EFF On IsoHunt: Bad Facts Make Bad Law

from the don't-get-distracted dept

As Gary Fung is seeking a rehearing of the IsoHunt case in the 9th Circuit, two amicus briefs were filed yesterday. The first from the EFF and the second from Google. Neither brief suggests that Fung should get off as innocent, or that he did nothing wrong. Rather, both are worried about how the broad ruling by the court for the specific situation regarding Fung and IsoHunt will lead to further abuse by copyright holders and massive chilling effects on service providers. The EFF notes that while Fung/IsoHunt may have been bad actors, it appears that the court used this to go way overboard in creating new and dangerous standards for copyright.

This Panel Opinion is a classic case of bad facts making bad law. Amicus Electronic Frontier Foundation does not file this brief to dispute the Court's factual conclusions regarding the conduct at issue in this case. However, the Panel Opinion went far beyond what was necessary to address that conduct. As a result, it has created new legal uncertainty for online service providers and their customers, undermining over a decade of legislation and jurisprudence designed to help reduce that uncertainty. A predictable legal environment has proven to be crucial not only the growth of the Internet generally, but the growth of innovative platforms for free expression, in particular. This case should not provide a vehicle to impede that development.

In particular, they're quite (reasonably) worried at the court's broad interpretation of causation here, in which the court suggests that the most minor example of inducement can lead to liability for all infringement, even if the site had nothing to do with it.

Most important, the Panel Opinion adopts a "loose causation theory" that disconnects the scope of inducement liability from the defendants' acts—raising the troubling possibility that a single inducing act (such as a message to one customer) could open the floodgates to liability for third-party infringement entirely unrelated to that act. The Opinion's loose causation theory conflicts with fundamental common law principles of proximate cause essential to both predictability and fairness. The Panel's decision to depart from those principles was apparently based on the unfounded assumption that the Supreme Court's decision in Metro-Goldwyn-Mayer Studios Inc. v. Grokster, Ltd., 545 U.S. 913 (2005) requires it. Not so. First, Grokster expressly recognized that secondary liability under copyright derives from common law principles. Second, given that Grokster's specific inducement standard was imported from patent law, it is more likely that the Court also intended to import the analytical framework patent law applies where, as here, a service is capable of both infringing and non-infringing uses.

Meanwhile, Google's focus is on the question of "financial benefit directly attributable" from infringing activities. The DMCA, of course, includes that as one of the prongs for testing whether or not a site gets safe harbor protections. Most courts have found that indirect profits don't make you lose safe harbors: i.e., if you're just making money on ads from a page that has infringing content, that's not "directly attributable". Most people recognize that for it to be "directly attributable" then it needs to be something like actually selling the infringing content, and the direct profits from that action need to be shown. Instead, copyright maximalists have tried to argue that if you have infringement on a site and some money is made (i.e., there are ads or affiliate links) then, that violates that prong of the test and you lose your safe harbors. Most courts have realized that's crazy. But the Fung ruling went very close to the maximalist view, and that (quite reasonably) has Google concerned. Specifically, it's concerned that the ruling could be read to mean that any "influence" a site has over content means it's liable for all of the content on the site:

There is a danger that this passage could be misconstrued to stand for a broader proposition that we do not believe the panel intended: that any time an online service provider is found to have exercised "substantial influence" over any user-submitted content on its service—no matter what that finding was based on—it thereby loses its DMCA safe harbor protections for all user-submitted content on the entire service. This is how some copyright plaintiffs have already tried to read the panel's ruling. In a recent submission to the Southern District of New York in the Viacom v. YouTube case, for example, the plaintiffs have asserted, citing the panel opinion, that this Court “made clear that where DMCA eligibility is unavailable due to the right and ability to control prong of the safe harbor, the DMCA defense is broadly lost as to all clips in suit.” Ltr. from Paul M. Smith to Hon. Louis L. Stanton at 2 (March 22, 2013) (attached as Ex. 1).

As Google right notes, this would lead to "absurd results."

Imagine, for example, a video-hosting service that was otherwise eligible for the section 512(c) safe harbor, but that on one occasion commissioned a particular user to upload a video that, unbeknownst to the service, turned out to be infringing. A court might conclude that the service exerted a “substantial influence” over that instance of infringement and, if the service earned a direct financial benefit from it, there would be grounds for denying the safe harbor for a claim based on that video. But it would make no sense to thereby disqualify the service provider from DMCA protection across the board—even for countless other videos whose posting it did not control or from which it earned no benefit.

Likewise, consider a search engine eligible for protection under the section 512(d) safe harbor for linking to infringing material online. If one of the millions of links provided by the search engine pointed users to infringing material that had been authored by the search engine itself and that users were charged to view, a finding of control plus financial benefit might be warranted for that particular link. But, again, there would be no plausible basis for categorically depriving the service of the safe harbor for the millions of unrelated links it delivers to material that it does not control or financially benefit from.

