Skip to content

A Billion Dollar Test of the DMCA Safe Harbors

Case and Court:

Viacom Int’l Inc. v. YouTube, Inc. (S.D.N.Y.)


On March 13, 2007, Viacom and several other television and film producers filed a copyright infringement lawsuit in the Southern District of New York against Google’s YouTube subsidiary. The complaint apparently came on the heels of protracted license negotiations between the parties and thus early commentators saw the case as merely a negotiating tactic on Viacom’s part.1 The complaint alleged at least $1 billion in damages and alleged that Plaintiffs had identified more than 150,000 unauthorized clips of their copyrighted programming.2 As the case has progressed somewhat slowly, some have speculated that the delays were caused by settlement talks, but these rumors have been rebuffed by Google, whose VP in charge of content partnerships reportedly said, “We’re going all the way to the Supreme Court.”3 If Google’s resolve is unwavering, then the case could be the biggest test of the Digital Millennium Copyright Act’s (“DMCA”) “safe harbors” that limit the copyright liability of internet intermediaries.

In many ways, Viacom’s complaint reads like sour grapes over the decision by Congress to provide internet intermediaries with limited liability when their users commit acts of copyright infringement. The notice and takedown system the DMCA created may indeed be as onerous as the complaint alleges for large copyright holders with numerous popular works to protect. However, it is the system Congress adopted and if YouTube shows that it is operating within the confines of that statutory scheme, then Viacom will have to direct its complaints to the legislative, rather than the judicial branch of government. The key question in this case then should be: does YouTube operate within a DMCA safe harbor?

However, Viacom does make an effort to allege that YouTube is itself engaging in acts of direct infringement by publicly performing Plaintiff’s copyrighted works, by publicly displaying individual thumbnail images taken from such videos in response to searches, and by reproducing and distributing unauthorized copies of Plaintiff’s works (through the routine operation of the site). Perhaps these issues will receive some attention, but it seems to me that each of these alleged acts of YouTube are either fair use when one is providing a transformative work such as a video website with a search engine or are really the result of YouTube’s storage of the content provided by its users, which brings us right back to the question of the applicability of the DMCA safe harbors.

The DMCA Safe Harbors

There are three steps to answering the question of YouTube’s eligibilty for the DMCA safe harbor:

1. Does YouTube meet the threshold requirements of 17 U.S.C. § 512(i)?
2. Is YouTube a “service provider” under one of the four safe harbors in §§ 512(a),(b),(c), or (d)?
3. Does YouTube satisfy all of the conditions of a specific safe harbor?

The Threshold Requirements of § 512(i)

As a threshold matter, a service provider is eligible for a safe harbor only if it has done at least two things:

(A) has adopted and reasonably implemented, and informs subscribers and account holders of the service provider’s system or network of, a policy that provides for the termination in appropriate circumstances of subscribers and account holders of the service provider’s system or network who are repeat infringers; and
(B) accommodates and does not interfere with standard technical measures.

Repeat Infringer Policy

As to the first item, YouTube has long had a “Repeat Infringer” policy.4 The question then would be whether YouTube has “reasonably implemented” its repeat infringer policy. The statute does not define “reasonably implemented” but there is some guidance from the Ninth Circuit in Perfect 10, Inc. v. ccBill LLC, 488 F.3d 1102 (9th Cir. 2007). There the court wrote,

We hold that a service provider “implements” a policy if it has a working notification system, a procedure for dealing with DMCA-compliant notifications, and if it does not actively prevent copyright owners from collecting information needed to issue such notifications… The statute permits service providers to implement a variety of procedures, but an implementation is reasonable if, under “appropriate circumstances,” the service provider terminates users who repeatedly or blatantly infringe copyright.

As to the first two items, a working notification system and a procedure for dealing with DMCA-compliant notifications, YouTube seems likely to be able to pass muster. Viacom’s Amended Complaint does, however, try to set up an issue with respect to preventing copyright owners from collecting information to issue such notifications by detailing the ability of YouTube’s users to share videos only with their “friends,” which in the case where a Viacom investigator is not among a user’s friends and that user shares a video containing unauthorized Viacom content, Viacom would not be in a position to issue a takedown notice for such videos. However, the Ninth Circuit derived this third item from the Aimster case, in which

Far from doing anything to discourage repeat infringers of the plaintiffs’ copyrights, Aimster invited them to do so, showed them how they could do so with ease using its system, and by teaching its users how to encrypt their unlawful distribution of copyrighted materials disabled itself from doing anything to prevent infringement.5

The district court had called Aimster’s repeat infringer policy “an absolute mirage”6 and explained that Aimster had voluntarily implemented encryption technology that made it impossible for them to determine whether their users were engaging in unauthorized transfers of copyrighted works and thus it was impossible for Aimster to actually implement their supposed policy. YouTube’s implementation of its policy stands in stark contrast to Aimster’s and thus it would be surprising if the Court found that Viacom’s inability to root out every unauthorized transfer among friends meant that YouTube had failed to reasonably implement a repeat infringer policy.7 Any site that provides users with the ability to store files and to control access to those files would face the same problem.

