Catching Up on 4 Months of Online Copyright Cases–Myxer, Hotfile, Megaupload, Flava Works, Zediva, Blue Nile, Perfect 10, Rojadirecta
By Eric Goldman
Online copyright cases have been coming at such a furious pace that I haven’t had a chance to keep up. This blog post wraps up the last 4 months of decisions.
Arista Records v. Myxer, Inc., 2:08-cv-03935-GAF-JC (C.D. Cal. April 1, 2011)
Myxer allows users to upload sound files and create downloadable ringtones up to 40 seconds long. The plaintiffs claimed 90+% of the ringtones were at least 34 seconds; but Myxer claims its average length is 25 seconds and that this is shorter than Audible Magic can reliably filter. Myxer claims to be a tool for independent artists to easily get their music into the mobile environment, but of course users can upload third party copyrighted material. UMG objects to the ringtones at Myxer because the downloadable ringtones disrupt its paid-download ringtones business. For reasons that aren’t explained, UMG apparently never sent a proper 512 takedown notice, and Myxer treated the complaint as a takedown notice.
In a perplexing but unexplained statement, the court says “The undisputed facts in the present record establish that Myxer has directly infringed at least one of Plaintiff’s exclusive rights, pursuant to § 106.” The court further says there is no “volitional” defense in the 9th Circuit. This shifts the battle over the 512 safe harbors.
Agreeing with the UMG and YouTube cases, the court says user downloading qualifies for 512 coverage. The adequacy of Myxer’s repeat infringer policy and the expeditiousness of its removal was deferred to trial. Citing Io, the fact that terminated users could re-register didn’t disqualify Myxer from the 512 defense. The court rejected UMG’s arguments that Myxer had generalized knowledge of infringement and says “to the extent that Plaintiff suggests that Myxer should have taken additional steps to filter infringing material on the Myxer Website, ‘the DMCA does not place the burden of ferreting out infringement on the service provider.’” With respect to direct financial benefit, the court says the test is “whether the infringing activity constitutes a ‘draw’ for users, not just an added benefit.”
Myxer also loses a fair use defense. It’s pretty clear that Myxer kept the maximum ringtone length fairly short as part of a fair use strategy, but the court circularly says Myxer took the best part, i.e., the chorus.
The most remarkable thing about this opinion is its length–75 pages. Easily, 17 USC 512 opinions are, on average, double or even triple the length of 47 USC 230 opinions. That’s because 17 USC 512 has so many more words–and elements–than the pithy 47 USC 230, and plaintiffs contest virtually every word in a 512 defense. This makes for long and sometimes ponderous opinions. As a reader of judicial opinions, let’s hear it for short immunities!
Disney Enterprises, Inc. v. Hotfile Corp., 11-20427-CIV-JORDAN (S.D. Fla. July 8, 2011)
Hotfile is a cyberlocker. Part of its business model is to give faster download speeds to paying members. Hotfile pays uploaders who post popular large files, which could naturally encourage people to upload infringing files.
Unlike the Myxer case, and following in the footsteps of the old Netcom case, the court rejects Hotfile’s direct liability: “the law is clear that Hotfile and Mr. Titov are not liable for direct copyright infringement because they own and manage internet facilities that allow others to upload and download copyrighted material…. nothing in the complaint alleges that Hotfile or Mr. Titov took direct, volitional steps to violate the plaintiffs’ infringement” [sic—the last word should have been “copyright”]. As a result, the court grants the defendants’ motion to dismiss the direct infringement claim.
The court distinguishes some 1990s cases (Webbworld and Russ Hardenburgh, among others) suggesting otherwise, saying that, in those cases, the defendants uploaded files themselves or used software to search for files. The court also expressly disagrees with Capitol v. MP3Tunes and Arista v. USENET.com. This is a pretty unpersuasive distinction based on the facts of the precedent cases, and I think it highlights the weakness of the “volitional” defense. The fact is that all but the most passive of hosts or conduits take some affirmative steps towards customizing the downloader’s experience, and trying to parse which of those steps constitute “volitional” conduct and which don’t is leading to the inevitable doctrinal incoherence.
Having said that, I thought we decided that web hosts aren’t directly liable for user-caused infringement 15 years ago. I can’t believe we’re still wasting our time on this issue in 2011.
Not surprisingly, the secondary infringement claims survived the motion to dismiss.
