's the only way to give meaningful protection to micro-valued works as
a disincentive to steal them. Without statutory damages, protecting the
underlying asset is impractical, thereby making investing in building
a business unlikely, if not impossible.
So, any business model that represents a creative work must have
that work be a registered copyright. Here's the problem with that: the
cost of filing a copyright registration online starts at $35. The paper
fee is $45. If you wish to correct a mistake, it is $115. If you wish
to record a transfer or lien, it is $95. To get a claim registered in 10
business days it is $685. But, even the lowest fee of $35 per
registration, it is financially impossible to invest in a business that
represents an unbounded, infinite supply of content. Remember, we're
talking about billions of works produced by millions consumers; no one's
going to spend money on copyright registration on a per-filing basis.
Not the publisher, and not the consumer. For reasons discussed later,
the rate of return on a per-unit basis is inconsequential, so any
investment on a per-unit basis will never be acceptable by anyone.
(Media companies register the works that they represent, which is why
they limit the works they choose to publish.)
In short, high registration fees is the first barrier to clear before
any business model can evolve. The second barrier is that, even if
the fees were zero, the copyright office doesn't have the resources to
process that many filings on a regular basis anyway.
So, how do you reduce the registration fees and increase the
administrative capacity for an unbounded number of applications?
Simple: Let someone else do ita lot of people. By outsourcing the
processing to private-sector companies, costs can come down (reducing
fees), and capacity increases.
While that sounds well and good, the devil is in the details. For such
a solution to evolve, we must start by setting up a basic infrastructure
for all these companies to work together. This requires both a technical
architecture, and policy mechanisms that also provides economic incentive
to stimulate both investment and participation by authors and buyers of
There's one more extremely critical factor: to avoid disrupting existing
internet culture of creating, sharing and exchanging content between
people in the most open way possible. This must be preserved in order
for the system to work at all, since it was that very foundation that
prompted people to create this content in the first place. If this can
be done while assuring copyright compliance, then the goal is achieved.
Introducing the Copyright Infrastructure
Whenever you have lots of "stuff" used by lots of people, what do you
usually have? An accounting system. And that's the elephant that's
missing from the middle of the copyright room: there is no infrastructure
for registered works to account for anything. So, the proposal begins
by defining a Copyright Infrastructure. This is more of a technology
backbone that includes a series of protocols and services that allows
for universal access to information about registered works. (Not all
works, just registered works.) Its function is to perform three very
basic, but critical services:
- Account for what's out there.
This is necessary so that anyone that wishes to use any given work has
access to a centralized information center that can identify it. It's
not so much that people would search for content like a search engine,
but rather, to identify information about content they already have or
want. This kind of identification is now possible through significant
advancements in "digital fingerprinting" for all sorts of media types.
There are various technologies that do this, and support varying degrees
of "quality," as well as varying media types. That aside for the moment,
the point is that any given work can not only have a unique ID, but
aspects about it can be analyzed to identify one or more other works
potentially embedded in it, were it to be a composite. For example,
a mash-up of different songs into one; or a composite of more than one
photograph. Digital fingerprinting is addressed again in the section on
For example, if you create a home video and overdub a small portion of
a popular song, and then upload it to YouTube, the site will prompt you
with a warning that it contains copyrighted material and that it could be
removed. Similarly, tineye.com can not only uniquely identify a photo, but
it can determine whether other works are embedded in it withing different
levels of proximity. It's beyond the scope of this article to get into
details of these technologies; the only point is that it is possible to
uniquely identify works for purposes of the Copyright Infrastructure.
- Certify and authenticate who owns works.
Here, it's critical that the owner of a given work is unambiguously
identified and authenticated. Remember, we're talking about registered
works, where this information is already required for certification.
Having this information in the Copyright Infrastructure not only
minimizes fraudulent selling or licensing of content, but protects
others from buying content from unauthorized sellers, who may do so
in order to entrap them into lucrative copyright infringement claims.
It's incumbent on the publisher to protect himself by assuring the
credentials of their sources, so the Copyright Infrastructure's role
is critical in this regard.
- Assign an authorized representative of the work.
This is the key to the economic functioning of the system. That is, just
accounting for information is only useful if the person querying the
system knows where to go to get it. The agent would do more than just
sell or license the work, but it could also act as a formidable negotiator
for both pricing and any legal claims. This would balance the playing
field between content owners and users, thereby providing incentives for
parties to strike larger-scale licensing deals. I get into the details
of this in the next section.
The easiest way to illustrate why these three components are essential
is by citing a similar kind of infrastructure used by Apple to manage its
relationship to music labels and television networks for distributing
songs and TV shows through iTunes. There are two things to observe here.
First, there's a lesson in behavioral economics: Consumers don't fear
copyright infringement consequences, companies do. Markets don't
grow by educating individuals about copyright; compliance is achievedand
business growsby creating convenient and automated mechanisms that make
both access to and use of content easier. The recent announcement by
Apple and record companies to remove copy protection mechanisms in songs
further reflects this economic reality.
Secondly, iTunes makes money not just because the songs are protected
by copyright, nor because these companies vigorously enforce those
copyrights, it's because there are economic incentives for these companies
to strike licensing deals to work together. True, those incentives
wouldn't exist without copyright, which brings me back to the Copyright
Infrastructure. One of its key goals is to create a series of mechanisms
and incentives that encourage companies to work together. If user-generated
content were registered works that were represented by formidable business
interests, these works would be participating in similar licensing
mechanisms along side music, TV shows and movies. Better yet, there would
be incentives for companies to build "convenient, automated mechanisms"
into their products that deal with this content appropriately. This would
not only reduce the number of infringements of user-generated content,
but would have a significantly positive economic effect for many players
along the supply chainall the way down to the urban kid that created
the song or photograph in the first place.
If such an opportunity for growth exists, why can't companies build
similar distribution mechanisms like iTunes? Put bluntly, they
can't. It's literally impossible to do anything with user-generated
content because there are no mechanisms to know anything about any
given work. If a marketing guy downloads a bunch of photos from
a photo-sharing site that he wants to use for his company's website,
but leaves them in a random folder on his computer for a few weeks
before he begins to design the site, how can he recall where he got
the photos? What work and effort is required for him to track all this
down? What incentive does he have to do so, other than the good will
in his heart? Put another way, what disincentive does he have not to
infringe? Currently, he isn't even aware he has to do anythinghe just
uses the photos and goes off to his next task. An infringement that few
people know about, and few will ever do anything about. It's also lost
economic opportunity that, in the end, also harm's the company.
To wit, consider if his production software were copyright-aware.