But, of course, that's crazy (even if it's exactly what many maximalists actually do seem to want). Hopefully, the court is willing to revisit these issues and recognize that its original ruling went overboard because of the situations in this case, and that could unfairly mess up other legitimate offerings.

Filed Under: causation, copyright, dmca, financial benefit, gary fung, inducement, liability, red flags, safe harbors, secondary liability
Companies: eff, google, isohunt

Google Defends The DMCA's Safe Harbors Against The MPAA's Attempts To Reinterpret Them In Hotfile Case

from the reasonable-brief dept

We've noted that the MPAA's case against Hotfile is surprisingly weak, and seems to be arguing that usage alone is proof of Hotfile's complicity in any infringement done by users. This is a strange argument, which is more smoke and mirrors than anything legit. It's as if the MPAA believes that if it just screams "but... but... piracy!" loud enough, the judge will forget to look at the actual law. However, in a bit of a surprising move, Google is trying to step in and inform the judge on one key piece of the case, with an amicus brief.

At issue is the standard used to judge whether or not the DMCA's safe harbors apply. Obviously, Google has a vested interest in having previous court rulings on the DMCA's safe harbor followed in this case, not just because those rulings protect Google, but because they're the only way the DMCA actually makes any sense. Google's argument is pretty clear and well-argued: as the DMCA safe harbors themselves, the massive DMCA caselaw and the Congressional history of the DMCA all show in pretty great detail, to lose the DMCA's safe harbors, a company has to have specific knowledge of infringement, not just general knowledge that its tool is used for infringement. The MPAA's argument is effectively the opposite -- and is completely nonsensical: that if it can show that enough people infringed, then it should be assumed that Hotfile could have stopped the infringement. As the Google argument explains simply, that's a clear distortion of the law. In fact, they point out that the MPAA is so far off the reservation on this one that it can't even find DMCA cases to support its position, instead choosing two cases that have nothing to do with the DMCA.

Amusingly (and ridiculously), the MPAA is so freaked out about Google explaining the law on this one key point that it's asking the judge not to allow the brief, suggesting that rather than providing a separate third party view, Google's brief is really just re-arguing Hotfile's position. That's an argument that makes little sense, though. Google's brief is pretty narrowly focused on just one key issue in the much larger case surrounding Hotfile: it's merely asking the court to make sure it follows the same standard used in most other courts in the country. It makes no specific statements concerning Hotfile's actions at all. The issue Google raises is important because this court and this Circuit have not specifically ruled on the DMCA safe harbors -- a fact you can bet the MPAA knows well. No doubt, the MPAA is hoping that a different ruling in this case can lead to split that would (it hopes) lead to all of those many other DMCA rulings protecting safe harbors being overturned.

I'm sure that some simplistic commenters may try to summarize this case as Google defending Hotfile, but the specifics of the filing make it pretty clear that's not what's going on at all. It merely points out the well accepted standards and practices for removing DMCA safe harbor protections -- which are quite different than the interpretation of the safe harbors that the MPAA gave the court in its motion for summary judgment.

Filed Under: actual knowledge, dmca, red flags, safe harbors
Companies: google, hotfile, megaupload, mpaa

Viacom Says That By Letting People View Videos On Phones, YouTube Loses DMCA Safe Harbors

from the crazy-talk dept

We've seen how the legacy entertainment industry is trying, desperately, to undo the DMCA's safe harbors, in an attempt to force internet companies to effectively kill off user-generated content. Let's be honest here, the goal is to turn the internet more into TV, so there's less competition for the legacy players, and then they can retain their gatekeeper role. Viacom has been at the forefront of this, making some rather crazy claims regarding the DMCA's safe harbors. In the ongoing case between Viacom and YouTube, the court asked both sides to comment further on the "red flag" provisions in the DMCA.

This is the crux of the fight here, and the real problem is that (surprise, surprise) the drafters of the DMCA didn't fully understand what they were drafting, and drafted parts of the law such that they could be read in conflict. The key points are that the DMCA safe harbors require expeditious takedown of content on receipt of a valid notice of infringement. The rules for a valid notice are pretty clear, and courts have said that invalid notices do not, then, create a burden to remove content. At the same time, there is a part of the bill that says you can lose safe harbors if there are "red flags" -- that the operator knew of infringing activities on the site. So the entertainment industry's argument is that if you are aware that there's infringement on your site, you lose your safe harbors. That, of course, makes no sense. The only way to read the law such that the two sections are consistent is to read it the way YouTube reads it: if you are aware of specific infringement, rather than just general infringement, then you need to do something. That makes sense.

The problem is that Viacom execs live in a world of denial, where they pretend that it's obvious what's infringing and what's not. Thus, to Viacom, YouTube obviously knew what was infringing and what wasn't... even though Viacom itself falsely reported hundreds of videos as infringing... even though they were not.

