Digital Millennium Copyright Act

The safe harbor provisions in § 512(c) of the Digital Millennium Copyright Act (DMCA) provide a mechanism that insulates online service providers from monetary damages for infringing materials posted or stored by their users.  To receive this protection, service providers must designate an agent to receive notice of claims of infringement with the Copyright Office and publicly post the agent’s contact information on the website.  A recent case in the Northern District of California, Oppenheimer v. Allvoices, Inc., examined whether service providers can avail themselves of the § 512(c) safe harbor for infringing acts that precede designation of such an agent.

Allvoices is an online service provider that maintains a community-driven platform for the exchange of ideas as well as graphical, written, and audio content.  Allvoices provides users with financial incentives to upload content to the site, and treats such users as “citizen journalists” and independent contractors.  While it began providing access to contributor content in 2008, Allvoices did not designate its DMCA agent until March 2011.

The plaintiff, David Oppenheimer, is a professional photographer whose photographs were posted on Allvoices’s website by contributors in January 2011.  Oppenheimer learned that his photographs had been posted on the Allvoices website in February 2011, prior to Allvoices’s DMCA agent designation.  Oppenheimer sent a cease and desist letter to Allvoices in August 2011, several months after Allvoices designated its DMCA agent.  While Allvoices eventually removed the photographs, Oppenheimer alleged that Allvoices failed to reply to his cease and desist letter and failed to terminate the accounts of repeat infringers, as required by the DMCA.

 Allvoices argued that it was entitled to the protection of the § 512(c) safe harbor for all alleged infringements, not just infringement occurring after it had designated its DMCA agent.  Allvoices did not cite any authority for this position, but maintained that, because the DMCA does not expressly carve out or preserve liability for pre-designation infringement, Congress had intended for the safe harbor to apply to such infringement.

The court rejected Allvoices’s argument and held that, under the plain language of the DMCA, an online service provider may invoke the § 512(c) safe harbor only if it has registered a DMCA agent with the Copyright Office.  According to the court, designation of an agent is a “predicate, express condition” for application of the safe harbors, so Allvoices could not avail itself of the safe harbors with respect to infringement that occurred prior to designation.  The court cited two previous Northern District of California cases that came to similar conclusions, Louis Vuitton Malletier, S.A. v. Akanoc Solutions, Inc. and Nat’l Photo Group, LLC v. Allvoices, Inc. (note that Allvoices was also a defendant in the latter case).  On the merits, the court held that Oppenheimer sufficiently alleged claims of direct, contributory, and vicarious infringement to overcome Allvoices’s motion to dismiss those claims.

A question remains regarding the period of time during which Allvoices may be liable for infringement of Oppenheimer’s photographs.  Specifically, the court did not address whether Allvoices’s potential liability is limited to the period during which the photographs were posted on Allvoices’s website prior to the date that Allvoices designated its DMCA agent with the Copyright Office.  Regardless, the message is clear: online service providers should designate a DMCA agent with the Copyright Office as early as possible in order to obtain the protection of the applicable DMCA safe harbors.

An aspiring actress moves to California and finds her life threatened. While standard fare for pulp fiction, the case of Garcia v. Google involves a twist on this well-worn plot line that not even the most imaginative Hollywood scriptwriter could invent.

Cindy Lee Garcia answered a casting call for a low-budget amateur movie with the working title Desert Warrior. The film’s writer and producer told her that it would be a “historical Arabian Desert adventure film.” Ms. Garcia received $500 for her performance in the film. It turns out the actress was misled by the producer, Mark Basseley Youssef (aka Nakoula Basseley Nakoula, aka Sam Bacile), a Coptic Christian from Egypt, who was reportedly working in conjunction with an American non-profit, Media for Christ. The filmmakers had no intention of making an adventure film; rather, the end product – titled Innocence of Muslims – is an anti-Islamic account of the Prophet Mohammed that many Muslims find highly offensive and blasphemous.