Whenever he opened a photo in, say, Photoshop, it could query a copyright
server online, which would identify the authenticated copyright owner, and
return a web link to the authorized agent for licensing. Photoshop could
then use this information to alert the marketing guy (via a dialog box)
that he was editing a registered copyrighted work, and it ask if he'd
like to license it for $.50 (as indicated by the automated licensing
agent's server). The guy could choose to either license it or not
after all, only he knows if he's using the work in a way that would
require licensing. If he ended up using the photo, but not licensing
it, the licensing agent's own web crawler might eventually find it and
look up whether it was properly licensed. If not, it would contact
the company and ask for proper payment, or be potentially liable for an
Since the marketing guy that used the photos was given prior notice
and opportunity to properly license them, opting out might be deemed
as a "willful" infringement by a court, which would escalate the damage
awards. Such automation would more likely encourage him to comply with
the initial licensing request, which wouldn't have been expensive or
time-consuming. Not only are there economic incentives to comply with
copyright, but if the process were automated through non-human means and
built into products and other services, then we are even more assured
of compliance: it's easy.
Scaling up such an automated system to support an unlimited supply
of works is not complicated, and it doesn't take an economics wizard to
appreciate how such a business with very low overhead can be very
profitable, even with micro-valued content like photos and songs.
The original copyright holder wins too, because he'll be collecting
revenue from the vacation photos he took on his Hawaiian vacation.
At this point, let me be clear: I am not concerned with teenagers that
download music and photos for personal creative experiments of mixing
and remixing to create derivative works that they share with their
friends. At least, not directly. It is only if those works end up in
the hands of a publisher that wishes to monetize that content, or
use it in ways that would have required licensing, does copyright become
One of the other prime directives is to avoid disrupting existing internet
culture, or of dissuading users from is creating and sharing content
freely, so the Copyright Infrastructure is designed to be agnostic of
finance, or of provisions of copyright law itself. It does not determine
"right and wrong," or permitted uses of works, or even of policy. It
is merely a grand accounting system; a mechanism by which the services
that sit on top of it define the objectives of various service providers.
Even free and public domain content needs to have some sort of accounting
mechanism to give assurances to those who wish to use it that it is
what it claims to be, and that it is, in fact, either free or in the
public domain. These are also issues I'll address in detail later.
Unfortunately, the Copyright Infrastructure isn't something that
an innovative entrepreneur can conjure up and implementat least,
not one that anyone would ever use. Remember, the two key barriers
to entry are high registration fees, and the limited capacity of the
copyright office to process all these applications. So, in order for an
infrastructure to be built at all, the US government has to put it into
motion by addressing these two problems.
When it comes to fees, the reason they are high is because the copyright
office is the sole entity that can process applications for copyright
registration. Therefore, it's a costly process. It also happens to be
the only entity that knows all of the works that has been registered,
and it also controls the registration fees and the timeliness by which
applications are processed. There are no third parties that can do any
of these functions at all. And that is the problem. So, my proposal is:
The copyright office should divest itself of its administrative
duties of registering copyrights, and institute a program where private
companies can become certified Copyright Registrars, where their role is
to accept and process copyright applications. Additionally, there should
be the creation of a worldwide, publicly accessible database that contains
all registered works.
Copyright Registrars would offer at least one basic service: process
copyright registration applications. Because there would be many such
companies, the higher volume of servicing agents would not only lower
the cost of processing applications (thus, lowering fees), but also
increase the capacity of applications that can be processed.
If this sounds complicated, it's already been implemented quite
successfully via domain name registration. The policy-setting and
oversight functions are performed within the not-for-profit public-benefit
corporation, ICANN. Their mission, as their own
website states, is to "keep the Internet secure, stable and interoperable.
It promotes competition and develops policy on the Internet's unique
The copyright office should have the exact same organizational structure,
including a similar relationship to Copyright Registrars. The government
sets policy and performs oversight, just as ICANN does over domain names
and numbers. A copyrighted "work" is, in fact, a unique identifier of sorts,
where the work is associated with an owner for a period of time. It can
be transferred, or even released into the public. All of this is possible
today with copyright, but the mechanisms to do so are opaque and onerous.
Similarly, the works themselves, like domain names, are held in a global,
unified, distributed database, which can be viewed by all, and updated by
the registrars. The moment a new work is registered with any registrar,
it's propagated to the internet much like domain names are. More on
The role of the copyright office, like ICANN, would be to "keep the
copyright registration process secure, stable, and interoperable, while
also promoting competition and developing policy on how copyright is
administered by its registrars."
Also, nothing has to change about copyright. The laws, legal
interpretations, court rulings and precedents, and almost everything
else remains in tact. There would be a number of things to address from
a policy perspective, discussed in the Policy section later. But it's
important to understand that the proposal is merely a shifting of
administrative duties from the government to the private sector, and
the formation of a grand network-based, distributed accounting system.
The copyright office's responsibilities should be to policy and perform
oversight duties, also addressed in the Policy section.
As for the database of all registered works, it would be centralized,
distributed, and world-readable, complete with a series of network
protocols and APIs for applications to mine it. Each work in the database
would contain information that's already required by the copyright
application process, including the work itself, the author's identity,
and so on. Once a work is registered, it would be immediately certified
(the need for a certificate should be eliminated), and it would immediately
show in up the copyright database for all to see using the web or any
other copyright-aware application.
Though people could look up what works have been registered by viewing
them directly, the visual representation should be as close to a
"read-only" format as possible. For example, most content could be
rendered as a lower-res visual and embedded with watermarks. If the
searcher wanted to find original source material, contact the authorized
NOTE: In my original proposal to the copyright office of this idea,
only the idea of the online database of registered works found its
way into the text of the Orphan Works Act of 2008. But it was very
loosely and ambiguously described. The act passed the senate, but
not the house.
There is one more item that would be in the only copyright database (not
necessarily part of the copyright "record" for the work or collection):
the legally assigned agent authorized to sell or license the work.
As I described in my example scenario above, if the marketing guy found
a photo that he wanted to use for his business, there would be automated
mechanisms that could query the copyright database for the authorizing
reseller agent, who would then conduct the transaction.
The fact that there is a reseller agent is critical to the economic
viability of the Copyright Infrastructure proposal. This is actually
where the incentives are for massive investmentnot in capital or
other resources, but in business opportunity with a very high ROI.
Technologically, a Copyright Registrar is very simple because accepting
and processing applications isn't very complicated, nor does it require
much capital. In fact, being nothing but a registrar wouldn't make
much sense, as competition would drive registration fees too low to
to make money. (Remember, low fees was one of the primary goals.)
Therefore, companies would have other incentives to be registrars, and
that would be centered around acquiring the right to be the designated
licensing agent for the works being registered. Therein lies the real
value, and how the business model evolves: the right to represent the
sales and other legal interests associated with massive amounts of
works has high valuehigh enough that it provides incentive to
offer these registration processing services in exchange for those
rights. The free market of many competitors would result in a likely
distribution of proceeds of the licensing fees to go back to the
copyright holder, which feeds back into the economic incentives for
people to want to register their works.
Note that a registrar can operate many different business functions
at the same time. Indeed, this is expected. It may also happen to
be a social networking site, or a content-sharing site, or a discussion
forum, or even a law firm. In fact, as will be discussed below, it's
likely that registrars would have multiple business models from many
different business sectors because the more services it offers, the more
likely it can gain representation rights to more content.
