Viacom v. YouTube Postscript—Copyright Infringement, Social Media and the Blurred Lines of the Digital Millennium Copyright Act’s Safe Harbors
By Fernando A. Bohorquez Jr. and David M. McMillan
Fernando A. Bohorquez Jr., a partner in the New York office of BakerHostetler, handles commercial, bankruptcy and intellectual property litigation, and advises startup and social media companies regarding business law, intellectual property and internet liability issues. David M. McMillan, an associate in the New York office of BakerHostetler, focuses on general civil litigation in the federal courts, with an emphasis on bankruptcy, securities and international law-related issues.
More than a year after the U.S. Court of Appeals for the Second Circuit’s seminal opinion in Viacom v. YouTube, the contours of copyright infringement liability in the social media space remain in flux. Viacom held for the first time that the Digital Millennium Copyright Act’s (DMCA) safe harbor provisions are inapplicable to an online service provider that is willfully blind to facts indicating a high probability of copyright infringement because willful blindness is tantamount to actual knowledge of infringement. The handful of decisions since then have highlighted the gray legal areas that remain, among other things: Who bears the burden of proving actual knowledge or willful blindness, and what facts cross the line from mere general notice of infringement (which carries no corresponding duty or liability for service providers) to actual knowledge or willful blindness (which clearly does)? Focusing on social media, this article gives service providers and copyright holders a primer on the DMCA, the continuing saga of the Viacom litigation and the current state of DMCA play in the courts on actual knowledge and willful blindness. It also highlights the complicated, blurred lines these issues will continue to present to copyright litigants in the foreseeable future.
Congress enacted the DMCA in 1998 to update federal copyright law in light of the new and unprecedented challenges presented by the Internet.1 Title II of the DMCA–the “Online Copyright Infringement Liability Limitation Act”–creates a series of safe harbors that help shield Internet service providers from copyright infringement liability and thereby encourage innovation in the developing social and commercial space. As Congress stated, and various courts have since quoted, “[m]any service providers engage in directing users to sites in response to inquiries by users or they volunteer sites that users may find attractive. Some of these sites might contain infringing material. In short, by limiting the liability of service providers, the DMCA ensures that the efficiency of the Internet will continue to improve and that the variety and quality of services on the Internet will continue to expand.”2
To that end, the DMCA established four safe harbors that significantly limit service providers’ liability for copyright infringement claims stemming from the provider’s “passive” activities–i.e., activities where the service provider has no control over, or interaction with, the infringing user-generated material. Specifically, the DMCA safe harbors cover the following activities: (a) “transitory digital network communications,” (b) “system catching,” (c) “information residing on systems or networks at direction of users,” and (d) “information location tools.”3 Each safe harbor represents a typical function performed by service providers that Congress intended to substantially insulate from liability, so long as the providers complied with certain requirements to help protect the rights of copyright holders. The latter two safe harbors are the most relevant to social media companies and user-generated content (UGC) websites, namely §§ 512(c) and (d). And it would not be an overstatement to say that without these safe harbors, there would likely be no Facebook, Twitter or even social media as we know it.
