PTO/NTIA access to rights information

The Government’s Role in a More Efficient Online Marketplace

Panel #1: Access to Rights Information

Moderator: Garrett Levin, Attorney‐Advisor for Copyright, Office of Policy and International Affairs, USPTO

Building the online marketplace is fundamentally for the private sector, and that process is well underway. Many comments stressed importance of private sector, and we agree, but gov’t may be able to facilitate the process.

Panelists:

Colin Rushing, SoundExchange

Building a database: we’re working on trying to make ISRC system work better, and working with counterparts around the world on better systems and processes for flow of data and money.

Prof. Pamela Samuelson, University of California at Berkeley School of Law

Berkeley hosted a conference on reformalizing copyright; many speakers discussed the importance of accurate and up-to-date rights information. Consensus: need more information through recording transfers. People who record transfers now do so voluntarily but incentives aren’t good enough.  Things one might do to create more incentives mentioned by her panelists (not a Samuelson endorsement): Jane Ginsburg among others mentioned making transfer unenforceable if unrecorded (shades of mortgages!); conditioning statutory damages and attorneys’ fees on recordation of transfer; condition availability of other remedies, including injunction, on recordation of transfer; condition right to sue on recordation of exclusive license or other transfer. Most intriguing, from Jane Ginsburg: unrecorded transfer would accomplish only nonexclusive license rather than exclusive license or assignment. That would be a pretty strong incentive to get the job done!

Matt Schruers, Computer & Communications Industry Association

CCIA is a trade ass’n of internet/tech companies. These panels are the answer to the previous panel. All the sticks in the world driving people away from access to unlawful content won’t be effective without access to lawful content. You can’t litigate your way to prosperity unless you’re a lawyer.  Must focus more on compensation than on control; too often we sacrifice compensation on the altar of control. How do you get to more yes?  Less sexy than polarized fights about notice & takedown or first sale; this is very technical and answers tend to be technical.

SoundExchange has interesting discussion about standards in its comments, as does Green Paper. Need to standardize how we store info, as well as how we access it through APIs.

Jim Griffin, OneHouse

Our job is to make it faster, easier, and simpler to pay, in hopes that more people will. The way to do that is create a market in registry services—to make it profitable. Role of gov’t is to create wholesale registries at the core that incentivize retail activity at the edge.  Should take notice of the best database in the world: DNS system brings millisecond answers worldwide.  Green Paper is positive about registries and databases, except for content.  Registering content is the key, and that will only happen when it’s profitable. Content industry sees registration as a cost, and as a risk. But tech loves databases and makes them profitable. All manner of investment into DNS because it’s profitable. In our industry we see it as a cost though.  Need to make every point of the registration chain profitable, and then advertise to get creative claims registered.  (There is so much space here for an analysis of the rhetoric of privatization and monetization!  Someone should get on that. But only for pay, because otherwise it’s meaningless.)

Jeff Sedlik, PLUS Coalition

Not enough people are involved in image licensing are here. Visual creators are the smallest of the small businesses. They aren’t able to represent themselves effectively; trade ass’ns struggle because it’s a disenfranchised industry. Information is easily stripped out, e.g. by screen capture. Connection between rights holder and image is lost, and you get an instant orphan. Inability to monetize is plaguing visual arts creators. Whole generation is deciding not to create because they can’t support themselves—can’t enforce/control rights because info isn’t there. Publishers are drowning in images, but can’t identify what rights they have and what rights they don’t because of fragility of metadata. Search engines can’t pass on rights info because they don’t have it—leads to overhesitancy of use or infringement. Cultural heritage sector: inefficient in amount of time to search for ownership—leads to trouble with preservation or hesitance to make use even though they should be used.

Lee Knife, Digital Media Association

Ideally people want to license everything at one fell swoop, but that’s not possible under today’s standards. 

Q: how could the gov’t promote the adoption of standards? What kind of standards?

Rushing: What doesn’t exist is a full list of sound recordings and associated ISRCs.  Makes system tricky to use.  Industry is addressing first problem: not a registry. Effort to create a useful/usable registry. Gov’ts role: support adoption of standards.  In registration, ISRC should be part of registration—a field allowing you to connect copyright records seamlessly with record company records. SoundExchange administers the statutory license, by regulation. ISRC isn’t required, but we’ve asked the Copyright Royalty Board to revisit that as an industry accepted standard. Gov’t can help standards become true industry standards.

Schruers: It’s not a burden overall when the gov’t requires standardization—increases the value of the entitlement if built into registration/recordation process.  Note also that mass licensing needs an API-like interface.

Samuelson: feasibility study of distributed registry systems would be justified.  Could standardize data collected, be interoperable; need information to be publicly available—how a DNS registry-type system might work is worth asking. Copyright Office could participate in developing standards for interoperability—would allow different types of creators to have communities they’re serving where they feel more connected to that than to the Copyright Office itself.  The copyright principles project recommended such a study. That can actually be done.

Griffin: goal is hierarchical inputs with nonhierarchical outputs: anyone anywhere on the globe can get a quick response, but the inputs have to be a certain way (like DNS). We’re all in this together: to exploit a musical work requires the album cover, lyrics. So we need a photo registry to help music, along with text. The one thing we know works with superspeed, which we need so we can do filtering (!!!! Nice aside!!!!), is domain name registry, and the investment is pouring into them because there’s a difference between the cost of entry at the wholesale level and what can be gleaned at retail. The only way we see necessary outreach to convince creators to register is that.  If it’s profitable, it will get the job done around the world. But mandatory registration for protection won’t work around the world.

