«THE DEATH OF FAIR USE IN CYBERSPACE: YOUTUBE AND THE PROBLEM WITH CONTENT ID TAYLOR B. BARTHOLOMEW† ABSTRACT YouTube has grown exponentially over ...»
Given the statistics on the processing power of Content ID,117 it is unlikely that YouTube will revert back to a manual copyright infringement system. It is equally unlikely that YouTube will hire an armada of tens of thousands of intellectual property lawyers to analyze fair use problems.
Nonetheless, it is also unlikely that fair use will be able to be automated under Content ID (or any algorithm for that matter).
The purpose and character of the use can only be partially determined using automation. Allowing Content ID to determine commerciality on the basis of monetization solves one half of the first prong of the analysis. But how does a bot evaluate a work’s transformative value based upon an algorithm? If YouTube wanted to utilize a bright-line test, it could allow Content ID to calculate the number of novel nuances introduced to the infringing content (superimposition of text, voice input, alteration of picture), identify the type of nuance, and make a preliminary determination of transformative use.
The nature of the copyrighted use is equally problematic. A computer program likely cannot determine if a use is for purposes of criticism. In its current form, Content ID cannot tell the difference between wholesale piracy and an in-person interview.118 As Angry Joe so complained: “My Tomb Raider interview with the Tomb Raider people has been claimed [on behalf of] Tomb Raider [by Content ID].”119 YouTube could allow Content ID to look at the title of the video and any supporting descriptions to determine what the uploader has characterized the video as, but this is unwieldy and easily circumvented by pirates uploading copyrighted content in bad faith.
Assessment of the amount and substantiality of the work used can certainly be automated. YouTube can likely introduce basic calculations to Content ID in order for it to determine percentage-based-infringement.
117 See YouTube Help, supra note 6 (“It’s like 36,000 people staring, without blinking, at 36,000 monitors, all day, every day.”).
118 Youtube Copyright Disaster, supra note 1.
88 THE DEATH OF FAIR USE IN CYBERSPACE [Vol. 13 Taking the above discussion of Angry Joe’s The Elder Scrolls V: Skyrim review as an example, Content ID would be able to determine that the review is infringing less than one percent of the original piece. The copyright owner could certify the length of the actual video game as part of their copyright ownership registration to YouTube’s database. Once both sets of data are collected by Content ID, it could cross-reference this with a judicially-created percentage, and determine eligibility of this prong of the analysis accordingly, or it could establish a presumption for its own purposes.
The effect on the value of the copyrighted work is another problematic element that most likely cannot be automated. Content ID would simply have to defer to certifications made by developers and publishers that allow derivative use of their work. Although this is not an independent determination, it is a better starting point than the current model of ignoring clear-cut cases of fair use altogether.
Content ID most likely cannot be automated given the current technology available to YouTube. But since there is relatively little in the way of defense for uploaders like Angry Joe, in stark contrast to the current Content ID model, an inelegant solution may be better than no solution at all. If the doctrine of fair use is to survive in cyberspace, then it needs a
defense of its own on YouTube.
CONCLUSIONIn theory, Content ID is a novel technology: it allows YouTube to simultaneously and efficiently protect a copyright holder against the unknowing theft of their content, while allowing users to continually create new content. In practice, however, it is a poor proxy for a case-by-case analysis of alleged copyright infringement. Content ID is ideal for a situation where, for example, a user uploads a full movie or song, without alteration, to the website: a blatant infringement of copyright. But infringement is not always this simple on YouTube. With Content ID in effect, YouTube is using a hammer where a scalpel is required. By ignoring fair use altogether in its faulty application, Content ID effectively shifts the neutral presumption of the fair use doctrine against the uploader as a content-creator and stifles the creation of any new works. Thus, Content ID ultimately undermines the doctrine of fair use, significantly impinging the encouragement of creativity that is a central tenet of copyright law itself.