To qualify under any of the safe harbors, an entity must first meet three threshold requirements: (1) it must be a “service provider”–i.e., a provider of “online services or network access, or the operator of facilities therefor”;4 (2) it must have “adopted and reasonably implemented” a policy to address and, in the appropriate circumstances, terminate repeat infringers;5 and (3) it must accommodate measures that copyright owners use to identify or protect their copyrighted works.6
After establishing the threshold criteria, a service provider such as a social media entity must then satisfy the requirements of a particular safe harbor. The §512(c) safe harbor, for instance, covers infringement claims arising “by reason of the storage at the direction of a user” of protected material on the service provider’s system or network. Section 512(c)(1)(A) protects a service provider who:
In addition to the knowledge-based framework above, §512(c) states that a service provider “[can]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 infringing activity.”10
Section 512(c)(1)(C) also requires social media companies and service providers to promptly remove or disable access to infringing material upon receipt of a written notice from a copyright holder that substantially complies with the requirements of §512(c)(3). The so-called “take down notices” under §512(c)(3) must, among other things, identify the copyrighted work allegedly infringed, provide information reasonably sufficient to permit the service provider to locate the material and contain a good faith statement that the work is protected.11 If a social media company receives a deficient take-down notice, then it has no duty to remove the content unless it possesses “actual” or “red flag” awareness of the infringement. And for purposes of determining actual knowledge under the safe harbor, courts may not consider substantially deficient take-down notices provided by the copyright holder.12 Finally, service providers do not have any affirmative duty to monitor for copyright infringement: the DMCA explicitly states that safe harbor eligibility shall not be conditioned on a “service provider monitoring its service or affirmatively seeking facts indicating infringing activity[.]”13
Viacom v. YouTube
In 2006, Viacom International Inc. (Viacom) sued YouTube and Google in the U.S. District Court for the Southern District of New York for copyright infringement based on roughly 79,000 video clips that appeared on Youtube between 2005 and 2008. The case was consolidated with a separate but related infringement action by the Football Association Premier League Ltd. against YouTube. Arguing that they had cured all specific instances of infringement of which they had either “actual” or “red flag” knowledge, the defendants moved for summary judgment and dismissal claiming DMCA protection. The plaintiffs agreed as to the specific infringements but contended YouTube either knew or should have known there were thousands of other protected videos on its site and that YouTube-sanctioned copyright infringement was widespread and common. The district court granted summary judgment for YouTube, holding that the DMCA protects service providers unless they knew of “specific and identifiable infringements of particular individual items” and that “[m]ere knowledge of prevalence of [infringing] activity in general” is insufficient. (Viacom I)14 The court found that the plaintiffs failed to create an issue of fact as to any item-specific knowledge on the defendants’ part.15
On appeal, the Second Circuit affirmed the trial court’s holding as to the DMCA’s requirement that service providers know of “specific and identifiable” instances of infringement but remanded because the trial court failed to examine the factual record in sufficient detail to satisfy the circuit court as to the absence of fact issues regarding YouTube’s knowledge. (Viacom II)16 The Second Circuit held that “the basic operation of §512(c) requires knowledge or awareness of specific infringing activity.”17 The circuit court also explained that the difference between “actual knowledge” in §512(c)(1)(A)(i) and the so-called red flags knowledge of §512(c)(1)(A)(ii) is that of a subjective versus objective standard.18 Whereas the actual knowledge standard turns on what the defendant subjectively knew, the red flag provision hinges on whether the provider was subjectively aware of facts that would have made the specific infringement obvious to a reasonable person.19 The record contained several e-mails among YouTube executives requesting the identification and removal of infringing clips, highlighting legal tensions over “blatantly” illegal clips and discussing specific copyrighted clips in the context of whether to remove the infringing material.20 These internal e-mails convinced the Second Circuit that Viacom “may have raised a material issue of fact regarding YouTube’s knowledge or awareness of specific instances of infringement.”21
Importantly, the circuit held as a matter of first impression that willful blindness may be applied to establish service providers’ actual knowledge under the DMCA. Willful blindness, the circuit court explained, is a deeply-rooted common law principle tantamount to knowledge. “A person is ‘willfully blind’ or engages in ‘conscious avoidance’ amounting to knowledge where the person ‘was aware of a high probability of the fact in dispute and consciously avoided confirming that fact.’ ”22 As the DMCA does not mention willful blindness, the Second Circuit had to apply the statutory construction maxim that a statute abrogates a common law principle only where the statute “speaks directly” to the issue at hand. The only relevant provision of the DMCA to this inquiry is §512(m), which absolves service providers of any duty to “monitor[ ] [their] service or affirmatively seek facts indicating infringing activity.” That said, although §512(m) eliminated a broad common law duty to monitor, “willful blindness cannot be defined as an affirmative duty to monitor.”23 Therefore, because the DMCA did not “speak directly” to the willful blindness doctrine, “[S]ection 512(m) limits–but does not abrogate–the doctrine.” Willful blindness therefore may be applied, the Second Circuit concluded in “appropriate circumstances,” to demonstrate knowledge.24
