YouTube describes its Content-ID anti-piracy filter as a state-of-the-art technology, but those who look closely can see that in some cases it creates a huge mess. The system invites swindlers to claim copyright on other people’s videos and make money off them through ads. It automatically assigns thousands of videos to people who don’t hold the copyrights, and its take-down process appears to be hugely biased towards copyright holders.
In recent years Google and YouTube have gone to extremes to protect copyright holders. Perhaps the greatest achievement thus far is their state-of-the-artContent-ID system.
Content-ID allows rightsholders to upload the videos and music they own to a central ‘fingerprint’ database. YouTube will then scan their site for full or partial matches, and if there is a hit the copyright holder can automatically take it down, or decide to put their ads on it.
Although the above sounds like a fair and honest solution, not everything Content-ID does goes to plan. Of course some errors are expected when pioneering a new system, but the problems are more severe than that. Welcome to the world of YouTube swindlers, mass misattribution of copyrights and an unfair bias towards stubborn copyright holders.
One of the problems appears to be that people with bad intentions can claim copyright on videos they have nothing to do with, and even run ads on them. In the YouTube support forums there are hundreds of posts about this phenomenon, also summarized by the PRV blog recently.
Although some swindlers may indeed be around, most of the “misattribution” problems seem to be the result of screwups and technical limitations. A good example is the case of the Dutch game review site Gamer.nl, owned by the publishing platform Sanoma.