Accommodating Standard Technical Measures

As to the second threshold requirement of accomodating standard technical measures, this should be a non-issue, because “standard technical measures” are specifically defined in § 512(i)(2) and the definition includes the requirement that these measures have been “developed pursuant to a broad consensus of copyright owners and service providers in an open, fair, voluntary, multi-industry standards process” and no such process has ever occurred, much less produced consensus.

Does YouTube Qualify as a “Service Provider” Under One of the Safe Harbors?

With the exception of the 512(a) Transitory Communication Safe Harbor, which has its own definition of “service provider” the remaining safe harbors use a broader definition of “a provider of online services or network access, or the operator of facilities therefor.” 17 U.S.C. § 512(k)(1)(B). This definition led the Aimster court to remark that,

“service provider” is defined so broadly that we have trouble imagining the existence of an online service that would not fall under the definitions…8

The most natural safe harbor to accommodate YouTube would be under § 512(c), which limits liability for infringement of copyright “by reason of the storage at the direction of a user of material that resides on a system or network controlled or operated by or for the service provider.” I will refer to this as the “hosting” safe harbor.

Does YouTube satisfy all of the conditions of the § 512(c) “hosting” safe harbor?

The hosting safe harbor has a number of requirements:

Designated Agent

The easiest requirement of the hosting safe harbor is that one is required to have a designated agent, registered with the Copyright Office, which YouTube has.

Expeditious Removal

The statute further requires that, upon appropriate notification of claimed infringement, the service provider “responds expeditiously to remove, or disable access to, the material that is claimed to be infringing or to be the subject of infringing activity.” While the Amended Complaint bemoans the constant game of whack-a-mole that this process subjects Plaintiffs to, I would be astonished if discovery showed that YouTube was not “expeditiously” responding to compliant notices. The facts I do know suggest the contrary, that YouTube has probabaly gone further than any other site in making the notice and takedown process less painful for copyright owners and more effectively removes material that is the subject of a compliant notice. For a site with such inconceivable traffic volume, one would expect outlier cases where a notice falls through the cracks or takes longer than expected, but, on a percentage basis, I would expect YouTube will make a solid showing on this issue.

Actual or “Red Flag” Knowledge

A harder requirement to satisfy requires that the service provider,

(i) does not have actual knowledge that the material or an activity using the material on the system or network is infringing;

(ii) in the absence of such actual knowledge, is not aware of facts or circumstances from which infringing activity is apparent; or

(iii) upon obtaining such knowledge or awareness, acts expeditiously to remove, or disable access to, the material;

Actual Knowledge

Cases interpreting the actual knowledge requirement have typically looked to whether the copyright owner provided notice to the service provider, which the service provider nonetheless ignored.9 Again, it is doubtful that discovery will produce many, if any, instances of a compliant notice being ignored by YouTube.

In a general sense, everyone has actual knowledge that unauthorized copyrighted works are available on YouTube (at least momentarily), because if YouTube is processing proper takedown requests, then so long as at least some of the requests are legitimate, then unauthorized material is on the site. However, this is not the sort of actual knowledge the statute intends, and with respect to any particular video, the issue is decidedly more complex. First, the sheer volume of material uploaded every second to YouTube makes it literally impossible for the humans employed by Google to review it all, so there is a large percentage of videos that YouTube does not know about at all. Second, even if a particular video were to come under the scrutiny of a YouTube employee, they face an impossible question: Is this particular use authorized? One’s authorization to use a copyrighted work is not visible on the face of the work. For all YouTube knows, the uploader works for or is the copyright owner and is legitimately uploading the content. In closer cases, the uploader may have a legitimate fair use argument, a multi-factor evaluation which also requires more factual information than is available on the face of the video.

“Red Flag” Knowledge

The Northern District of California addressed the “red flag” standard as follows,

In determining whether a service provider has such awareness, “the question is not ‘what a reasonable person would have deduced given all the circumstances.'” [citations]. “Instead the question is whether the service provider deliberately proceeded in the face of blatant factors of which it was aware.” [citation]. In other words, “apparent knowledge requires evidence that a service provider ‘turned a blind eye to ‘red flags’ of obvious infringement.'” [citations]. Io Group, Inc. v. Veoh Networks, Inc., 586 F. Supp. 2d 1132, 1148 (N.D. Cal. 2008).