The Second Circuit also reversed the district court’s holding that §512(c)(1)(B)–which prohibits financial benefit from infringing activity that the service provider controls–has a specific knowledge requirement.25 The circuit court held that the “control and benefit” safe harbor does not require specific knowledge but requires “something more than the ability to remove or block access” to protected material.26 The circuit court remanded the issue to the district court to determine whether a reasonable jury could conclude that YouTube had the right and ability to control the infringing activity and directly received a financial benefit.27
Actual Knowledge and Willful Blindness
Willful blindness is known by many names–“conscious avoidance,” “the ostrich defense,” “deliberate ignorance” or, for fans of 1960s sitcoms, the Sergeant Schutlz “I-hear-nothing, I-see-nothing, I-know-nothing” defense. The doctrine is based on the theory that those in a position to know the truth but deliberately avoid it are just as culpable as those who in fact actually know. In all its permutations, willful blindness has two essential components: “(1) the defendant must subjectively believe there is a high probability that a fact exists and (2) the defendant must take deliberate actions to avoid learning of that fact.”28 Some courts hold that the first component (the actor’s subjective belief) has an objective sub-component, centering on what the actor “must have known,” given his or her sophistication, expertise or experience.29
Although willful blindness’s application to the DMCA may have been one of first impression in Viacom II, its use as a proxy for actual knowledge is nothing new.30 Courts regularly invoke the doctrine across a broad range of legal areas from securities,31 to intellectual property,32 to most commonly criminal law.33 The doctrine was also recently applied to bankruptcy avoidance actions.34 And in the online trademark infringement context, the Second Circuit has made clear that the willful blindness doctrine applies to a defendant who has “reason to suspect that users of its service” are infringing a trademark, but “shield[s] itself from learning of the particular transactions by looking the other way.”35
Courts across all contexts also recognize that questions of knowledge and intent are fact-specific and therefore generally inappropriate for summary judgment.36 The same is true for willful blindness as a stand-in for actual knowledge.37
There are also certain common themes that cut across the willful blindness cases. For example, there is usually some showing that the willfully blind defendant had a motive to turn a blind eye to what he or she strongly suspected was improper.38 Usually that motive is a financial one, such as when a company’s officers suspect–but intentionally ignore–fraudulent transactions by a junior employee that generates profits for the firm.39 Other times, a willfully blind defendant may be motived by safety concerns,40 a desire to maintain goodwill,41 or simply to maintain plausible deniability.42 Another thread is the defendant’s sophistication or expertise: because willful blindness turns in part on the defendant’s subjective awareness (i.e., on what he “must have known”), the defendant’s ability to comprehend the facts before him is crucial.43 This can come into play where the defendant possesses a unique skill-set or industry-specific knowledge.44
Post Viacom Year One
This past spring saw the first lineup of post-Viacom II DMCA cases in the Southern District of New York and the Ninth Circuit. On remand in Viacom itself, the trial court again held that YouTube was protected by the safe harbor provisions of the DMCA and rejected Viacom’s renewed summary judgment motion. (Viacom III)45 Specifically, the court found that Viacom failed to show that YouTube had actual knowledge of specific infringing content concerning the clips-in-suit, or that YouTube was willfully blind to any facts warranting further inquiry as to specific instances of infringement.46
Meanwhile on the West Coast, a pair of appellate decisions from the Ninth Circuit–Columbia Pictures v. Fung47 and UMG Recordings, Inc. v. Shelter Capital Partners,48–canvassed opposite ends of the safe harbor spectrum. The Ninth Circuit had no problem finding that Fung–a distributor whose websites induced third parties to download infringing copies through peer-to-peer file sharing protocols–was ineligible for DMCA protection as he was clearly aware of red flags of infringement under §512(c)(1)(A)(ii).49 Conversely, the Shelter Capital court, in adopting Viacom II, had little difficulty holding that the video sharing website Veoh–which employed a robust hash filtering software and repeat infringer policy–fell within the DMCA’s safe harbor as it lacked knowledge–red flag or otherwise–of specific instances of infringement that it failed to address.50 And finally, there was the post-Viacom do-over Capitol Records Inc. v. MP3tunes LLC, where the trial court reconsidered and vacated its pre-Viacom grant of summary judgment for mp3 locker storage service MP3tunes, holding that Viacom II required a fact-intensive inquiry into issues of actual knowledge and willful blindness that were inappropriate for summary judgment.51
Burden of Proof
The range of holdings in these cases highlights the fault lines that remain after the Second Circuit’s decision in Viacom II. At the outset, the decisions leave open the threshold question of who bears the burden of proving “actual knowledge” or “willful blindness.” In Viacom III, the district court directed the parties to submit evidence concerning whether YouTube had knowledge or awareness of specific infringement.52 YouTube submitted a list of more than 60,000 clips-in-suit and challenged Viacom to identify what adequate notice had been provided to YouTube of the specific infringements.53 Viacom could not.54 (Notably, none of the e-mails relied on by the Second Circuit to remand the case dealt with the clips-in-suit before the trial court).