As for the role of the licensing agent, it's not rocket science either.
In the simplest case, selling and licensing can involve little more than
a shopping cart application. As the market evolves, even that aspect
is likely to be automated by an auction-based pricing system similar to
how online advertising rates are derived. For Google's adsense product,
you display the ads, and let the open-market determine the price you
get for clicks. Same thing with copyright: you take the pictures, and
let the open market determine the price someone pays to license it.
I'll discuss the more complex effect on market pricing in detail later.
As for the legal representation, the registrar can perform all its duties
and responsibilities itself, or it can outsource various tasks to third
parties. For example, copyright infringement claims can be farmed out
to law firms. If the concern of a world full of litigating lawyers causes
your eyebrows to rise, see the Policy section later.
Whatever the registrar's rights and responsibilities, it doesn't have to
affect the copyright holder's actions with his own works. In fact, that's
the whole intent, as will be described more fully later. He still owns
them, so he can still post them on social network websites or send them
in email to friends. It's only if someone wanted to license a work would
the copyright database be queried to determine who the licensing agent is.
Registrar Business Incentives
Though it would probably shock most people to believe that some guy's
vacation photos to Hawaii will yield him any money in licensing fees
over the course of time, this represents an old, bygone era of economics.
The financial model in being a registrar has its basis on a proven
and modern principle that didn't exist until the internet came about:
long tail economics. To many, this is a fundamental paradigm shift.
Long-tail economics states that everything has potential value,
and profit is made when the per-unit cost of sales and of warehousing
becomes independent of quantity. Amazon.com brought it to public
lightand the term was coinedwhen an editor for Wired Magazine once
noted that amazon's sales of little-known books exceeded the sales of
very well-known, "best-selling" books. The reason for this was that sheer
volume of these little-known books, sold in quantities of ones and twos,
was so numerous, that it outpaced the smaller number of top-sellers,
even though those per-book sales volumes were quite high.
When it comes to non-physical goods like user-generated content, the
model is even more compelling. That is, if the cost of doing business
is roughly the same whether you represent 10,000 photos or 100 trillion
photos, you might as well take in as many as possible. As the per-unit
cost quickly approaches zero, why limit the supply of content available
for purchase at all? Therefore, the real objective is to acquire as
much contentbetter yet, the rights to itas you can, and let
the buyers find and pay for what they want through an automated process.
To forecast where this is going, consider these scenarios:
- Record companies could make billions more dollars if they
accepted all songs ever submitted to them, provided that they didn't
spend money to market them. Instead, they merely "sold" whatever someone
was interested in buying.
- Photo agencies could make billions by accepting all photos ever
submitted to them, provided they didn't have to do anything to promote
them. They merely handled the transactions in the event that someone
wanted to buy or license them.
- Movie studios could make billions if they accepted all videos,
films and screenplays if they didn't have to actually do anything with
them, other than to license them to potential buyers the same way as
the other publishers of creative media.
The skeptic would counter that there are two fundamental flaws. First,
"no one would submit their creative works to any of these publishers
that would 'do nothing' in return." The response to that is that people
already submit these works freely to the internet everydayand they
do so in return for nothing. Consumers may not understand or care about
copyright, nor the economic value (or potential) of their creative works,
but they put their stuff online anyway. Now, imagine if they were told
that they could do exactly the same things they do today, but could
potentially make money as well? (In a survey I've done of over 300,000
random visitors to my website, 91% said they would participate in such
a plan. I'm guessing the other 9% didn't quite understand it.)
Better still, in the event that someone did want to buy or license
a given work, the consumer who snapped the picture would be much happier
if someone more experienced, influential and powerful did the price
The second thing the skeptic would say is, "these publishers don't
represent unlimited content because it would be cost-prohibitive to
manage." That would be true if the management process was the same as
how they handle their existing library of works. But that wouldn't be
the case in the scenario I propose here. These companies wouldn't have
to do a thing with all content submitted to them, other than reply to
requests to buy them through automated mechanisms. If it turned out
they were representing works that someone else expressed interest in,
they would be more than happy to assess its value and monetize it.
None of this suggests they would otherwise cease or alter their other
business activities. Music labels would still promote bands that tour
well and sell tons of songs; movie studios would still spend hundreds
of millions of dollars making and promoting blockbuster action films;
and stock photo agencies would still make invest in their featured pro
photographers. They can do all this and also monetize content submitted
by people who create content and upload it to the internet.
The concept of any work having potential valueand that value being
protectedpresents a completely new and different marketing paradigm
of this next "new economy."
Traditionally, people have sought "publishers" of various sorts to not
only represent and sell a work, but to promote it. But here, it is
the creators of such works who do this, just as they already have been
throughout the Web 2.0 era. The very act of uploading content and
advocating others to share and distribute them, or even to modify and
redistribute them, only serves one's own interests. It doesn't matter
whether a work is emailed around, or sitting on someone's personal
computer, the Copyright Infrastructure would always be able to identify
it, even if it were combined with other works. In the event someone would
use the work in a way that would require licensing, the work would be
identified, the licensing agent would conduct the transaction, and the
author would be compensated.
Any other use of the work, such as those that wouldn't require licensing
(as permitted under Fair Use provisions), would just be yet another form
of distribution and promotion that, again, increases the chance that
someone else would want it someday. In a system where there is proper
accounting, protection and enforcement, freely sharing and distributing
content actually rewards people.
Regardless of the author's intention when he creates a work, there would
be many registrars to choose from. Choosing one would be based on either
financial or social objectives. Perhaps even both. Note that any given
registration filing represents one or more works at a timeno one
is obliged to use the same registrar for all his works as a general
ruleso he may have multiple personal objectives.
Below are some of the candidates likely to become registrars that would
be most desirable by content creators:
- Search Engines and Domain Registrars
Google, Yahoo, MSN and other sites have the infrastructure and wherewithal
to immediately and instantly scan the copyright database (if not help
set it up in the first place), index it, and serve up content relevant
to search queries. They have the administrative and technical resources
to realize the backbone very quickly. Creators of works may choose these
entities simply out of convenience and reliabilitythey're first, they're
there, and more people use search engines than any other site, making
them first in line to attract sign-ups. The search engine may choose to start
some level of media promotion for the content it represents, which itself
would attract even more people.
- Social Networks and Content-sharing Websites
These already have strong name brand recognition, thereby attracting
a lot of interest from their existing end-user community. Photo and
video sharing sites are already heavily trafficked by companies that
seek content hosted there, so it's a natural sales pitch for consumers
to use them as registrars.
- Media Agencies and Publishers
As has already been discussed, record labels, movie studios, television
networks, photo agencies, and other publishers and producers of media
content would be good candidates to be Copyright Registrars. Technically,
their lack of significant traffic for user-generated content may appear
to put them behind the 8-ball, but they already have inroads with many
media companies, and also have more sophisticated licensing models
that may help translate to higher license fees, which translates to
more income for contributors.
- Special-interest Groups
Industry trade associations for creatives could each act as registrars
for individuals who create particular or specialized forms of content.