Perfect 10, Inc. v. Megaupload Ltd., 2011 WL 3203117 (S.D. Cal. July 27, 2011)
Megaupload is another cyberlocker with a terrible brand name. Its business model appears similar to Hotfile’s. Like Hotfile, it also sought a motion to dismiss the direct infringement claim. Unlike Hotfile, the court sees enough volitional conduct to survive the claim. It would have been great if the opinion did a compare/contrast with the Hotfile opinion, but sadly Hotfile wasn’t mentioned at all–but, naturally, the MP3Tunes and Usenet.com cases (the ones the Hotfile court trashed) were. Sigh.
As usual, Perfect 10 apparently failed to send a proper 512(c)(3) takedown notice. They sent 22 notices in all, but allegedly 21 related to stuff that wasn’t theirs. (Huh?) The court can’t do much with this on a motion to dismiss and tells Megaupload to put it in its answer. The court does dismiss the vicarious infringement claim, saying that Perfect 10 didn’t properly allege the requisite supervision over infringing activity beyond alleging supervision of the system.
Perfect 10’s trademark infringement claims get Dastar-ed, but the dilution claims survive. Perfect 10’s allegation is tarnishment because its high-quality photos are intermingled with junk photos. Seriously?
Flava Works, Inc. v. Gunter, 2011 WL 3205399 (N.D. Ill. July 27, 2011)
This case involves myVidster, another lousy trademark. After the cataclysmic flameouts of Napster, Grokster, Aimster, Friendster and so many others, who thinks the “-ster” suffix is still cool??? myVidster is small operation, mostly focused on porn, that allows users to “bookmark” a video. From the description, it sounds like myVidster is like a delicious-style link aggregator that embeds the linked videos, but the court confusingly goes out of its way to say myVidster isn’t a linking site:
MyVidster does not simply link to video files displayed on another site; it embeds the files on its own site at the direction of users. In other words, when a visitor to myVidster clicks on a video that is posted there, the video plays directly on myVidster, and the visitor remains on the myVidster site; he or she is not taken to the site that hosts the video file.
Right, that sounds like a linking site to me. It’s not clear to me how defense counsel failed to get the judge to grasp this essential fact.
The court describes that myVidster adds some metadata. The court shares this panicky observation: “In cases where a myVidster user bookmarks another user’s bookmark, the ‘source code’ will be a myVidster URL even though the original file of the video may be hosted elsewhere.” Quelle horreur! Exactly where would the court like that link to go? The court might not have realized that linking user #1 might have added its own metadata or other users may have posted comments that linking user #2 would want to incorporate.
Flava produces gay ethnic porn. It discovered that myVidster users were embedding lots of its content. It sent multiple DMCA takedown notices. Flava claimed that myVidster’s response to its takedown notices was erratic–sometimes it would take everything down, sometimes it would remove the link but leave up a thumbnail, and sometimes it would ignore the notice. myVidster responded that it did a better job than that and produced a chart showing its responsiveness. The court discounts the chart because it was not prepared on a timely basis and wasn’t sufficiently credible in its preparation, but in a footnote, the court makes it clear it just didn’t believe myVidster’s principal:
In any event, although it does appear that it was sometimes like pulling teeth to obtain full compliance from Gunter, as discussed infra, the crux of the problem here is not so much the removal of the infringing videos; it is Gunter’s attitude toward copyright protection and his related refusal to adopt measures to prevent or reduce copyright infringement on myVidster as well as to adopt and implement an appropriate policy regarding repeat infringers.
This quote shows that it’s not a good idea to displease judges hearing online copyright cases. As we learned in Io v. Veoh, websites that go beyond the DMCA and proactively fight copyright infringement get a lot of extra credit from judges. For more on this in the trademark context, see Stacey Dogan’s new article, “We Know It When We See It.”
In myVidster’s case, its principal took the highly unusual position that anything publicly posted to the Internet was fair game for linking–irrespective of its underlying copyright status. For example, the principal said: “my policy on repeat infringers are those who are using myVidster as a ways and means to distribute content that is not publicly available.” He repeatedly told copyright complainants to take it up with the video host, because removing the links from his site wouldn’t remove the file itself from the Internet. This is true, but it’s not really sensitive to the emerging legal obligations of intermediaries.