Viacom insisted, however, that its lack of evidence of specific infringement was of no moment because it was YouTube’s burden to establish its affirmative defense of lack of actual knowledge.55 The court dismissed this burden-shifting argument as “ingenious but … anachronistic[.]”56 Relying on the DMCA’s legislative history and purpose, as well as the statutory scheme placing the burden of the take down notice squarely on the copyright holder, the court held that the “burden of showing that YouTube knew or was aware of the specific infringements in suit cannot be shifted to YouTube to disprove. Congress has determined that the burden of identifying what must be taken down is to be on the copyright owner, a determination which has proven practicable in practice.”57
Although the trial court’s holding appears consistent with the DMCA’s history and purpose, service providers should at most find respite as opposed to finality in the decision. Several courts before Viacom II held that the DMCA’s safe harbor provisions are in fact affirmative defenses and that, as is historically true with affirmative defenses, the burden is on the defendant–not the plaintiff–to prove.58 Most recently in Fung, the Ninth Circuit stated that “[b]ecause the DMCA safe harbors are affirmative defenses, Fung has the burden of establishing that he meets the statutory requirements.”59
There is little question that whichever way this scale tips, it will be a game changer. Burden of proof is always critical, but it carries especially significant weight in the DMCA context due to the sheer volume of copyright infringement material on social media sites. It is therefore no surprise that the Viacom plaintiffs have placed the burden-shifting argument front and center in their “Groundhog Day” appeal to the Second Circuit of the trial court’s latest summary judgment dismissal.60
Line Between Duty to Monitor and Willful Blindness
After Viacom II, another pervading issue is the “tension between the doctrine of willful blindness and the DMCA’s explicit repudiation of any affirmative duty on the part of service providers to monitor user content.”61 As discussed above, the Second Circuit found that Section 512(m)–which releases service providers from any duty to monitor infringing activity–limited, but did not abrogate, willful blindness’s application to the DMCA. Some courts since Viacom II have reconciled the perceived tension by latching on to Viacom II‘s knowledge of “specific and identifiable instances of infringement” language. The district court in Viacom III, for example, explained that “under appropriate circumstances, the imputed knowledge of the willfully avoided fact may impose a duty to make further inquiries that a reasonable person would make,” 62 and in the context of the DMCA, that means “blindness to ‘specific and identifiable instances of infringement.’ ”63
And according to the Viacom III court, general knowledge that infringing works may exist–without knowledge of their specific locations–is insufficient to cross the line into willful blindness and trigger a duty to investigate.64 Because Section 512(m) relieves YouTube of any duty to affirmatively search for and locate the material, an internal YouTube memo discussing the existence of infringing clips generally, but with no specifics, was insufficient evidence of YouTube’s willful blindness.65In other words, even knowledge of infringing clips fell short of triggering a duty to inquire, as it was the copyright owner’s obligation to point YouTube to where the infringing clips were located.66 The Shelter Capital court dismissed a similar argument that Veoh “should have taken the initiative to use search and indexing tools to locate and remove from its website any other content by the artists identified in the [takedown] notices” provided by UMG to Veoh.67 The Ninth Circuit explained that “the DMCA recognizes that service providers who do not locate and remove infringing material they do not specifically know of should not suffer the loss of safe harbor protections.”68
The court in Capitol Records took a subtle but decidedly different approach, explaining that Viacom II only “hints in a footnote that knowledge of general infringing activity would likely not constitute willful blindness.”69 In that case, the court focused on the Second Circuit’s admonition eschewing bright line rules in favor of “explicit fact-finding” on the issue.70 When viewed through that lens, e-mails from users pointing to specific infringing materials–and not necessarily the identifiable location of such materials–were sufficient to create an issue of fact for a jury to decide whether they “impos[ed] a duty to make further inquiries into ‘specific and identifiable’ instances of possible infringement.”71 One e-mail stated that “everything I post is in clear violation of the DMCA” and asked the site to remove “any MP3s that [were] linked to that site,” whereas another e-mail stated “if you search for ‘the clash I fought the law’ …you will get 5 results ….2 of which point to the website www.officerjellynutz.com.”72
The takeaway here is that the tension identified by the Capitol Records court between §512(m) and the doctrine of willful blindness is one that will have to be played out in the record. As noted above, whether a defendant knows of the high probability of a fact and deliberately avoids confirming it is a fact intensive inquiry generally inappropriate for resolution on a motion to dismiss or even at summary judgment–Viacom I and III notwithstanding. Practically speaking, unless the facts are plainly egregious, service providers and copyright holders alike should brace themselves for high litigation costs through discovery and up to, at a minimum, summary judgment motion practice.