These organizations may provide value-add by having re-distribution
or re-licensing relationships to others that may benefit content
creators. They would act somewhat like a meta-agency (an agency that
licenses work to other agencies).
- Existing Content Aggregators
News outlets, magazines, newspapers and other media outlets could attract
more content from contributors and at all levels, including their readers.
Like the media publishers above, creators of works might choose any of
these entities because of their specialized positions within particular
industries, but may also be the go-to place for authors that don't want
to work directly with a single media agency or publisher.
- Specialized Technology Companies
Companies like Picscout and Idee (who develop image-recognition
technologies and provide some extra services) already charge fees to
track whether registered works are used by others on the web, but they
may provide additional incentives to those who use them as copyright
registrars as well. Their decisive advantage would be to find
- Creative Commons
For those who wish to make their works available for free (in exchange
for credit and/or attribution), the Creative Commons could be an option
as a registrar, should they choose to become one. The CC is an
organization that promotes and advocates the free use of intellectual
property (including copyrighted works). Here, the CC would assure
compliance by "licensees" of these works (even though no money would be
exchanged) by enforcing the creator's copyrights via the registration
process. In other words, if attribution isn't given for using a
CC-licensed work, the user could be liable for copyright infringement and
be awarded statutory damages. It's not likely such a thing would happen
if the use of the work was free in the first place, and the internet
would be laden with web crawlers looking for such infringements.
So, let's consider an example: Say a musician registers his songs with
a record company as his Copyright Registrar. He then posts his actual
songs on MySpace (in addition to other sites), where he promotes it
and gains traffic. If someone wanted to buy it, the user would be
automatically directed to the record label's ecommerce page where the
transaction is made.
Why would the musician pick the record label and not MySpace as the
Copyright Registrar? His choice might be that he also gets grouped
together with other similar bands where he may receive extra exposure by
association. Or, the label itself may do more broad netcasting of itself
and the kinds of bands and music it represents, which may also stimulate
sales of his particular songs. Maybe he was given other incentives by the
As for MySpace, it may have seen the lost opportunity, but it would be
impossible to stop it. Identifying the source of copyrighted material
would be enabled by copyright-aware applications, which would include
the web browser, and would have nothing to do with MySpace. While the
MySpace page might not display the specific link for the buyer to use,
no site can stop a user from navigating away to another location.
Here, merely displaying content does not require the site to also
display copyright information or list the licensing agent. See the
Policy section for further discussion. The web browser's support of
the Copyright Infrastructure would be a more likely conduit for this
kind of information anyway.
Remember, no site is required to host or display content it doesn't
want to. Here, MySpace could adopt a policy of only permitting content
that it, itself, is the licensing agent. Or, it could strike co-marketing
and co-licensing relationships with other websites or registrars. That's
just a matter for the free-market to work out on its own.
But there is the market reality at hand. There's still quite a bit of
money to be made in the old Web 2.0 model, where merely having the content
reside on a site attracts traffic and advertising revenue. A site like
MySpace would have to do some critical analysis to be so draconian as
to not host content it doesn't necessarily have the rights to.
Ultimately, content creators will help registrars discover which models
work best. The same musician may choose to register different songs
with different registrars, just to test which performs better insofar
as revenue is concerned. (The Copyright Registrar only gets to represent
the works registered with them; it may not necessarily choose to promote
him.) In the end, competition is good.
Because the Copyright Infrastructure would be designed so that registrars
would have exceedingly minimal costs in processing registrations, and
a great deal of financial incentive to act as the representatives for
such works, it follows that the nature of competition will be such that
"fees" that creators would pay to register copyrights would be either
zero, or absorbed into the existing "membership fees" members pay anyway.
This is extremely critical for the system to be accepted; it won't
take very long for many consumers to realize their works aren't going
to generate much (if any) income, even though it's never possible to
predict what kind of sales diffusion might result. But, since the laws
of long-tail economics state that it doesn't cost any more to represent
an infinite amount of contentsor save any money to accept lessit's
always in the best interests of registrars to encourage everyone to
participate, even the losers. If participation fees are either free,
or included along with other desirable or useful services consumers would
otherwise already sign up for, then this helps with innovation uptake.
On a related not, there may be those who wish to only register their
works, and not give up representation rights. that is, people may want
to represent themselvesfor example, professionals, like photographers.
Here, the Copyright Registrar could perform the service for whatever
fee the market would bear.
Of course, in order for Copyright Registrars to pass on low fees, the
copyright office should charge as little as possible to registrars. It's
too soon to know what the copyright office would need in order to fund
its minimal operations, but the economic objectives must be kept in
mind in order for the entire system to succeed.
Evolution of Pricing Models
It would be impossible to predict how prices would evolve. However,
classic Economic Evolutionary theories can predict the path of price
evolution under the conditions proposed here, because they are similar
enough to other models. The greatest unknown is the time it takes for
buyers and sellers to become "rational players"that is, when they
understand that content is to be licensed, not stolen. This when I would
start the comparisons with other models.
One such model is the pricing volatility associated with online advertising
rates before Google stabilized the market. Prior to Google, online ads
were sold the same way as most online content is today: peer-to-peer,
with no "intelligence" in pricing structures. Some sites were getting
thousands of dollars for banner ads, whereas others were getting only a
few dollars, exactly the same way photo licensing prices swing today.
(My own sales data for my stock photography licensing corroborates
broader industry data.) In old online advertising, pricing inefficiencies
were aggravated by a lack of standards to test and verify traffic
metrics; nor was there a true understanding of what end-users actually
wanted when they visited a site (a concern that ultimately lead to
keyword search data as a metric for setting advertising rates). The
same metrics are currently unavailable for stock photo licensing as well.
What Google did to stabilize advertising pricing volatility would
be similar to what might happen with license fees for arbitrary
media content online, once empirical sales data was collected under
the new Copyright Infrastructure model. Here, that evolution of pricing
models will include many components that may share some attributes with
advertising rate analysis, such as site origin's ranking and historical
sales trends, keywords, and so on. Another component for automated
pricing structures for media would include the use of automated
content-recognition algorithms that recognize certain patterns, subjects,
and various other attributes and assessing which sell for more or less.
When combined with appropriate metadata and other known features,
categories emerge that can assist in producing the beginnings of pricing
formulas that could apply broadly to many different media types.
Evolutionary Economics also predicts that pricing tiers would naturally
evolve later as well. One can see this very phenomenon in the music
industry, where Apple's new licensing deals with record companies
are migrating away from a flat $.99 price to a tiered structure:
new releases would be priced at $1.29, traditional music sold at the
current $.99 rate, and older music would be priced around $.69. It is
entirely likely that licensed content used for business purposes would
follow a similar path towards eventual tiered pricing. Here, factors
would depend on usage terms such as advertising, promotion, editorial
publishing, and products and services. Additional variables may later
incorporate the timeframe in which the work is used, terms of exclusivity,
and size and/or "resolution" of the media.