Based on the view that anything publicly posted was freely linkable, user-posted links implicated myVidster’s repeat infringer policy only if they went to password-protected sites or private URLs. I find it hard to believe that a lawyer recommended this standard. Because of this policy’s incredibly narrow scope, myVidster only had a single user who got a repeat infringer warning. The court didn’t have kind words for this policy: “Gunter’s ‘repeat infringer’ policy is in fact no policy at all, at least with respect to copyright infringement….His definition of ‘repeat infringer’ does not encompass copyright law.”
The court has little trouble establishing prima facie contributory copyright infringement. Users posted links to infringing videos. myVidster’s principal knew of these links from the DMCA takedown notices. He didn’t remove the links (or only partially remove them) upon notice. The court says his view about the free linkability of publicly posted but infringing videos “is the epitome of ‘willful blindness.’” The court also cites his failure to implement filters to prevent repeat infringement or to take recourse against the infringing users.
The general operation of a linking site counted as a material contribution. It also cited the fact that the site offered paid hosting as a way to avoid the possible breakage of links (the court said that “encouraged” infringement), and the site’s principal also encouraged infringement through his own personal “favorites” list, which included “Star Trek,” “Crank 2,” and “Hancock”–and at a hearing, he stubbornly said that only the copyright owner could tell if the movie posting was infringing. The court also cited the website’s failure to discourage users from infringing copyrights—this was a doctrinal shortcut by the court.
Needless to say, myVidster doesn’t get a 512 defense. The court says brusquely: “It is difficult for us to understand how defendants can argue with a straight face that they have adopted and reasonably implemented a ‘repeat infringer’ policy.”
It’s frustrating to see a ruling like this. myVidster had a legitimate chance of legal success if it got competent legal counsel from day 1 (or, if they did, if they listened to their lawyers) and baked in even the minimum industry-standard anti-infringement efforts. But the misguided view that anything online was linkable couldn’t be defended in court, and the extremeness of that view (combined with the principal’s mule-like refusal to give any ground) gave the court too much liberty to say doctrinally unhelpful things as part of a judicial bodyslam.
Although it’s largely mooted by the summary judgment opinion, the court’s opinion on the motion to dismiss is also interesting. The court dismissed the direct infringement claim (correctly IMO), and on the vicarious infringement claim, echoed the (uncited) Perfect 10 opinion by saying that “To sufficiently allege the element of “right and ability to supervise,” plaintiff will have to allege more than the mere ownership and operation of myVidster.” Similarly, the court follows in the (uncited) Myxer court’s footsteps by saying the plaintiff “does not allege that the presence of the infringing material on the site enhances the site’s attractiveness or draws customers.” As a result, the court tossed the vicarious infringement claim too. The inducement claim also dropped out because of “formulaic” pleading. Finally, the court tossed the trademark claims because they didn’t show how myVidster made a use in commerce. Yet, all of these defense-favorable rulings that narrowed the case didn’t save myVidster from a disastrous contributory copyright infringement ruling. But, for other defendants, it shows how victory was attainable for myVidster had it followed a better recipe for clean living.
Warner Bros. Entertainment Inc. v. WTV Systems, Inc., 2:11-cv-02817-JFW -E (C.D. Cal. Aug. 1, 2011)
Unlike most of the other cases here, this is not principally a secondary infringement case. However, it bears some commonalities with the other cases.
Zediva was trying to find a way to take advantage of the Cablevision case to offer a video “rental” service online. Zediva would acquire DVDs and essentially check them out to paying customers such that only 1 customer could enjoy the DVD at any one time. Frankly, if the Cablevision case is good law, Zediva has a point. However, I’ve been troubled by the Cablevision precedent, and this opinion shows that Cablevision isn’t a very robust precedent as the court basically says “nyet” to a similar application in a new context.
Zediva’s main failing is that it doesn’t maintain separate copies of the downloadable file for each viewer as Cablevision did. Of course, that probably wouldn’t be profitable for Zediva, because it presumably needs to amortize the DVD’s purchase price over multiple viewers. (Unlike Cablevision, which had already paid for the license fee to obtain the digital bits through its cable retransmission license). Then again, it would be really lame if the legal rule is that we have to build wasteful redundant systems to avoid copyright infringement. That was the implicit rule from Cablevision, and this court doesn’t let Zediva cut that corner.
