Although Google has recently stepped up its anti-piracy efforts, go to YouTube (that Google owns), and you’ll still see thousands of copyright infringing clips from movies or TV shows like the Simpsons or Seinfeld. Despite being uploaded by individuals with no affiliation with the copyright holders, the media hasn’t been removed. Instead, Member’s of Google’s Content ID Program can use those clips to generate ad revenue, with user’s getting to both see free content or use the content to make their videos.
The Content ID program is a pretty simple concept, although the technology is quite sophisticated. Here is how it works from a user perspective: Members of the Content ID program submit the copyrighted works that they want the Content ID program to track. Each day, Google searches over 400 years worth of videos looking for any content that match the member’s submissions. When one of found, the member is notified. The member can then either ask for the work to be removed or instead, leave the content intact and run ads against it. The Content ID Program currently has 500 members, with the majority having decided that monetizing the infringements is better than banning the content. The program has been very successful and lucrative with over $1 billion dollars paid to program members since 2007.
The Content ID program works through a technology called Digital Fingerprinting. Every piece of media has identifying features that can be spotted by software and turned into a code that is unique to that piece of media. Digimarc, a company that specializes in digital watermarking and fingerprinting, explains the technology this way: Fingerprinting digital media works very much like fingerprinting people, relying on innate characteristics of the subject. Human fingerprints are characterized by patterns of loops or whorls; digital media fingerprints are comprised of clues such as audio waveforms and/or video characteristics. In both cases, a database of fingerprints must be maintained, against which to compare fingerprints “found in the field,” and technology is needed to rapidly match fingerprint “evidence” to fingerprints on file.” The database of fingerprints is generated when Content ID members upload content that they wish to track. Digital fingerprints help not just to identify media but also can help distinguish between versions. Imagine 50 pictures of the Eifel Tower from the same exact angle. A digital fingerprint will look at color, hues, contrast, as well as the people walking around, to create that unique fingerprint.
That’s great for users, who want to create mash-ups, karaoke versions of songs, add new voiceovers to popular movies, and any other types of user-generated content based on copyrighted material. According to YouTube’s Matthew Glotzbach, “Content ID has empowered creators to remix, curate and celebrate their favorite songs and videos resulting in videos that are both entertaining legions of fans and rewarding rights holders with revenue”.
Some Users are Upset About Content ID
That’s not to say that everything is perfect with Content ID. There are still a lot of takedown’s generated by the program. In some cases, Content ID flags users who believe their use of copyrighted material is fair use. The result is that their content is either removed or filled with ads, despite the result being a false positive. Fair use, however, is rarely a bright line. Instead, the court must decide if the use is a fair use. So the copyright holders have the right to remove the alleged infringing material pending the outcome of a lawsuit. While digital fingerprinting may be a method for discovery, legally, the copyright holder may be standing on firm ground when it takes an action.
Certain groups also tend to have a greater incidence of takedowns than others. Gamers, for example, have been particularly upset about takedowns of their gameplay videos by what they say are overly aggressive copyright holders. The problem with gaming and other multimedia type content is that the works include multiple rights holders, such as the video and music. While the gaming company may want videos of the gameplay to remain online, the music rights holder may want the clip removed. Since the licensing agreements usually don’t include user-generated video content, the videos must be removed if requested. Mash-up videos are also particularly vulnerable to this problem.
But overall, the Content ID concept seems to be working as intended. Not all copyrighted material posted to the Internet is a bad thing. Not every instance is a lost sale. U0ploading of copyrighted material, while an infringement, may also hold promotional value. On most sites, the copyright holder has only two choices, leave the content up or have it removed. The Content ID program gives the marketplace another choice by adding a new layer of monetization for copyright owners. The more sites that adopt programs like Content ID, the more free content users will have access to, which is good for the consumer.