As pricing becomes more "intelligent," it would likely happen that
Copyright Registrars might get out of the business of setting prices
themselves and instead, outsource such "market rate" intelligence to
third parties in exchange for a percentage of the revenue. Again, this
is similar to Google taking a cut of ad revenue in exchange for providing
Any concern over price manipulation would be alleviated by virtue of there
being many competitors, and the likelihood of a universal auction-rate
mechanism, similar to how Google's ad rates are calculated.
"Fee or Free"? Ownership is Separate
There's another important factor to consider when it comes to price:
the concept of free. First, understand that copyright ownership and
monetary remuneration are entirely independent from one another. If you
take the picture, you own the picture. You may want to give it away
for free, or charge a lot of money for it, but the role of copyright
registration is independent of that. In fact, the ability to enforce
pricing, including "free", is weaker without having registered the work.
Consider the buyers: Can they trust you? Not just that that you won't
sue them, but that you actually own the work and have the right to charge
nothing for it? How do they know the work really isn't owned by someone
else who will then sue them for infringement? Because of this risk,
buyers have to be wary of any work they acquire in an environment where
legal action is potentially heightened. This, especially if one promotes
"free" usage of otherwise valuable contentit could easily be a ploy
On the other hand, the "buyer" could also be a bad actorhe could claim
(also by lying) that the work did have a CC license, which would
allow him to avoid license fees. This would complicate the copyright
holder's efforts in winning an infringement claim.
It is for reasons like these that courts look at one and only one thing
when deciding copyright infringement cases: documents between the parties
that grant consent by name. This documentation does not have to be complex
or even written. It can be a digital transaction on a website. The terms
of use can be a CC license agreement, but it's the direct granting of
consent that's critical here, and the role of the "licensing agent"
in my Copyright Infrastructure proposal gives assurance to both parties
that the exchange is authorized.
Digital Rights Management
So long as someone has a business interest in selling content, it implies
an equally important interest in protecting it. There are different kinds
of entities that infringe, which require different protection methods and
models, but they essentially differ as "consumer infringements" and
For consumers, there are two ways to deal with infringements: psychology
and technology. On the tech side, there are Digital Rights Management
(DRM) technologies that publishers use to assure that people only
gain access to content they purchased. On the psychology side, we have
behavioral economics, which, as I've already discussed, treats the problem
using more carrots than sticks. Apple's recent announcement that it will
no longer use DRM mechanisms is an example of this approach.
But the reason why the Copyright Infrastructure model helps both sides is
because it is rooted in mechanism, not policy. That is, the mechanisms
are open-ended enough so that people can enforce whatever policies they
want within the same infrastructure. Having policy-agnostic mechanisms
like this is becoming increasingly more useful because more and more
applications and devices are internet-enabled. This gives them access to
copyright servers where they can query copyright records for any given
content in real-time. Works themselves do not have to have restrictions
embedded in themrather, their "status" (and that of the end-user
and/or device) can be held elsewhere (i.e., in the cloud).
For example, if someone tried to play a movie on an iPod that was
connected to the internet, the following would take place:
- The movie's digital fingerprint (a value that can be determined
in real-time using open-systems algorithms) would be sent to the Copyright
Infrastructure for identification.
- If the work were a registered copyright, the iPod would be given
a link to a URL associated with the licensing agent for that movie.
- The iPod would exchange information with it, such as the ID of
the iPod player and/or other identifying data for the device or user
(including the ability to enter a password that may be affiliated with
a debit account for playing movies).
- The licensing agent would then return a confirmation or rejection code.
Implementation of the communications protocols over the wire would
almost assuredly be open-source software and made available for free
to any application and device maker; embedding it into applications
would be effortless. This doesn't guarantee that Apple or anyone else
would cooperate in such a mechanism, but its simplicity makes it a
whole lot easier to "switch on" later if license agreements were later
established. The advantage is that decisions can be made at any time
without having to revamp technology, or alter or regenerate digital
content, or update firmware.
This greatly simplifies the entire architecture and implementation of any
DRM technologies that anyone would choose to develop. Such simplification
could also unite many of the fractured entities battling it out to see
whose technology would be used. One such example is a consortium of
companies called the Digital Entertainment Content Ecosystem (DECE),
which almost everyone is a member of, except for Apple. But since
Apple commands more than 70% market share for at least music sales, the
DECE's influences are minimal. However, the Copyright Infrastructure's
potential to simplify and even unify approaches to DRM technologies,
it would completely reshape the way companies cooperate on DRM.
The goal here is not necessarily to settle DRM disputes within other
consortiums. It's broader than that. Future media will come from a lot
more user-generated content, not just movie studios and record labels.
And those consumers who generated this content have exactly the same
rights as the larger media companies in the eyes of copyright law.
Gadgets like the iPod will have to deal with an infinite number
of licensing agents, not just bigger companies, simply because these
user-generated would be represented by formidable registrars, ready to
pounce with punitive infringement claims for non-compliance.
People don't currently understand that today because most consumers
don't have powerful companies behind them with batteries of lawyers to
assert their rights. This is a failing of the copyright system itself,
not its laws or enforcement procedures. Hence, the goal of the Copyright
Infrastructure is to remedy this.
The role of government is to establish frameworks (laws and policies)
that stimulate growth through incentives for participation and good
behaviors, and disincentives for disruption. It can do this best by
providing mechanisms to build and innovate new business models that
can expand, and setting policies that rewards behaviors that have a
positive social or economic effect. The purpose of the Copyright
Infrastructure is to provide the means and specifications for companies to
participate in the copyright process without promoting or advocating any
particular licensing model. The marketplace can do that. However, in order
to provide those mechanisms, the copyright office must assure that
certain business-critical issues are addressed, or this infrastructure
cannot be realized.
One of the main goals is to provide as many incentives as possible for
people to participate in the copyright system: for creators to want
to register their works, and for users to want to use those works.
By not mandating anything, but instead, providing incentives, then
the unexpected nature of the free market can do whatever makes sense.
In keeping with this spirit, the copyright office should attempt to
derive its revenue in ways that don't discourage creators from participating
in the process, and the primary factor here is the minimum registration
fee. By keeping it as low as possibleby incentivizing Copyright Registrars
to do so as well by not charging them per-unit feesthis goes a long way
towards that goal.
It is also important to recognize that the registration of works can
be a very powerful deterrent for people to infringe, but it can also be
a tool for bad actors to spoil the system. Therefore, policy needs
to assure that people's interests are protected, for those on
both sides of a work: the creator and the end-user of it. For example,
policies should be written to ensure that registrars do not engage in
practices that dupe contributors into unknowingly granting permission
to represent their works, or by using opaque and difficult to understand
user-agreements. Full disclosures must be made as to the implications of
granting representation rights, and the penalties of fraudulent actions as
dictated by copyright law. This would be coupled with the contributor's
requirement to provide personal identity information, a requirement also
mandated by copyright law. Users of all sorts should also reminded that
registration of a work is never mandatory, that not every work on the
internet would have to be registered, nor are potential users of works
required to use "only registered works."