In July 2012, Mr. Youssef posted a 14-minute trailer of the film to YouTube, which is owned and operated by Google. Ms. Garcia appears for about five seconds in the trailer. The film overdubs her voice with lines she never actually spoke. In September 2012, an Egyptian cleric issued a fatwa against all involved in the film, calling on Muslims to “kill the director, the producer, and the actors and everyone who helped and promoted the film.” Ms. Garcia claims that she began to receive death threats and was forced to take precautionary measures at great expense to protect herself from retribution.

Sending takedown notices under the Digital Millennium Copyright Act, Ms. Garcia demanded that Google remove all copies of the trailer from YouTube. Google declined to do so. In September 2012, Ms. Garcia sued Google, later also naming YouTube, asserting claims for copyright infringement. In October 2012, Ms. Garcia moved for a preliminary injunction, seeking to have Google take down all copies of the movie trailer from YouTube. Continue Reading Google Ordered to Remove All Copies of Anti-Islamic Film From YouTube After Actress With Bit Part Threatened by Outraged Muslims; Decision Puzzles Copyright Attorneys

If you want to use those pictures you found on Twitter, beware. A federal judge in New York recently held that taking photos from Twitter to use for a commercial purpose infringes the photographer’s copyrights. On January 14, 2013, Judge Alison Nathan ruled that Agence France Presse (AFP), which provides subscribers with access to photos though an international wire and databank, and the Washington Post (“the Post”) infringed Daniel Morel’s copyrights to photos he posted on Twitter.

In January 2010, freelance photographer Daniel Morel uploaded to his TwitPic account a number of photos he took in Haiti in the immediate aftermath of the earthquake. An individual named Lisandro Suero took those photos from Morel’s Twitter account, reposted them to his own Twitter account, and tweeted that he had exclusive photos of the earthquake. AFP got the photos from Suero’s Twitter page, attributed the photos to Suero, and began distributing them to users of its wire and databank services. Getty Images (“Getty”) received the photos through AFP’s wire service. The Post received the photos from Getty. Getty and the Post published the photos on their websites, with captions that attributed them to Suero.

When Morel’s exclusive agent found out that AFP, Getty and the Post were using his photos, his agent complained. While at least some efforts were made by AFP, Getty and the Post to address Morel’s agent’s complaint, those efforts in most respects fell far short of what is required under the law.

In March 2010, AFP sought a declaratory judgment that it did not infringe Morel’s copyrights, and Morel counterclaimed for copyright infringement against AFP, Getty and the Post. During the course of the case, Morel moved for summary judgment on his copyright infringement counterclaim. In response, the defendants argued that pursuant to the Twitter Terms of Service (TOS), Morel provided them a license to use the photos by his very act of tweeting the photos.

Judge Nathan disagreed. Judge Nathan found that the Twitter TOS provides that users generally retain their rights to the content they post—with the exception of the license granted to Twitter and its partners. Twitter’s “Guidelines for Third Party Use of Tweets in Broadcast or Other Offline Media” further underscored that, while the Twitter TOS permit users to retweet posts, the Twitter TOS was not intended to let the “world-at-large” remove content from Twitter and commercially distribute it. Rebroadcasting tweets in their entirety is now a news program staple and actively encouraged by Twitter. Twitter’s TOS, however, do not permit media outlets to rip copyrighted material out of tweets and use it for some other purpose. Because AFP and the Post put forward no defense other than their license defense, Judge Nathan granted Morel’s motion for summary judgment and found them both liable for copyright infringement.

Unlike AFP and the Post, Getty argued that it was entitled to the benefit of the safe-harbor provisions of the Digital Millennium Copyright Act (DMCA) that protect service providers from liability for copyright infringement. Judge Nathan held, however, that genuine issues of fact existed as to whether Getty could take advantage of the DMCA safe harbor, noting that companies like Getty that are in the business of selling copyrighted material may not be shielded from copyright liability under the DMCA’s safe harbor. Thus, it remains to be seen whether Getty will also be found liable for copyright infringement.

