viz.

Recently, one of my YouTube videos was automatically removed for "copyright violations." I decided to take a closer look into YouTube's policies and found they may be dissuading users from exercising their Fair Use rights. 

In partnership with Megan Varelmann and Vincent Robert-Nicoud, I created a video illustrating a few of Lev Manovich’s concepts from The Language of New Media. The scene YouTube found questionable was an illustration of the Language of Selection through a clip of the

IKEA scene from Fight Club:

Our use of this clip was in my mind well within Fair Use, so I filed a dispute. The video was restored within a week, but it got me thinking about some of the silencing effects from YouTube’s automated system and their approach to “educating” consumer-producers (or “prosumers,” to use Alvin Toffler’s  term). The visual rhetoric of both the interface for filing a dispute and the cartoon copyright violators are forced to watch subtly argue for copyright as mainly for owners’ rights, rather than a balance between owners’ and users’ rights.

First, though, a bit of clarification on YouTube’s policies. In 2007, Viacom, Inc. brought a $1 billion lawsuit against YouTube for mass copyright infringement. Viacom argued that YouTube knowingly allowed users to upload copyrighted content, and Viacom wanted its share of the multibillion-dollar pie that is YouTube. In 2010, Google (the owner of YouTube) won the case, but only because the judge ruled YouTube is protected underneath the “safe harbor” provisions of the Digital Millennium Copyright Act. In these provisions, online service providers are protected “from liability for information posted or transmitted by subscribers if they quickly remove or disable access to material identified in a copyright holder's complaint." Thus arose YouTube’s fairly recent measures to remove offending content.

YouTube created their Content Verification Program, where copyright holders could submit material to a database that YouTube uses to automatically screen uploaded videos. A match from the database then means the copyright owner can block, track, or monetize the content. Blocking the content automatically removes the video. Tracking allows the owner to keep an eye on the offending video for future action. Finally, the most controversial option allows the owner to “monetize” the offending video by placing ads around it. All of these options happen automatically (and impressively, from a computing standpoint – see this TED talk for a brief explanation of the system). That is, there is no human intervention in these stages, and there have been numerous complaints about YouTube’s crawlers assuming an offense before the facts are reviewed, especially since three violations equals a blocked channel. Even Lady Gaga ran afoul of this automated system when her channel was temporarily suspended for posting videos of her own copyrighted music performances. Fear not, though, humans: we can still petition our machine overlords.

If a video is flagged, users then have the option to dispute the automatic takedown for one of three reasons: 1. the Content ID program made a mistake (the video contains no copyrighted material); 2. the user doesn’t need the owner’s permission to use the content (the portion used falls under Fair Use); or 3. the user has the owner’s permission to use the copyrighted content. When a user files a dispute, YouTube automatically restores the video, but the user has just made herself legally liable should the copyright owner choose to sue for infringement.

If you’re feeling a bit confused, you aren’t alone. The process can be somewhat daunting for a casual YouTube user, and this is only the first stage in copyright complaints. Further, the dispute form uses some interface choices that serve to silence the user’s input into the process. The takedown notices from YouTube are vague, giving only a short statement that in my case read “Your video, Visualizing Manovich's New Media, may have content that is owned or licensed by FOX. No action is required on your part; however, if you are interested in learning how this affects your video, please visit the Content ID Matches section of your account for more information.” Of course, no action was required on my part had I chosen to not dispute their claim.

But I did want to dispute their claim. The video in question was made for a class, and I stood to gain nothing financially from challenging their claim. However, I wanted to exercise my Fair Use rights and learn more about this process. When I visited the Content ID Matches section to dispute the takedown, I was informed by YouTube that the length of the copyrighted material may not matter. While this is technically correct, the length also may matter in Fair Use disputes. On many other pages about copyright on YouTube, the company takes a hands-off approach, stating they “can't give [me] advice on either of these topics [i.e., Fair Use and Fair Dealing], and if [I] do plan to use even a small portion of copyrighted material in [my] video [YouTube would] strongly advise [me] to take legal advice first." But here, YouTube makes a point to tell users that the length used may not matter, thus pre-empting a common conception about copyright that may actually hold true.

YouTube cautions against frivolous disputes, as such actions could make me legally liable in civil suits (and should the “10 Strikes Bill” circulating in the Senate pass, possible criminal penalties in the future). However, it’s a computer program that chose to flag my video without any context to the situation, and thus no consideration of Fair Use rights. The program works in the owners’ favor by acting first and asking questions later, presuming guilt instead of innocence. Say, for example, that Fox had wanted to monetize my video instead of blocking it. Would Fox still receive revenue from my video between the time it was flagged and the time I acted to dispute the claim?

Even beyond the advantage given to complainants by YouTube’s Content ID system, the notice gave too little information to mount a thorough counterclaim. All YouTube told me was that my video contained some material copyrighted by Fox. My video actually contained clips from four other films, and only by searching the Internet Movie Database did I learn Fight Club was the only one Fox owns. In the larger arena of mashups, parodies, and home videos, a single uploaded video could contain a lot of copyrighted material that falls under Fair Use. What parts did YouTube’s Content ID system match? Was it a clip from a film? The television program running in the background of my humorous cat video? A couple of bars from a song? To effectively argue a position, both parties need information, and the current YouTube system hands more information to the owners while limiting the information given to users.

The dispute process interface further reinforces this imbalance of information. If the user files a dispute based upon Fair Use, YouTube gives the option of one small text field to explain the use:

In my case, I explained how my video fared under each aspect of the four-part test, and the text went well beyond the field given. Thus, it was difficult to edit my text and added to my frustration with the entire process. If the copyright holder rejects the dispute, though, no explanation is needed whatsoever. The user only receives a notice that the claim is rejected.

While YouTube doesn’t deny users their Fair Use rights, as such a practice would be illegal, they certainly frame the debate in such a way to make exercising Fair Use difficult. YouTube’s Content ID system is designed to shield YouTube from liability in copyright cases, and it’s understandable that any company would want to do so. YouTube must sift through the twenty hours of video uploaded each minute. Mistakes in such a massive system are inevitable, and I’m amazed by YouTube’s programmers for even being able to address such a large dataset. However, even when “educating” the public about copyright, YouTube errs on the side of copyright for owners’ rights. Next week, I’ll address YouTube’s Copyright  School, a kind of “traffic school” for those it deems “copyright violators.”