Submitting Works to a Copyright Registrar
Currently, copyright law specifies that a "best edition" of a work to
be submitted when applying for copyright registration, but this is a
product of the print-media world. In today's environment, the copyright
office no longer uses this terminology for online registrationin
fact, it's ambiguous. The leading presumption is that authors submit
works in a "sufficient resolution necessary to identify the work
unambiguously in an infringement claim."
Needless to say, this policy needs to be nailed down and codified. The
goal should be to allow copyright holders to register the minimum
size necessary, thereby avoiding making a commercial-usable edition
of the work publicly available. A photo would only practically need a
lower resolution for most cases; a song would not need to be the highest
bit-rate or at full audio quality; a movie or video wouldn't have to be
HD format or even full frame-rate. So, if a user uploads his photo to a
social-networking site that also acted as the user's Copyright Registrar,
he shouldn't be required to upload the full-resolution edition.
There may also be reason to permit a method of a "hidden filing," where
the content is submitted, but not visible through the open-access viewing
mechanisms. Traditionally, such a thing has never been necessary because
copyrighted works have never been publicly visible through any means
anyway. But, once an online database were instituted, there could
be many reasons why copyright holders (or their clients) may need to
shield the visibility of works from public view. However, since this
is antithetical to the intent of the Copyright Infrastructure and its
underlying incentives and motivations, hidden views should be more than
exception than the ruleperhaps requiring special application subject
to approval? Note that if such a thing were considered, the work itself
would still be in the database, and it should still come up as a "match"
if someone submitted a copy of it to the database mechanisms to query
whether it has been registered. After all, if someone already has a
copy of the work, determining whether it's been registered should be
permitted. These functions are fundamental to the operation of the
Access to the Copyright Infrastructure
Section 1202 of the copyright code prohibits impeding or circumventing
copyright management systems. This was originally designed to stop
people from messing with devices like DVD players and iPods so they
could play stolen movies or songs. However, this section may need to be
amended to address issues associated with the Copyright Infrastructure. It
should never be required for any given application, device or website to
access a copyright database server, but if any such item attempts to
do so for purposes identifying a work, or to communicate with a licensing
agent, that access cannot be impeded or circumvented. The presumption
should be that anything trying to access the server is essentially a
copyright management system, and should be treated accordingly with the
provisions of 1202.
Implementing this could be tricky, especially because there is often a
long chain of devices that lead to the internet from any given starting
point. For example, if someone with an iPod is passing by someone's
house or office where there is a closed network, one should not find that
network to be in violation of 1202 because it doesn't permit access to
the copyright server.
There are similar considerations for each node along the network: even
if access up to any given node is possible, there's always the next
layer up that could have similar restrictions. It is the right of any
given portal to block access to any IP address, network mask, or domain
that it wants, but assuming that one is actually on the network in
the first place and can access "something" on the net, the question
is whether authorized copyright servers should be exempt from filtering
or blocking. At a minimum, one should not be permitted to block access
to copyright servers as a matter of policy. Policy-makers should also
look into nuances of how interested parties could achieve the same end
(blocking access to copyright servers) without appearing to.
Lastly, there should be severe penalties for schemes akin to phishing,
where users or devices themselves are duped into believing they are
talking to the Copyright Infrastructure cloud, whether it's a registrar,
a licensing agent, or parties related to the transaction.
This document is a proposal and economic analysis of a Copyright
Infrastructure. It is not a technical document. Though I've mentioned
that the technology to build and deploy this infrastructure is not
difficult, that doesn't mean there aren't going to be problems that policy
needs to address. For example, I mentioned earlier that there are many
technologies that can do "digital fingerprinting" of various media
types. The problem is that there needs to be uniformity on which
technologies are used by clients and servers (between the applications or
devices and the copyright database server in the cloud). If an iPod uses
one technology to determine the fingerprint of a song, the Copyright
Infrastructure has to use the same one when determining if it's a match
to a known copyrighted work, or they'll each come up with different
values, resulting in a failed match. The policy question is whether the
government wants to review all available technologies that can do certain
technical functions like this and come up with a "standard," or should the
protocol remain "open" so that the nodes on either end of a communication
channel to negotiate which to use during the initial handshake?
Copyright Infringement Claims
The issue of infringements is a big one. Because the internet is not new,
and content is already online, and third parties are already using it,
the sudden deployment of a copyright database that already has millions
of registered works (with billions more to come online in short order)
would create an instant shift in public awareness and liability.
Studies already show that millions of photos alone are already infringed
upon by companies that should be licensing them (see
for data). Under today's environment, these are incredibly hard to find;
even sites like picscout.com and tineye.com have only been able to
identify a small fraction of the internet's contents to find potential
infringements. However, once the copyright database became live, the rate
of infringement filings would probably spike to unprecedented numbers.
Anticipating such an event is a policy matter that should be taken up by
the copyright office, though it's not clear how it should be handled.
Should there be a temporary moratorium of statutory damages? A period
of time where existing infringers would have the right to legitimately
license whatever works they would otherwise by sued for? Would there
be new statutes of limitations during this period? All this should be
discussed in open debate.
Whatever that time frame window might be, it should be very short.
The very same event that unearth millions of infringements would also
cause an abrupt end to future infringements, since everyone would be
aware of it and want to avoid liability, and it is desirable to have
this happen sooner than later.
While infringements as we know it would go away, there would be a new
concern: false claims of infringements. To be clear, I'm not talking
about ambiguous claims, where it isn't clear whether the use of a
given work falls under the Fair Use provisions of copyright law. Such
occurrences happen all the timethat's why we have courts. Rather,
I'm talking about rogue licensing agents who file false claims in the
hope of gaining settlements. This is where the copyright office's policy
needs to be clear and effective.
The copyright office's oversight responsibilities should monitor the
actions of registrars to be sure they are not trying to make a business
out of filing infringement claims, or misleading or coercing people
into thinking they need to license works, or to take a position on
whether any given use requires licensing. These are, and always have
been, open to debate. Though educating people is a different challenge,
that is not something that the copyright office or Copyright Registrars
should be engaged in as a matter of conflict of interest. That said, the
concern on how and whether people will be properly informed will likely
evolve as a natural byproduct of there being many parties involved in
the process, representing those on both sides of a transaction.
Since copyright law and Fair Use provisions are complicated, it's
certainly expected that conflicts will arise and actions will be
taken. While monitoring Copyright Registrars is important, it's not
easy because rarely do claims reach the courts; settlement payments are
made instead, and they can add up to millions of dollars before someone
were to take notice. Therefore, Copyright Registrars should be subject
to periodic (and mandated) monitoring, including an accounting of the
number of infringements claims made. Obviously, since settlements often
contain nondisclosure language to protect companies from public exposure,
such monitoring of claim activity among Copyright Registrars would have
to be kept private as well, unless malfeasance is discovered, in which
case, appropriate remedies would need to be made to affected parties.