In one bright spot for AFP and Getty, Judge Nathan granted summary judgment in their favor on the proper method for calculating statutory damages under the Copyright Act, which can result in awards of up to $150,000 per work infringed. Morel claimed that he was entitled to a statutory damage award “in the tens or hundreds of millions of dollars” against AFP and Getty. Morel argued that, because AFP and Getty distributed the photos to many of their subscribers, each downstream infringement by one of their subscribers would entitle him to an additional statutory damages award. Judge Nathan disagreed and held that any award of statutory damages against AFP and Getty could not be multiplied based on the number of infringers with whom they may be jointly and severally liable.

This decision clarifies that Twitter users do not lose ownership rights to their content by posting it to Twitter. Although you may have the right to retweet or publish tweets in their entirety, you don’t have the right to take someone else’s content and use it for commercial gain.

Popular online marketplace CafePress.com suffered a legal setback recently when a U.S. District Court in the Southern District of New York denied CafePress’s motion for summary judgment against claims of trademark infringement. CafePress operates an online “print on demand” service that allows users to upload designs which CafePress then prints on a variety of items. The users receive a share of the money that CafePress makes when it sells items displaying the users’ designs. These items include everything from coffee mugs and beer steins to iPhone cases and flip-flops. In 2009, guitar neck manufacturer Born to Rock Design Incorporated (BTR), which owns a federal registration for the trademark “Born to Rock,” sent a letter to CafePress asking the site to stop selling merchandise displaying the mark. Since 2003, CafePress had produced a number of different items displaying the “Born to Rock” phrase, all based on designs provided by users. These designs included the following:

After CafePress refused to remove user designs incorporating the phrase, BTR filed a complaint for, among other things, trademark infringement. Following discovery, CafePress filed a motion for summary judgment, arguing that the “Born to Rock” designs were not used in commerce (an element of trademark infringement) and that, even if they were, CafePress’s use was “fair use”—i.e., a descriptive or ornamental use of the phrase “Born to Rock” in a non-trademark sense.  The court struck down the first argument outright, stating that CafePress was being “facetious” in arguing that it did not use the mark in commerce given that CafePress actually imprints the designs on merchandise and ships that merchandise to customers. In considering the fair use argument, the court acknowledged that certain uses of the “Born to Rock” designs may constitute non-trademark fair use (e.g. “Born to Ride / Born to Rock”), but concluded that CafePress could not rely on fair use as a blanket defense for all of the designs.

Legal scholar Eric Goldman has pointed out that CafePress can raise other, stronger arguments in the future, including that the trademark is invalid and that consumers were not likely to be confused by CafePress’s use of the mark. Nonetheless, the district court’s denial of summary judgment does send a message to trademark holders: you can sue online service providers for trademark infringement based on user-generated content and you just might win.  The Digital Millennium Copyright Act (DMCA) creates a safe harbor for online service providers who promptly remove user-generated copyright-infringing content after receiving takedown notices, but there is no equivalent safe harbor for content that infringes trademarks (although chillingeffects.org, a website devoted to the DMCA, does note that “in the absence of any caselaw on the subject, should a trademark holder bring a claim for contributory infringement, an [online service provider] might be able to mount a valid defense by analogy to [DMCA] section 512(c).”).

Social media sites in particular may be easy targets for trademark claims based on user-generated content. Such sites often host “community pages” that serve as fan pages for brands without any authorization from the companies involved (for example, compare this official Facebook page established by a trademark holder with this community page run by a fan). In the wake of the case against CafePress, social media sites and other websites that host user-generated content should be aware of these trademark-related risks and the fact that the DMCA safe harbors do not apply to trademark claims.