The court says Zediva is publicly performing the videos. Its main citation is to the On Command case from 2 decades ago and the Redd Horne case from over a quarter-century ago. How’s this for some circular logic: “Defendants’ transmissions are ‘to the public’ because the relationship between Defendants, as the transmitter of the performance, and the audience, which in this case consists of their customers, is a commercial, ‘public’ relationship regardless of where the viewing takes place.” Huh? The court distinguishes Cablevision because each viewer had his/her own dedicated copy of the video, unlike the shared copy of the DVD in this case.
Citing the Myxer case (discussed above), the court says that the Ninth Circuit hasn’t adopted a volitional defense.
This opinion is pretty bad, but IMO the worst part is the court’s discussion of irreparable injury. Check out this parade of horribles:
* Zediva might “jeopardize the continued existence of Plaintiffs’ licensees’ businesses” (really?!) because consumers find Zediva’s service more attractive than the crappy options officially sanctioned by the movie studios.
* “Defendants’ service threatens the development of a successful and lawful video on demand market and, in particular, the growing internet-based video on demand market. The presence of Defendants’ service in this market threatens to confuse consumers about video on demand products, and to create incorrect but lasting impressions with consumers about what constitutes lawful video on demand exploitation of Plaintiffs’ Copyrighted Works, including confusion or doubt regarding whether payment is required for access to the Copyrighted Works.”
* “Defendants’ service also threatens the development of a successful and lawful video on demand market by offering a sub-optimal customer experience and, thus, tarnishing customers’ perception of video on demand as an attractive option for viewing Plaintiffs’ Copyrighted Works.” In particular, the court cites Zediva’s inability to let multiple customers share the same DVD simultaneously against it, saying that telling customers that a video is “out of stock” will turn off video-on-demand customers permanently.
After reading this, I get the sense that the movie studios think the video-on-demand industry is more fragile than a 1980s Jaguar.
The court spent a lot of time discussing the movie studios’ windowing. This reminded me of recent language from the Second Circuit Barclays’ case:
The adoption of new technology that injures or destroys present business models is commonplace. Whether fair or not, that cannot, without more, be prevented by application of the misappropriation tort.
Perhaps copyright law can prevent it.
Blue Nile Inc. v. Ideal Diamond Solutions Inc., 2011 WL 3360664 (W.D. Wash. Aug. 3, 2011)
This is a piercing-the-corporate-veil case. Chasin is the principal of IDS, an outsourced online jewelry website operator for offline jewelry retailers. Blue Nile (remember them?) claims IDS republished its copyrighted product shots. Chasin defended that he personally didn’t do it, so he should not be personally liable. The court grants summary judgment to Blue Nile.
On direct infringement, the court says (footnotes omitted):
There is no question that IDS was the “brainchild” of Larry Chasin, that IDS “was a small company”, and that Chasin “controlled the corporate affairs”. In addition to creating and controlling IDS, Chasin licensed the development of the infringing websites, and had the power to direct the removal of infringing content.
Any lack of knowledge goes to damages as an innocent infringer, not to the merits. Of course, if the photos had a copyright notice, the innocent infringement defense may be unavailable, so I’m not sure the court did Chasin any favors there.
On vicarious infringement, the court says (cites omitted):
Chasin admits that he had the ability to remove the infringing content and that he controlled the corporate affairs of IDS; thus he had the right and ability to supervise the infringing activity. He also admits that he personally invested “over $440,000 cash” into IDS and that he received salary and benefits from IDS, thereby giving him a direct financial interest in IDS.
Without more insight into the source of the allegedly infringing photos, it’s hard to know just how bad this ruling is. I believe that officer/investor liability for a company’s copyright infringement remains undertheorized, and I’d love to see more attention paid both to the doctrinal fine points and to the policy implications of treating someone like Chasin as a direct and vicarious copyright infringer. If you’re looking for a paper topic, this might be worth considering.
I also continue to howl that product shot-related lawsuits are a ridiculous tax on the entire retailing industry, and an area that desperately needs reform. One idea I’ve been kicking around is a commons-style repository of product shots. If you’re interested in kicking this idea around, contact me.
Perfect 10 v. Google, No. 10-56316 (9th Cir. Aug. 3, 2011)
Perfect 10 is in the Ninth Circuit again. This Ninth Circuit panel (the opinion was written by Judge Ikuta) made it very clear that it did not want to discuss the main event–i.e., the merits of Perfect 10’s lawsuit against Google. I think that’s because the panel knows that the case’s substantive issues are coming back to it soon enough (it’s interlocutory now), and this panel didn’t want to skew the substantive analysis when the time comes.