Sedik: it is a global issue—images are available all over the world; you won’t know which one to search, so you need to connect them all. My organization is example of public-private cooperation. Marybeth Peters told us: If industry doesn’t pull together with users and distributors for visual content you won’t do well in the future.  API based way of finding who owns the information, so a machine could do it.  Europe is ahead of the US on some of this—Linked Content coalition on how rights info is communicated for various types of media.

Griffin: the only problem with LCC is the question of embedding the content within the file. Others can then change that info. We need a roughly centralized database so that it can’t be tampered with.

Sedik: agree: the position w/r/t visual works is that there’s presently no other way to communicate rights info other than embedding it. We will transition to identifiers linked to remotely stored information, with public and private metadata when we can.

Schruers: additional benefits to metadata could deal with other problems, e.g. in music industry, where it’s not clear that people collecting for uses of works are authorized to do that. A database could do that. Disputes about digital media services: there are two realities. On the one hand, digital media services are paying out billions. On the other, artists claim they aren’t getting paid. Where is the money going? SoundExchange is fairly transparent, but many other institutional licensors aren’t. Metadata could be a painless solution to that.

Griffin: the day of using artist name, track name, and album name is over. Moving into many countries with different character sets, and there are multiple ways to write the single name “Bee Gees.” Black box of money is divided by market share and arrives as unattributed income. We’d all agree that we want to end black boxes and orphans the right way—not through exceptions (quelle horreur!) but through finding the people and giving them money.  When I ask music services why they don’t report with the ISRC code (more than 95% of the money), the reason is there’s no database of them; we can’t verify them; they’re unreliable. We’ve waited 2 decades for the market to solve this. The gov’t can build a wholesale market around which profit-making activity can occur. Profit-motivated operators would not allow this.

Q: what role can the gov’t play in interoperability? 

Schruers: leading by example: altering recordation/registration processes to match best standards. Having APIs and then going out to industry to ask them to do the same.

Samuelson: Office doesn’t right now have the tech/infrastructure. We can talk, but without resources, it’s not going to happen.

Knife: Without turning our back on Berne, we could do more than lead by example. We could have a requirement that to get extra benefits of registration you have to comport with certain standards. Private entities have market motivation to solve these problems—if we could put in idea that responsible entity like Copyright Office would review the standard we wouldn’t be locking ourselves into something that works in 2013 but not 2017. But we’d still motivate people to register with that dataset.

Samuelson: flexibilities within Berne do exist. Jane Ginsburg & Daniel Gervais wrote papers on the subject. Especially for recordation of transfers, formalities are not a problem under Berne. Don’t start with saying that Berne means we can’t do anything with formalities. We need to do what’s right, not just ignore formalities.

Griffin: Brazil has registration and it’s not in violation of Berne. Enormous depths of problem—we’re not keeping up with the databases now. Societies around the world report lots of people joining and expecting to be paid. Best registry is 8 million songs; 2 unions that take 5% of money have only 800,000 songs in their databases.  Others report databases of 1-1.5 million tracks, which they think is impressive. To get what we need, we need 250 million song database. And photos are much more, but we don’t have a solid global unique ID. If we take the numbers we have now and shoot for them we’ll miss the mark radically. Have to get trillions of works in every part of the industry. That requires public private partnerships—private capital, outreach, and advertising to get the word out to people who aren’t represented: if you aren’t in this database, you aren’t getting paid.

Sedlik: looking at music example for visual arts. All stakeholders agree that databases should be separate from licensing; stakeholders should make user-controlled system, not gov’t, and allow any sort of system to connect to it, and connect to copyright offices around the world. We have participants in 130 countries. Solution must be global.

Rushing: ISRC ID’s sound recording. But who has the right to license it? Turns out to be unbelievably complicated; can be multiple entities depending on the use. Then you start crossing borders and everything gets even more complicated.

Knife: a truly central database is impossible. There are entities who control certain pockets, and control sometimes seems more attractive to them than giving access to the data. If we can’t have it all housed in one spot, of all of those rights, they are all owned by somebody. Eventually you can get someone on the phone who will say they control the rights in that country.  (Say what?)  That information is out there—rights exist and are striated. We need to collect that info and create access even if we don’t centralize that information itself.

Schruers: some constituencies view registration as a cost—but those are the ones who can navigate the system. Distributional costs to complex system: favors incumbents, keeps out smaller entities. Be wary of concern trolling about decomplexifying system, because that allows in smaller competitors who may be able to get a larger share of what they’re entitled to because things become smaller, more transparent, democratizing.

Griffin: entrepreneur sees more people who could pay if you start complexifying the info by allowing each person to register a claim such as the person who played the guitar. What one person/gov’t sees as complexity/problem, another sees an opportunity to lower cost and increase information.  Someone in domain name business wouldn’t complain that there were more domain names to register; to the contrary, domain name numbers are increasing.  And people do speculate in domain names.  (I do not think that analogy works at all.  It is the difference between a right that encompasses an entire work with nothing left over (as ownership of a domain name is presently exclusive) and a right to control a tiny chunk of a big thing (that is, the problem we now face with patents.)) We will be successful in registry when we watch the Superbowl and see “register your involvement in a creative work—you might get paid and you’ll get credit.” Without outreach, it won’t happen.