Over the past year, a number of courts across the country have decided cases involving contributory infringement and the application of the Digital Millennium Copyright Act’s § 512(c) safe harbor in the social media context. Unfortunately for those who favor a uniform approach to the law, the precedent being developed is in many ways inconsistent. On one side of the country, the Ninth Circuit solidified § 512(c)’s protections for social media sites in UMG Recordings, Inc. v. Shelter Capital Partners LLC by holding that social media sites are not liable for infringing user-posted material subject to compliance with the DMCA’s notice and takedown procedures.  Several months later on the other side of the country, the Second Circuit addressed similar questions in Viacom Int’l, Inc. v. YouTube, Inc. Judge Cabranes’s opinion introduced the possibility that a social media site-owner’s “willful blindness” to infringing activity may trigger liability, thus raising the specter of (very) expensive litigation. The Seventh Circuit has now held in Flava Works, Inc. v. Gunter that online service providers are protected from contributory infringement liability—and therefore need not depend on the DMCA’s safe harbors at all—where they do not actually host allegedly infringing material or encourage copyright infringement but merely link to such material.

In an opinion written by one of the country’s preeminent circuit judges and cat fanciers, Richard Posner, the court addressed whether to uphold a preliminary injunction against social bookmarking site myVidster for contributory copyright infringement. myVidster allows users to “bookmark” videos they find on the Internet, such as videos from YouTube or Vimeo. myVidster automatically retrieves the “embed code”—code that permits the video to be viewed in a browser window separate from the original website (for example, when you link to a YouTube video on your Facebook page, the site automatically embeds the video so that your friends can view the video on Facebook rather than having to go to YouTube). myVidster then creates a new page for the embedded video, replete with advertisements.

Plaintiff Flava Works is an entertainment company that produces and streams adult videos through various websites. Flava allows its customers to download its content solely for personal use. Users are not permitted to upload Flava’s videos to other sites or to create any additional copies of the content. Thus, in Judge Posner’s view, a user who copies Flava’s videos by downloading them and then uploading the copyright-protected video to a third-party website is a direct infringer of Flava’s copyright. Because myVidster didn’t upload the infringing videos, the court found that myVidster did not directly infringe Flava’s copyright.

The court next considered whether myVidster should be held liable for contributory infringement based on such copying by Flava’s users. Posner disregarded the oft-cited Gershwin Publishing Corp. v. Columbia Artists Management, Inc. definition of contributory infringement in favor of a more succinct standard from Matthew Bender & Co. v. West Publishing Co.: contributory infringement is “personal conduct that encourages or assists [direct] infringement.” The court ultimately held that myVidster was not liable for contributory infringement for two reasons.

First, myVidster does not make any copies of Flava’s videos—whether on its own initiative or at its users’ direction—but instead links to videos on servers controlled by third parties. In bookmarking offending videos, myVidster’s users were not copying such videos. And by embedding those videos on its site, myVidster was not furthering any copying. Rather, the court found that myVidster effectively acts as an exchange, connecting the server hosting the video and myVidster’s users. Posner wrote:

[The user’s] bypassing Flava’s pay wall by viewing the uploaded copy is equivalent to stealing a copyrighted book from a bookstore and reading it. That is a bad thing to do (in either case) but it is not copyright infringement. The infringer is the customer of Flava who copied Flava’s copyrighted video by uploading it to the Internet.

Second, the court found that myVidster had done nothing to encourage uploaders to upload Flava’s videos. Therefore, myVidster did not “encourag[e]” infringement and was not a contributory infringer. As a result, myVidster had no need to resort to the § 512(c) safe harbors.