Instead, this opinion is about the standards for a preliminary injunction in copyright cases. The district court denied preliminary injunctive relief to Perfect 10, while at the same time ruling mostly in Google’s favor. In this opinion, the Ninth Circuit weaves eBay v. MercExchange into the standards for preliminary injunctions in copyright cases. The court holds:
We therefore conclude that the propriety of injunctive relief in cases arising under the Copyright Act must be evaluated on a case-by-case basis in accord with traditional equitable principles and without the aid of presumptions or a “thumb on the scale” in favor of issuing such relief.
As a practical matter, this opinion ought to make it harder for copyright plaintiffs to get preliminary injunctions, And unlike many recent Ninth Circuit opinions, I believe there’s a realistic chance that other Ninth Circuit panels will honor this holding. The opinion is well-reasoned and a logical extension of the Supreme Court’s eBay decision.
The court goes on to say that it wasn’t an abuse of discretion to deny a preliminary injunction to Perfect 10, even if it meant they would go out of business. (We can only hope). Basically, the court called BS on Perfect 10’s claim that Google is wrecking its business, noting (among other things) that Perfect 10 “failed to submit a statement from even a single former subscriber who ceased paying for Perfect 10’s service because of the content freely available via Google.” It is really hard to be too sympathetic towards a copyright owner who seems far more passionate about litigating than in giving consumers reasons to patronize it.
Puerto 80 Projects SLU v. USA, 1:11-cv-04139-PAC (SDNY Aug. 4, 2011)
Rojadirecta is a Spanish operation that runs linking sites. Naturally, some links are to infringing material. Rather than send takedown notices or sue Rojadirecta for infringement, the US government (DHS ICE) just took the domain names on the theory that they were being used to commit criminal copyright infringement. There are numerous problems with ICE’s seizure, including jurisdiction (Rojadirecta is legal in its home country), doctrine (the US government will have a tough time showing criminal copyright infringement), procedure (all of this was done without an adversarial process) and the Constitution (the domain name was the functional equivalent of a printing press for Rojadirecta). The DHS ICE’s efforts to shut down a purportedly “rogue” site has, in fact, caused our own government to go rogue itself.
In this lawsuit, Rojadirecta tries to get its domain name back. The applicable statute was designed to govern physical chattel, not virtual printing presses. In a remarkably tone-deaf opinion, the court has none of it. The statute at issue effectively puts the burden on Rojadirecta–an odd place for the burden to rest, given that the US government still hasn’t proven anything–and the court doesn’t see enough reason to disturb the status quo.
Consider the tone-deafness of the court’s response, keeping in mind that the underlying allegation is criminal copyright infringement, where the government’s burden should be higher than in a civil case. Rojadirecta argues that it’s losing users who can’t find it due to the seized domains. The court responds that Rojadirecta has other domains, and “Rojadirecta has a large internet presence and can simply distribute information about the seizure and its new domain names to its customers.” Well, yes, it can bang the drum through other media, but by definition there’s no way to ensure the publicity reaches folks who only knew of Rojadirecta at the seized domains. Contrast Judge Kozinski’s discussion in the Toyota v. Tabari case, a trademark lawsuit over a domain name:
the Tabaris needed to communicate that they specialize in Lexus vehicles, and using the Lexus mark in their domain names accomplished this goal. While using Lexus in their domain names wasn’t the only way to communicate the nature of their business, the same could be said of virtually any choice the Tabaris made about how to convey their message.
In general, it’s repugnant under the First Amendment for the court to second-guess a publisher’s choices of how to communicate with its audience. Here, the court does just that, and ignores any lost audience because ICE has blocked the publisher’s preferred communication method. Corynne at the EFF has more to say about this point.
The court makes the same error when it expressly discusses the First Amendment. The court says “the fact that visitors must now go to other websites to partake in the same discussions is clearly not the kind of substantial hardship that Congress intended to ameliorate in enacting § 983.” That is true, but only because 983 contemplated the government would seize physical chattels that putatively facilitate crimes, not virtual printing presses. So when the government misuses its power to reach speech-facilitating chattels, the court should modulate accordingly.
The judge doesn’t permanently shut the door on the First Amendment issues or other relevant defenses, but he’s apparently not yet appreciated the magnitude of the government’s errors either.