But both social media companies and copyright holders can glean some important and helpful guideposts from the willful blindness doctrine’s traditional application. For one thing, motive is crucial to the willful blindness analysis. Examining a service provider’s motive is consistent with Congress’s aim of distinguishing legitimate businesses that host user content from pirate sites that trade of off protected material.73 And whether courts state it explicitly or not, at least some seem to be weighing motive in the DMCA knowledge analysis. In Fung, where the Ninth Circuit had no problem granting summary judgment for the plaintiff, the court’s skepticism of the DMCA defense seemed driven by disdain for the defendant’s business model, which centered on the display and distribution of infringing material on its website.74 And in Shelter Capital, where the court granted summary judgment for the defendant service provider, the Ninth Circuit went to great lengths to highlight the steps Veoh had taken to filter copyright infringing material and to point out that Veoh promptly removed all specific infringing material of which it knew.75 Such conduct does not suggest a defendant with a bad faith motive.
Further, given the emphasis courts often place on a defendant’s sophistication or expertise in evaluating willful blindness, these factors may also surface in the DMCA context. Whether the service provider is a budding startup with little industry expertise or a social media goliath with an extensive and deep bench of industry knowledge, these facts may likely weigh in a determination of whether the social media company had actual knowledge or was willfully blind. Parties should be wise to capitalize on these factors to tell the most compelling story for their side.
The Second Circuit’s decision in Viacom II makes clear that service providers cannot willfully turn a blind eye to indications of specific copyright infringements and still claim DMCA protection. But as the post-Viacom II decisions show, it is unclear which side bears the burden of making that case, and it is even further from clear exactly what circumstances will obligate a service provider to identify and remove protected content outside of a proper take-down notice given the DMCA’s explicit repudiation of any duty to affirmatively monitor for infringement. At bottom, the question is one of fact, and in crafting the competing narratives about a social media company’s culpable knowledge, it may be helpful to incorporate common themes from willful blindness case law such as motive and sophistication.
4 17 U.S.C. §512(k)(1)(B). Courts have routinely interpreted the term “service provider” broadly to include just about anyone under the Internet umbrella and especially social media companies. See, e.g.,UMG Recordings Inc. v. Shelter Capital Partners LLC, 718 F.3d 1006, 2013 BL 67727, 106 U.S.P.Q.2d 1253 (9th Cir. 2013) .
28 See, e.g., Global-Tech Appliances Inc. v. SEB S.A., 131 S. Ct. 2060, 2070, 2011 BL 142067, 98 U.S.P.Q.2d 1665 (2011) ; see also Viacom II, 676 F.3d at 35 (“A person is ‘willfully blind’ or engages in ‘conscious avoidance’ amounting to knowledge where the person ‘was aware of a high probability of the fact in dispute and consciously avoided confirming that fact.’ ”) (quoting Aina-Marshall, 336 F.3d at 170).
34 Picard v. Katz, 462 B.R. 447, 454, 2011 BL 247245 (Bankr. S.D.N.Y. 2011) (departing from the traditional objective “inquiry notice” analysis of a defendants’ “good faith” defense and endorsing a subjective willful blindness test).
36 See, e.g., Thomas H. Lee Equity Fund V, L.P. v. Grant Thorton LLP (in re Refco Inc. Secs. Litig.), No. 07-8663 (S.D.N.Y. Mar. 28, 2011). (“Determining scienter is ordinarily a fact-intensive question and is thus an unlikely candidate for summary judgment.”).