One of the most interesting aspects of the Flava Works opinion is its discussion of the various flavors (flavas?) of contributory infringement. Google and Facebook submitted an amicus curiae brief  in which they argued that the connections between myVidster’s (and other social bookmarking sites’) activities and any copyright infringement by users are simply too attenuated to constitute either direct or contributory infringement. They argued that myVidster was, at most, “contributing to contributory infringement.” Thus, myVidster’s potential infringement was not “secondary,” but rather, tertiary: the direct infringers are those who uploaded Flava’s copyrighted material, those who bookmarked the videos are arguably “secondary” infringers, while myVidster might be a “tertiary” infringer. Posner dismissed this argument, finding that common law notions of remoteness were sufficient to deal with this “contributing to contributory infringement” situation: “An injury will sometimes have a cascading effect that no potential injurer could calculate in deciding how carefully to act. The effect is clear in hindsight—but only in hindsight.” For Judge Posner, even in social media situations, there’s no need for the direct-secondary-tertiary “layer cake” model; there is simply infringement, contributory infringement, and non-infringement. And regardless of the theoretical “level” of removal of myVidster from the underlying direct infringement, myVidster was not “materially contributing” to that infringing activity—that is, myVidster’s actions were too remote from the uploader’s infringement—and was therefore not liable for contributory infringement by copying.

Judge Posner also addressed whether myVidster might be liable for contributory infringement based on public performance of Flava’s videos. The Copyright Act makes it unlawful “to transmit or otherwise communicate a performance . . . of the work . . . to the public . . . whether the members of the public capable of receiving the performance . . . receive it in the same place or in separate places and at the same time or at different times.” Posner identified two ways in which myVidster might infringe Flava’s performance right: “performance by uploading” and “performance by receiving.” On the “uploading” interpretation, “uploading plus bookmarking a video is a public performance because it enables a visitor to the website to receive (watch) the performance at will[.]” On the “receiving” interpretation, the performance occurs (or is, in other words, finalized) when the user clicks on and plays the video.

Posner dismissed the “uploading” view, arguing that myVidster is simply “giving web surfers addresses where they can find entertainment[,]” much like TimeOut and the New Yorker  list the details of various social events happening in the real world. According to Posner, the only infringer on the “uploading” view is the uploader himself. myVidster does not interfere with the data streaming directly from the host to the viewer, so myVidster did not contribute to the uploader’s infringement of Flava’s public performance right.

On the “receiving view,” the infringing act occurs when the myVidster users click “play” on Flava’s videos. Flava argued that, by providing an exchange that makes Flava’s videos available to myVidster’s users, myVidster provides “‘support services’ without which ‘it would [have been] difficult for the infringing activity to take place in the massive quantities alleged.’” Posner, however, was not persuaded by the “receiving” argument either. First, myVidster was not selling the allegedly infringing videos and thus had no direct pecuniary motive for pushing visitors to view Flava content bookmarked by the site’s users. Second, there was no substantial evidence that the videos were being accessed via myVidster rather than other websites. Thus, in Posner’s view, there was no basis to hold that myVidster was “abet[ting] others’ infringements” of Flava’s public performance right.

Judge Posner left open the possibility that myVidster had invited users to post infringing material, in which case it could be liable for inducing infringement. Similarly, he stated that myVidster’s now-discontinued sideloading service constituted direct—not secondary—infringement. Sideloading typically involves the transfer of data between two local devices. myVidster’s service allowed premium members to back up bookmarked videos on myVidster’s servers. As at least one commentator has noted, this raises interesting issues for sites like Pinterest and other social networks that periodically sideload copyrighted material posted by users on the presumption that such sideloading is authorized by, and therefore done at the direction of, the user. If such actions constitute direct infringement, the §512(c) safe harbors may not be available.

In today’s information economy, content owners are faced with a challenging decision regarding digital content. On the one hand, the viral nature of social media can mean unprecedented exposure as digital content is shared. On the other, that opportunity can come with significant legal risk if companies take an insufficiently careful approach to intellectual property clearance issues. One luxury clothing brand, Burberry Ltd., recently discovered just how substantial that legal risk can be.