The Ninth Circuit’s ccBill decision also provides guidance on what constitutes “red flag” knowledge of infringing activity. There the Court was extremely reluctant to impute awareness of apparent infringing activity based on the name of a site or its status as an apparent password-hacking site. Such an approach in this case might suggest that even were title of a video “Stolen Copy of SpongeBob Squarepants the Movie” or “Illegal Copy of Mean Girls” or even were the uploader’s username “IllegalUploader” or similar, then these facts alone would not amount to “red flag” knoweldge since the title or username could be a hoax or an attempt to increase the videos’ appeal rather than an admission of unauthorized activity.

In short, “deliberately proceed[ing] in the face of blatant factors” is a high standard and it again would be surprising if discovery produced such evidence.

Direct Financial Benefit and Right and Ability to Control

To maintain its limited liability under the hosting safe harbor, YouTube also will need to show that it

does not receive a financial benefit directly attributable to the infringing activity, in a case in which the service provider has the right and ability to control such activity;

In a case involving a video sharing site, the Court explained that having the right and ability to control one’s system is not the same thing as having the right and ability to control the infringing activity. Io Group, Inc. v. Veoh Networks, Inc., 586 F. Supp. 2d 1132, 1151 (N.D. Cal. 2008). Thus, the fact that YouTube has a Terms of Service that prohibits its users from violating the copyrights of others and that it has the ability to block and remove access to materials should not satisfy the “right and ability to control” prong of § 512(c)(1)(B). Instead the standard “presupposes some antecedent ability to limit or filter copyrighted material.” Tur v. YouTube, Inc., No. 06-4436, 2007 U.S. Dist. LEXIS 50254, 2007 WL 1893635 at *3 (C.D. Cal., June 20, 2007). This is another area where the facts are likely to favor YouTube as its efforts to provide fingerprinting and filtering technologies, which go beyond the requirements of the DMCA, will be hard to square with the portrait of an unabashed infringer that Viacom paints. Thus it seems likely that the Court will find that to the small extent that YouTube has the right and ability to control infringing activity, it is exercising that ability to prevent it, and otherwise does not have such a right and ability.

However, even were the Court to hold that YouTube does have the right and ability to control infringing activity, it would also have to “receive a financial benefit directly attributable to the infringing activity.” Viacom tries to lump YouTube in with Napster and Grokster as companies who built their business around the draw of infringing material. This argument just does not fit the facts of YouTube, which probably has far more videos of housecats than it does of Spongebob Squarepants. More to the point, it is unclear what “direct financial benefit” to YouTube could be attributed to the infringing videos, because, as far as I can tell, ads only appear on pages where the copyright owner has authorized them.


YouTube’s position is not without some risks, as most of the cases, mentioned above, that have interpreted the safe harbors are not binding precedent for the Southern District of New York, thus it is always possible this Court might see one of the above key issues differently. It is also conceivable that discovery in the case will reveal that YouTube’s outward appearance of making serious efforts to implement a first-rate notice and takedown regime will actually turn out to be full of missteps and half-measures–but I wouldn’t bet on it. The case has been going on for over two years now and the discovery process has been extensive and therefore costly. This makes me believe that the parties really do want to see this through to a decision, because if it is just a chess piece in a larger licensing negotiation, it has become a very expensive one.

See also:


  1. Anne Broache, Viacom sues Google over YouTube clips, ZDNet (Mar. 13, 2007). []
  2. The Amended Complaint, filed Apr. 24, 2008 contains a 1,812 page Exhibit listing 17,314 allegedly infringing YouTube videos. However, many of the videos involve the same copyrighted works, so the number of works at issue is no doubt large, but substantially less than 150,000. []
  3. AP, Google says it will not settle copyright issue with Viacom outside court, The Hindu (May 7, 2008). []
  4. YouTube Terms of Service, Sec. 7.A. (Retrieved on Aug. 25, 2009). []
  5. In re Aimster Copyright Litig., 334 F.3d 643, 655 (7th Cir. 2003). []
  6. In re Aimster Copyright Litig., 252 F. Supp. 2d 634, 659 n18 (N.D. Ill. 2002). []
  7. See also Perfect 10, Inc. v Cybernet Ventures, Inc., 213 F. Supp. 2d 1146 (C.D. Cal. 2002) (AVS had not reasonably implemented repeat infringer policy). []
  8. In re Aimster Copyright Litig., 252 F. Supp. 2d 634, 658 (N.D. Ill. 2002). []
  9. Io Group, Inc. v. Veoh Networks, Inc., 586 F. Supp. 2d 1132, 1148 (N.D. Cal. 2008). []

Categories: Copyright, DMCA Safe Harbors.

Comment Feed

No Responses (yet)

You must be logged in to post a comment.