Fact #1: In order to notify Google of an infringement, you first need to find the infringement. But Google places artificial limits on the number of queries that can be made by a copyright owner to identify infringements. These limits significantly decrease the utility of Google’s take down tool given the vast nature of the piracy problem today and the number of titles we are trying to protect. The number of queries they allow is miniscule, especially when you consider that Google handles more than 3 billion searches per day. Yet Google has denied requests to remove this barrier to finding the infringements.
Fact #2: You can’t notify Google about the scope of the problem if it limits the notices it will accept and process through its automated tool. And that is what Google does. On top of the query limitation, Google also limits the number of links we can ask them to remove per day. Google has the resources to allow take downs that would more meaningfully address the piracy problem it recognizes, given that it likely indexes hundreds of millions of links per day. Yet this limitation remains despite requests to remove it.
Fact #3: One needs to consider these numbers and Google’s activities in context. Google says it received requests to remove 1.2 million links from 1000 copyright owners in one month. But consider that Google has identified nearly 5 million new links posted in just the last month in searches for free mp3 downloads of just the top 10 Billboard tracks. The constraints Google has placed on the tools they promote to deter infringement are well below what is necessary to identify and notice infringements on the Billboard top 10, much less the entire catalog of the American creative community.
Googl
Fact #4: Google’s “transparency report” calculates the percentage of a site that is infringing – but this data is flawed and of little value on its own. Specifically, Google claims that the DMCA notices it has received for a site represent less than .1% of the links it had indexed for the domains at the top of this list. But this number is misleading given the constraints imposed by Google on a copyright owner’s ability to find infringements and send notices to Google. If these constraints did not exist, how many more links on these sites might be identified? For example, Google calculates that infringing links account for only .1% of links on filestube, a notorious source of infringing links. For anyone who knows filestube, this seems unlikely, especially given that Google’s data doesn’t include DMCA notices sent directly to the site. Moreover, Google’s methodology fails to account for the percentage of traffic to the infringing portion of the site compared to any potential non-infringing portions. Let’s give copyright owners the ability to access all the pages on a site and take down all the infringing links, and then let’s rationally discuss how to categorize the sites.
Fact #5: Google’s data shows why its interpretation of the DMCA makes it ineffective. Let’s take a step back for a moment. Everyone – including Google – knows that the worst sites are repopulated with links to infringing files of the same content as quickly as links are taken down. For example, in a recent one month period, we sent Google, and the site in question, multiple DMCA notices concerning over 300 separate unauthorized copies of the same musical recording owned by one of our member companies. Yet that song is still available on that site today, and we reached it via a search result link indexed by Google. This highlights the futility of the exercise: if “take down” does not mean “keep down,” then Google’s limitations merely perpetuate the fraud wrought on copyright owners by those who game the system under the DMCA.