Burberry approached social media with an innovative concept: “historical timelines” on its various social media pages, including Facebook, Twitter, and Instagram. These timelines featured photos of celebrities wearing Burberry’s iconic trench coats, scarves, and other products. Among Burberry’s chosen photos was a shot of Humphrey Bogart from the final scene of Casablanca, in which Bogart’s Rick, clad in a timeless Burberry trench, sends Ingrid Bergman’s Ilsa off to Brazzaville. While Burberry acquired permission to use the photo from photo agency Corbis, which manages the rights to various stock photos from Casablanca, Burberry failed to clear its use with Bogart LLC, which owns the actor’s publicity rights. Applicable law allows a celebrity to object to use of his or her name or likeness in a commercial context, particularly if the use is likely to cause members of the intended market to believe that the celebrity endorses the product. Bogart LLC alleged that Burberry’s use of the photo falsely implied that Bogart had endorsed the brand, thereby violating Bogart LLC’s publicity rights. Burberry countered, arguing that its timelines constituted “a historical positioning of the image within an educational project along with numerous other photographs of people wearing Burberry apparel over the last century.”

Although Burberry and Bogart LLC settled their pending state and federal cases for an undisclosed amount, this case provides a good example of the unexpected issues that can arise when brand managers fail to consider the full spectrum of rights that may be implicated by the use of photographs and other content. While the content industries have spent the last decade educating the public on copyright law’s effects in the digital media world, less attention has been paid to other areas of potential liability, such as trademark infringement and privacy and publicity rights violations, and their respective effects on the social media experience.

For example, in 2007, Virgin Mobile Australia (VMA) launched an advertising campaign using amateur photography culled from the social photo-sharing site, Flickr. The photos used by VMA were licensed under a Creative Commons “Attribution” license, which requires only that the original creator—that is, the copyright holder—be given credit. VMA chose for its campaign a photo of then-15-year-old Alison Chang, taken by her church youth counselor and uploaded by him to Flickr. Although VMA had appropriate copyright clearance to use the counselor’s picture under the Creative Commons license, Chang’s parents sued VMA for failing to get permission from Chang or her parents to use Chang’s name or likeness. Although the case was dismissed on procedural grounds, the incident illustrates how easily (and often) clearance procedures are overlooked when it comes to Internet-based content.

Similar cases have raised complex issues relating to federal preemption of state law claims. For example, in Laws v. Sony Music Entertainment, Inc., the plaintiff sued Jennifer Lopez and LL Cool J, alleging misappropriation of her name and voice through use of a sound recording on which the plaintiff’s voice was featured. The defendants had obtained a license to use the sound recording on which the plaintiff’s voice was featured, but had not obtained from the plaintiff the right to use her voice. Nonetheless, the Ninth Circuit held that, on this set of facts, the federal Copyright Act preempted the plaintiff’s state law right-of-publicity claim. Thus, her permission was not required for the defendants to use the validly licensed sound recording. By contrast, a different Ninth Circuit panel in Downing v. Abercrombie & Fitch, Inc., held that the Copyright Act did not preempt the plaintiffs’ state law publicity claims based on Abercrombie’s advertising use of a photo of the plaintiffs taken after the 1965 Makaha International Surf Championship in Hawaii. Thus, Abercrombie should have sought the plaintiffs’ permission in the first instance. The preemption inquiry is fact-bound—the Copyright Act preempts state law publicity claims in some circumstances, but not others.

While the details of these preemption cases exceed the scope of this article, suffice it to say that a company’s social media marketing personnel may not have the expertise to wade through such complex clearance issues. A clearance system that focuses narrowly on copyright issues and doesn’t consider other forms of intellectual property may therefore result in unexpected claims. It is also worth noting that the safe harbors provided by the Digital Millennium Copyright Act apply only to copyright claims, not other types of claims such as those mentioned above, and in any event, provide protection only with respect to user-generated content, not content posted by a company’s own employees. Therefore, companies should not assume that the DMCA will shield them from all liability for content posted on their social media pages.

Social media is an exciting new channel for reaching both current and prospective customers. But from a rights-clearance perspective, the old rules largely remain in force. Accordingly, companies’ review procedures for company-driven social media content should, to the extent possible, mirror the process they undertake for print ads and other traditional media. And where that may not be feasible (given the speed and flexibility often required on social media platforms), companies should institute rigorous policies and train marketing associates on how to avoid potential liability.