Registrars may need to be given specific guidelines and parameters on
how they may file claims. Use of specific language, contact information
for suspicious activity, and so on. At the same time, Copyright Registrars
should also have minimum requirements to actually engage in protecting
works they represent. Copyright holders should have some way of forcing
Copyright Registrars to take action, and ways to back out of their
legal relationship with them. Policy language and procedures for this
should also be discussed in open debate.
It should be noted that companies that already engage in representing
the copyrights of many otherssuch as stock photo agencies and the
likehave never shown a preponderance towards massive litigation
against corporate users. As a matter of course, companies like Getty,
Corbis and even smaller microstock companies have taken a much more
lax approach to copyright infringements, often settling matters without
legal action, or even punitive claims. The motivation has been largely
to preserve relationships with buyers to stimulate new business. This
very same motivation may be a similar incentive in a world where there
are many Copyright Registrars, and where competition is heavy.
Choosing and Changing Representation
Another matter to be considered is that of changing which registrar a
copyright holder uses as a licensing agent. (This only applies if
the registrar is also the work's licensing agent.) There may need
to be an ordained mechanism by which people can use in the event that
they aren't happy with the performance of their current registrar.
At the same time, registrars need protection in the event they are
already in the process of licensing negotiation with a buyer. And
then , what do you do about disputes between these parties.
Policy needs to assure that only authorized agents can act on behalf
of copyright holders, and that those people are not duped into changing
representation improperly. On a related matter, there is also the matter
of predatory practices by some companies trying to persuade copyright
holders to change their legal representation. Because the online
database would be made available to everyone, there would be those
constantly scouring the net, looking for high-value content creators,
or of potential infringements, whereby those companies would try to
inappropriately convince people to switch representation. This was
a serious problem with phone companies trying to get people to change
carriers all the time, a practice that the government eventually was
able to stop through legislation.
If authors opt to represent themselves or hire their own legal
representatives, they wouldn't be subject to the same litigation
policies noted above. This is how the law exists today, and is
sufficient. Copyright Registrars need to be treated specially because they
would be representing billions of pieces of works, millions of people,
and could therefore be able to do great harm. Thus, their behaviors and
actions must be more closely legislated and monitored.
Duplicate and Derivative Works
One potential area that may need to be addressed by policy is the
handling of bad or ambiguous content. This may be due to either quirks
of the data being submitted, or by bad intentions. For example, if
someone tried to register a work that isn't theirs, especially one that
has already been registered by someone else, it is currently considered
a federal code violation, punishable by fines that go up to $2500. More
punitive punishment can be asserted for gross violations as well. But
this may be rather excessive when dealing with mass registration
applications by consumersmistakes will likely be frequent. At the
same time, there must continue to be strong deterrents for those with
Complicating matters is the question of ambiguity: it is possible to
register works that appear similar (or even identical) to others,
so long as they are not literally the same physical work submitted by
two separate people. So, the copyright office would have to define new
procedures for handling these issues, including that of dispute resolution.
Next, there is the matter of "derivative works," which is when a given
work is derived from other copyrighted works, often called "mash-ups",
"re-mixes" or "enhanced modifications." The new works are themselves
considered unique and can be copyrighted separately, provided that consent
was given by the original copyright holder(s). The fact that the new work
may or may not have to be properly acquired from the original artists is
a concern, but it has traditionally not been a matter under the purview
of the copyright office. It is (and always has been) strictly a legal
matter handled by the courts, which would only come up if someone filed
a claim against another. Yet, if policy is going to address duplicate
submissions and other anomalies that trigger errors in the submission
process, it needs to treat them all consistently, which raises the
question about the matter of derivative works here.
I mention these because people will no doubt use them as insurmountable
problems, but the fact is, any and all problems that people might cite
in objection to the Copyright Infrastructure idea, already exist in
today's system. The copyright registration process has traditionally
been used solely as a vehicle to register works. If policy is going to
change to engage the copyright office ahead of the courts, that could
be a large undertaking.
Long ago, the government recognized that, in order to stimulate participation
in online commerce, there should be sales tax exemptions from transactions
conducted online (except for within-state sales). Many consider this as having
a stimulative effect on the adoption and growth of the internet itself by
the consumer public. There are similar tax-related issues here that would
also affect adoption and growth of copyright participation and compliance.
Since so many people would be engaged in potentially monetizing their
works, they would potentially have income that would otherwise be taxable.
However, if the government were to tax them, then people would then be
eligible to claim all sorts of things as tax deductions for creating
those works. The guy who took a vacation to Hawaii and ended up getting a
check (or credit) for $10 from his Copyright Registrar would then try to
claim the cost of his vacation as a "business expense." The government
would likely lose considerable tax revenue because people's expenses
would clearly exceed the income derived from it.
Having an income tax moratorium on commissions derived from Copyright
Registrars (as opposed to self-representation or by the use of
other licensing agents) for a period of time, provided that the income is
below a certain minimum (and does not include proceeds from infringement
claims) would help incentivize people to participate in the system.
Modification of Copyright Law
Though the intent of this proposal is to be merely an infrastructure
that sits on top ofand is agnostic ofcopyright law, its existence
requires that certain aspects of the law be re-examined. For example,
the requirement for having a certificate for a registration to be
"complete" seems obsolete In fact, copyright law currently states that
having a certificate is necessary in order to file an infringement claim,
although courts have ruled differently on this. (One claiming that the
date of application is sufficient; another threw out a claim because
the copyright holder had not yet received his certificate.)
Consider the eye-fi memory card that can be placed in digital
cameras. It has wi-fi built into it, so the the moment a photo is taken,
it's instantly uploaded to the net through whatever wi-fi network it can
find. The image can be directed to on any number of photo-sharing sites,
or a personal site, a stock agency, or even a news service. The need
for instant copyright registration and certification is already past due.
Because copyright is at the heart of so many very powerful industries
and business interests, most of those companies have already spent lots
of money to build infrastructures to support their "current" business
models. And many of them consider these to be their competitive edge
in their particular industry. Therefore, there are those who would not
like to see many changes in copyright.
These are valid concerns, which is, in part, why my proposal requires
no change whatsoever to copyright law, nor does it compel anyone to do
anything they don't already do. It is merely an infrastructure designed
to address pitfalls in how copyright operates within a modern economy.
It provides means and mechanisms by which the free market can build
new businesses that correct for past economic inefficiencies, while
providing more incentives for people and companies to be more compliant
Still, there are those who will still oppose the Copyright Infrastructure,
for a variety of other political or economic motivations. I've heard
many of the dissenting arguments, so I'd like to address them here.
Some have argued that there are already those who have a business interest
in helping consumers, both with information and assistance in copyright
registration and protection. There are law firms and small companies
that offer these services, but what they do is all over the map. Most
not only do not perform copyright registration; they lure users into
thinking they have protection, when, in fact, they don't. In order to
meet a price consumers are willing to pay, they don't register works,
but instead, provide a "timestamp" for when works are sent to them. Their
"pitch" is that it keeps people from infringing because the timestamp
that the company received the submission predates the infringer's use.