37 See Chanel, Inc. v. Italian Activewear of Florida, Inc., 931 F.2d 1472, 1476 (11th Cir. 1991) (“[w]hether a defendant has been willfully blind will depend on the circumstances and, like intent itself, will generally be a question of fact for the factfinder after trial.”).
38 See In re Fishbach Sec. Litig., No. 89-5826 (S.D.N.Y. Jan. 15, 1992) (“[I]n most cases of ‘reckless disregard of the truth,’ defendants have had a motive for deliberately remaining ignorant of the facts in question[.]”).
40 See United States v. Heredia, 483 F.3d 913, 920, 2007 BL 12584 (9th Cir. 2007) (drug possession defendant claimed she failed to pull over and investigate suspicious odor in car due to highway safety concerns).
41 See In re Leslie Fay Cos. Inc. Sec. Litig., 835 F. Supp. 167, 174 (S.D.N.Y. 1993) (defendant’s “long and profitable history” of providing accounting services cited as possible motive for defendant turning a blind eye to the company’s fraudulent accounting).
43 See, e.g., SEC v. Credit Bancorp, Ltd., 195 F. Supp. 2d 475, 494 (S.D.N.Y. 2002) (“Those who hold themselves out as professionals with specialized knowledge and skill to furnish guidance cannot be heard to claim youth or inexperience[.]”) (quoting SEC v. Hasho, 784 F. Supp. 1059, 1108 (S.D.N.Y.1992)).
44 See, e.g., SEC v. Elliot, No. 09-7594 (S.D.N.Y. Aug. 11, 2011) (defendant’s “knowledge of the federal securities laws from his years of experience as a broker and investment advisor” made it all the more unlikely that lacked scienter).
46 Id.The court also found that Viacom had failed to raise a triable issue of fact as to whether YouTube lost protection under the “control and benefit” safe harbor of section 512(c)(1)(B). Viacom III, 107 U.S.P.Q.2d at 1166. To constitute “something more” than just ordinary control over the website, the provider must “influence or participate in the infringement” to lose the benefit of the Section 512(c)(1)(B) safe harbor. Id. at 1164. YouTube’s decision to restrict its monitoring efforts and use of technology to locate and remove certain infringing clips did not meet that standard. Id. at 1165.
58 See, e.g., Wolk v. Kodak, 840 F. Supp. 2d 774, 746, 2012 BL 3899 (2012) (“to meet the eligibility requirements of 17 U.S.C. §512(c) ‘safe harbor’ provision, [plaintiff] Photobucket must show … that it did not have actual knowledge of the material or an activity using the material on the system or network is infringing and was not aware of facts or circumstances from which infringing activity is apparent.”).
59 Fung, 710 F.3d at 1039; but see Shelter Capital, 718 F.3d at 1023 (affirming that under the “red flags” safe harbor provision “the burden remains with the copyright holder rather than the service provider”).
62 Viacom III, 107 U.S.P.Q.2d at 1162. Several commentators have argued that the court’s articulation of the “willful blindness” standard conflates it with the red flags statutory standard further confusing an already complicated jurisprudence.
67 Shelter Capital, 718 F.3d at 1024. Similarly, in the context of the court’s “something more” analysis under the “benefit and control” provision of §512(c)(1)(B), the trial court in Viacom III found that YouTube did not have an obligation to use filtering or other technology to locate infringing content and could restrict its monitoring and takedown efforts as it saw fit for business reasons. Viacom III, 107 U.S.P.Q.2d at 1165.
68 Id. Notably, notwithstanding Shelter Capital’s conclusion that the plaintiff lacked any evidence that Veoh had “buried its hand in the sand” to specific infringement, the court did comment that the plaintiff “came closer to meeting the §512(c)(1)(A) requirements with its evidence of emails sent to Veoh executives and investors by copyright holders and users identifying infringing content.” Shelter Capital, 718 F.3d at 1024. However, the court rejected those e-mails because (a) as they came from a copyright holder–as opposed to a user–they were subject to the notification requirements of §512(c)(3) and deficient notices cannot be considered under the §512(c)(1)(A) analysis; and (b) even if deficient notices could be considered, the record suggested that Veoh promptly removed the offending content. Id. at 1040.