Roughly 25 years ago, Google started its business as a simple and straightforward search engine.
The startup swiftly captured a dominant market share and branched out into other businesses, including online advertising and video streaming.
Google is a leading player in all of these markets today. This brings in a yearly revenues of hundreds of billions of dollars, an amount that continues to go up at a rapid pace. At the same time, however, complaints from copyright holders have grown as well.
To deal with these complaints, Google’s parent company Alphabet has implemented a wide variety of complimentary anti-piracy measures. An overview of these technologies was shared with the U.S. Copyright Office recently, which revealed some things we weren’t aware of.
The submission (pdf) is part of the Copyright Office’s inquiry into technological copyright protection measures, which could be used as input for a new and improved version of the DMCA law.
According to Google, there is no need for new legislation. The company points out that it’s already taking extensive voluntary measures to curb piracy and promote a healthy Internet.
Punishing Pre-Release Piracy Searches
This arsenal of anti-piracy measures includes the demotion of pirate sites in search results. Google started doing this ten years ago. It basically means that if Google receives a high number of takedown requests for a particular domain name, other results from the same site are downranked as well.
This system has been optimized over the years. For example, Google can now detect if a site switches to a new domain name so that the demotion signals are carried over.
One DMCA takedown feature that’s new to us is that rightsholders can now indicate if the reported content is unreleased or still-in-theater. When that’s the case, Google can take even stricter measures.
“[W]e have added a ‘still-in-theaters/prerelease’ tag for DMCA notices involving this category of content to enhance the Search demotion signal,” Google writes.
We assume that this “prerelease” flag will result in a stricter downranking punishment, but Google doesn’t provide any further details on that.
Another search-related measure that’s new to us involves Google’s advertising business. When the company receives a takedown notice for an infringing URL in its search engine, the response doesn’t stop there.
If the reported site has any Google ads, these will be automatically disabled in conjunction. At the same time, searches, where the reported URL appeared, will be stripped from ads as well.
“When a URL is delisted from Search following a DMCA notice, any Google ads running on that page are automatically disabled. We also will not run ads on Search that link to delisted pages,” Google writes.
It’s interesting to see how a takedown notice targeted at search results automatically affects another Google product. It certainly amplifies the punishment, but it also results in more collateral damage for wrongly identified URLs.
At the same time, we wonder where this integration stops. Could the next step be to block these URLs in the Chrome browser as well?
Fingerprinting & Hash Matching
The overview shared by Google also includes other widely known anti-piracy technologies such as YouTube’s Content ID. This system processes four million claims per day and handles 98% of all copyright issues on the video platform.
Automated recognition tools are not exclusive to YouTube, however. Google Drive also uses hash matching to prevent content that was previously reported as copyright-infringing from being publicly shared.
This hash matching, which also takes place on YouTube, ensures that infringing content stays down, or at least out of public view.
This technology isn’t perfect. The hash filter recently flagged text files containing only a 0 or a 1 as copyright infringement. And it appears that macOS ‘.DS_Store’ files are erroneously being flagged as well.
It’s safe to say that Google finds itself in a difficult position. The company has to find a balance between helping rightsholders and keeping its customers happy. Most people don’t mind that infringing content is removed as long as it’s done without collateral damage. And based on recent experience, that’s easier said than done.
The public’s aversion to automated takedown tools was also apparent in the Copyright Office’s consultation. This triggered thousands of responses from the public, with many taking a critical position in respect of upload filters and similar technologies.