This provides absolutely no protection at all; in fact, it's meaningless
in the eyes of copyright law.
Some copyright registration services do properly register works,
but of course, what they do is procedurally no different what
someone can do for himself on the copyright office's website at
http://www.copyright.gov/eco. In the best case, it's sort of like
tipping a bag handler at a hotel to take your luggage from the taxi to
the front door ten feet away: it's nice to have, but it doesn't solve
the real problem. With filing fees at $35 per registration (at the
minimum), that's not something that works for billions of works created
But, by and large, the loudest voices in opposition to the Copyright
Infrastructure are those who oppose many aspects to copyright itself.
They feel that copyright interferes with artistic freedoms and free
speech, or that copyright can be exploited for unfair monetary gain,
and of course, it may compete with their own economic objectives. My
goal here is to illustrate how copyright actually helps each achieve
their objectives, not harms them.
The first group to address is the one feels that copyright is antithetical
to the free flow of ideas: that it obstructs creativity because creative
ideas cannot be owned and controlled. Creativity is often built from
prior works and ideas, which "inspire" new works. If someone can own and
assert control over their ideas, so the argument goes, then others cannot
build upon them.
The misunderstanding here is based on confusing copyright itself with
"copyright law," which are two separate things. Copyright is
merely a source of identification. Think of copyright as an official
passport that associates a work to its origin. Its "official" status
is from the fact that it's federally recognized, which itself implies
a higher degree of credibility than if were identified by an arbitrary
online registry from an unregulated entity. Obstruction does not come
from the fact that a work has an origin, it comes from the goals of the
It is copyright law where some feel that it harms creativity and
freedom of speech. Some want to adjust the law to permit all sorts
of adaptations of prior works to encourage the creation of more works.
Irrespective of whether copyright law is adjusted, copyright itself
still exists underneath it. However the law may be revised, one still
needs an accounting system that identifies a work so that it can be
more efficiently and effectively protectedthis, whether it's sold
or freely used.
It should be noted here that the Fair Use provisions of copyright law
already permit most of the kinds of creative expressions that some
feel aren't permitted. The common example they use, which I noted
earlier in this article, is the teenager who mixes and remixes content to
create new works and share them. Despite common misunderstanding (and
disinformation), these actions are largely legal and do not trigger
copyright violations. It is only certain forms of publishing those works
that consent may be required from the copyright holders. Fair Use is a
very broad (though complicated) set of provisions, and courts face new
cases all the time in determining whether infringements take place. It
is impossible to expect to eliminate conflict like this by removing
or altering copyright law without having an extremely disproportionate
effect on global economies.
Copyright is needed within any transaction involving creative works for
the simple reason of assuring identification and authorization. There has
to be appropriate checks and balances that assure accountability, that
people are who they say they are, that they own what they claim they own,
and that their price is what they claim it to befree, or a gazillion
Then there are those who object to statutory damages under the premise
that people use them as a form of easy entrapment into extortion.
Publishers, media companies, search engines and web portals, for starters,
are all subject to infringement claims on a regular basis, the vast
majority of which are frivolous and/or baseless. But therein lies their
concern: by having more registered works would raise the likelihood
of more and higher infringement claims.
This fear is at the heart of many companies' concerns, which is reasonable
on the face. But the Copyright Infrastructure would be far more effective
as abating this problem than aggravating it, especially if there were
policies instituted proscribed above. There's no question there are
"trolls," but it's not always made up of those looking to file claims.
There are also those who troll the internet looking for content to use
for freeto infringe. Filing legitimate infringement claims, even
frequently, is often necessary for those who produce copious amounts of
highly desirable work. This is especially problematic now, since the
majority of online content is unmonitored and rarely pursued.
Real "copyright trolls" use deceptive tactics to entrap people
into bad behaviors, such as using leading or deceptive promotional
language that entices people into using works in a way that lead to
infringement claims. But, one needs to discern between those acts and
that of protecting one's own economic interests.
Another argument copyright detractors use is that you don't need copyright
registration to provide protectionany kind of public registry can
serve that purpose. For example, The Creative Commons has just such
a registry where the "owners" of works can submit their content to be
registered as freely available. I put the word owners in quotes because
there is no assurance that the person who submitted the CC attribution
is, in fact, authorized to do so. Such registries can be exploited by
bad actors on both sides of a transaction. A company that wished to use a
work for free could easily manage to have it show up in the CC registry.
Without secure and trustworthy authentication, who's to know the better?
If the original author of the work were to file an infringement claim
saying he didn't register the work with the CC, how would a judge really
know? Is the author trying to exploit the end-user? Or, is the end-user
trying to get to use the image for free?
The reason why copyright registration is the better mechanism for claiming
ownership is because it's backed by federal law, which ratchets up the
penalties and consequences substantially, which in effect, gives more
credibility to the data within the registry. The "federal" aspect to
copyright also reduces attempts by bad actors because such people go out
of their way to avoid federal crimes. (This is why smart criminals don't
use postal mail, or cross state lines for certain illicit activities.)
As for the Creative Commons, it is a perfectly reasonable and adequate
mechanism for stating "license terms" for works. Once works are registered
with the copyright office (irrespective of the existence of a Copyright
Registrar, which may very well be the Creative Commons), then the CC
registry would be adequate for expressing your desire to have others
use your work freely.
The main objective of my proposal for reorganizing the copyright office
is basically very simple: the copyright registration process needs
to open up to private industry so that market forces can help people
participate in both the economic development and in the protecting of
their creative works. Once that model is spun into motion, new industries
will take off in a directions that will have unbound economic stimulative
effects for years to come.
What secures that economic development is one that behavioral economists
can understand. When you institute policies that reward good behaviors
and punish bad ones, the byproduct is a cultural shift. When consumers
realize they can make money from their photos or other works by doing
exactly the same things they currently do online today, they will be
more participatory in the system, and therefore, more aware, changing
the mentality of how copyrights are perceived.
And this mindset percolates up the food chain to the corporate level:
A guy works at a big company, and his job is to produce brochures,
advertisements and reports. Does he search the web for photos and use
them in work-related documents without seeking permission? How might
his views change if he's also the one who's registered his own stuff with
the copyright office and is on the watch for infringers of his works?
Even if we don't necessarily trust him to turn over a new ethical leaf
for his personal actions at home, there's the simple fact that he's now
aware of the new domain, and that there's a new industry of tracking
and policing how works are used.
It should also be pointed out that this proposal is not to create an
Orwellian-style police state, where the government spies on and prosecutes
people for minor and frivolous infractions. This is instituting an
economic infrastructure that can percolate throughout every business
sector from the consumer to the Fortune 500 companies. It creates both
economic incentives to participate, and disincentives for cheaters.
When the risk of getting sued for copyright infringement is a more
realistic disincentive than what is available today, especially since
the cost of legitimately licensing works is incredibly cheap (even free
in many cases), the goal of why copyright law was conceived in the first
place will have been achieved.
Click to recommend this page: