Friday, June 25, 2010

The Real Issues in Viacom v. YouTube


The district court opinion in Viacom v. YouTube gives summary judgment for YouTube. It's a big win for YouTube's parent company Google.

On its face, the issue in the case is the construction of the "safe harbor" requirements of section 512(c) of the Digital Millennium Copyright Act. In Web 2.0 applications that allow end users to upload content, there is always the chance that some of this content will violate somebody else's copyright. Under the DMCA, online service providers like YouTube are not liable for copyright infringement (including contributory or vicarious infringement) if they take down clips alleged to be infringing when they are notified by the copyright holder.

The online service provider must notify, if possible, the person who originally uploaded the clip, and if that person wants to restore the clip they must give counter-notice: their address and their consent to receive service of process for litigation to settle the copyright issue. The basic idea of the notice-and-counter-notice scheme is to get the two parties together to settle the matter; the intermediary is held harmless if it facilitates this process. In practice, however, counter-notice is relatively infrequent: most end users don't want to identify themselves or risk a lawsuit. Therefore the larger practical effect of 512 concerns the situations under which clips alleged to be infringing will be removed.

Viacom argued that a very large number of copyrighted Viacom clips were regularly posted to YouTube. Viacom identified these clips for YouTube to take them down, and YouTube complied. However, Viacom argued, YouTube knew that people kept uploading infringing clips of various Viacom shows even before it received specific requests to take them down. Therefore YouTube had knowledge of infringing activity (even if not specifically identified clip by clip) and should have taken proactive steps to remove infringing clips before Viacom gave it a takedown notice. (Indeed, because YouTube knew of infringing activity, a take-down notice was superfluous). Moreover, because YouTube makes money from advertising, and because infringing clips attract viewers, YouTube financially benefits from the infringement, it cannot take advantage of the 512 safe harbor in the first place.

The district court held that section 512(c) requires specific identification of each clip and where to find it. To make use of the notice-and-takedown procedure Viacom must identify each clip claimed to be infringing if it wants YouTube to take them down. It is not enough that YouTube knows that there are some infringing clips from Viacom on its servers. YouTube enjoys the safe harbor unless it has actual or constructive knowledge of specific and identifiable individual items. The fact that YouTube receives advertising revenue from attracting viewers to the clips uploaded to the site does not change matters. This holding preserves the status quo in the industry.

The court pointed out that "over 24 hours of new video-viewing time is uploaded to the YouTube website every minute" and that section 512 was designed to relieve the online service provider of a duty to look for clips that might be infringing. Of course, this cuts both ways; because so much information is being uploaded, the burden of identification falls on Viacom. The argument is that Viacom is better able to identify its content than YouTube is (since YouTube would have to identify content from every possible content provider that complained about infringement, not just one).

Although the decision seems to be about copyright and the DMCA's standard of knowledge for an online service provider, this real issues concern innovation policy: who creates technology and who pays for it.

Viacom can digitally watermark its video content (and does so) so that Google can filter any Viacom clips uploaded to YouTube, identify them as Viacom files and block them. In other words, instead of using individual notice and takedown, the entire system can be automated using watermarking plus filtering.

Therefore the real issue in the case is whether Google must design and install such a filtering system and who will pay for it. If Viacom wins the litigation, Google must design and install such a system for every content provider who requests it at its own cost; and even if content providers do not digitally watermark their clips, Google must find some way of identifying them. (Thus, Google might have to pay content providers to watermark their clips so that they will be easier to filter using Google's system).

On the other hand, if Viacom loses this litigation, Google may still filter video clips for copyright infringement. Google can strike deals with content companies in exchange for implementing customized filtering for their content. Google's "Claim Your Content" program is an example of what such a customized filtering system will look like. Customized filtering systems become an additional source of revenue for Google.

Whether Google installs filtering as a legal requirement, or whether it does so for a price, the effect will be to make the counter-notice provisions of the DMCA-- designed to protect the fair use rights of end users-- increasingly irrelevant. After all, if Google filters content from content companies as soon as it is uploaded (which Google probably has a right to do under its terms of service policy)-- then there is no obligation to notify the person who uploaded the content. The counter-notice provisions of Section 512 don't apply at all. The fact that the end user might have a fair use claim is simply irrelevant.

Thus, the district court's opinion is a victory for Google, but it is not necessarily a victory for end users, because Google's interests and theirs may differ in important ways.

On the other hand, the district court's argument ultimately does benefit end users, but in a different way. There is a strong argument that the district court's decision is pro-innovation (in ways that ultimately benefit end users). It allows companies to come up with new Web 2.0 platforms for end users without being hit by lawsuits from content companies who insist that shortly after they are launched the companies will have constructive knowledge of infringing materials uploaded by end users. Thus, if we want more innovation in Web services-- which will ultimately benefit end users--we should construe section 512 strictly as the district court does.

Viacom might reply that the decision is not pro-innovation at all: if Google is held harmless, it will have no incentive to innovate to come up with new methods of identifying infringing content. There will be too many platforms that do nothing to stop infringement. Thus, the opinion gives insufficient incentives for the right kind of innovation.

Ultimately, I don't buy this argument. We must distinguish between established online service providers like YouTube (now owned by Google, an enormous and wealthy company) and startups (like YouTube in 2005). YouTube in 2005 could not have survived such a lawsuit by Viacom, and it might not have been able to develop filtering technology simultaneous with the initial creation of its product. For one thing, it might not have gotten content providers to cooperate by watermarking all of their content beforehand. (YouTube might have to pay content providers for this service). In short, if Viacom's proposed standard had been adopted in 2005, YouTube might never have gotten off the ground. On the other hand, once YouTube becomes an important part of digital culture, many different content providers have economic incentives to work with YouTube (now owned by Google) to develop technologies for filtering content. By allowing small startups to get going, the district court's decision does not necessarily prevent the development of innovations valuable to the content industries later on. The only difference is when these innovations occur and who pays for them.

Older Posts
Newer Posts