In filing the response to the court's questions, however, Viacom takes this argument even further. Showing off its technological cluelessness, it argues that the mere act of YouTube setting up an API that allows YouTube videos to show on mobile phones and tablets outside of the browser, means that YouTube loses its safe harbors.

YouTube’s argument that its syndication must be protected because it merely "makes videos accessible" proves far too much. YT Ltr. 8. Any propagation of user-uploaded copyrighted material increases accessibility in some sense. If the DMCA immunized any act of infringement that increases access, an operator could keep the safe harbor merely by showing that its infringement enabled others to view the work, notwithstanding copyright law’s grant of exclusive rights to the copyright holder. This is why Congress limited the safe harbor to acts of infringement that occur "by reason of storage at the direction of the user."

To say that this is stretching the law to ridiculous levels would be an understatement. Effectively, what Viacom is arguing is that if someone uploads a video to YouTube, they only intend to let people watch it via the web. And the idea that YouTube might make such works available on mobile devices somehow goes against the wishes of its users, and shows YouTube making active decisions for the distribution of content. That's insane. And technologically clueless. All YouTube has done is optimize the content for the device viewing it. That's it. People uploading videos to YouTube do so because they want those videos to be seen. They are platform agnostic. Viacom's desperation to break the safe harbors of the DMCA here are really quite crazy, since Viacom itself relies on such safe harbors too -- and is putting its own business at risk out of some sort of quixotic obsession with YouTube.

Filed Under: copyright, dmca, red flags, safe harbors
Companies: google, viacom, youtube

YouTube's Reply In Viacom Case Demolishes Each Of Viacom's Key Arguments

from the good-luck,-viacom dept

It surprised no one last year when Viacom appealed the complete smackdown it got from the district court in its lawsuit against YouTube. As I noted when the appeal was filed, Viacom's arguments made very little sense, and in YouTube's response (which apparently hasn't yet been filed, but will be shortly), it seems that the company does a nice job debunking each of Viacom's arguments. The key point in this has been focused on the whole "red flag" question, of whether or not the red flag parts of the DMCA meant that YouTube gave up its safe harbors, despite its proactive stance on taking down content when alerted to it being infringing. Viacom is trying to change the plain meaning of the statute into meaning that a service provider must proactively investigate and see if it can find infringing material. YouTube points out that this is not what the law says.

Congress crafted such a “high bar for finding ‘red flag’ knowledge” because it recognized that it is challenging for service providers to distinguish infringing from non-infringing material. UMG II, 665 F. Supp. 2d at 1111. No red flag exists, therefore, where circumstances leave uncertain whether the material is protected by copyright at all, or whether a particular use is licensed or, if unlicensed, a fair use. H.R. Rep. 105-551(II), at 57-58. A service provider has no “investigative duties” to make such determinations. CCBill, 488 F.3d at 1114. Indeed, “if investigation of ‘facts and circumstances’ is required to identify material as infringing, then those facts and circumstances are not ‘red flags.’” UMG II, 665 F. Supp. 2d at 1108.

There is no room here for an approach that would convert generalized information that unspecified infringement is occurring somewhere into red-flag knowledge. Red flags, instead, are necessarily specific: they indicate unmistakably that a given piece of material or a given activity using that material is infringing. And they create an obligation for the service provider to expeditiously remove that particular material. That does not render red-flag knowledge duplicative of actual knowledge. The red-flag provision does independent work, establishing that a service provider can lose DMCA protection by failing to remove a given item, even if the plaintiff cannot prove that the provider subjectively thought the item was infringing.

The filing also goes through many of the arguments we've seen before, including debunking the simply ridiculous claim from Viacom and its supporters that knowing that some people could use the service for infringing purposes means a loss of safe harbors. If that's the case, then there are no safe harbors. The filing also notes that, contrary to the "straw man" suggested by Viacom, saying that a service provider needs specific knowledge does not mean that the red flag provisions are meaningless. It just means that YouTube must have specific knowledge that the works were infringing, and the court reasonably established that YouTube did not have that with the clips that Viacom was complaining about (in part because even Viacom wasn't always sure which clips were infringing, and which it had uploaded itself).

The filing is pretty direct at points, not just in mentioning things like "straw men" arguments, but also in highlighting that Viacom's argument in one part "makes no sense." That claim was in reference to Viacom's weak attempt to claim that because YouTube very briefly had a feature to let users flag "infringing" works, but then took it away, that meant that YouTube had willful blindness. YouTube debunks that, pointing out that they wouldn't have even created the feature if it wanted to ignore infringement. The reason they got rid of the feature was because it didn't work, because users don't know if a work is infringing or authorized. At another point in the filing, YouTube more or less mocks Viacom's attempt to reinterpret the Grokster standard, quoting Viacom's language (which is quite different that the Supreme Court's) and says "whatever that means...."

The filing also reiterates the key point that highlights the pointlessness of Viacom's claims that YouTube must have known what was infringing: which is the extensive efforts by Viacom to upload its own stuff but make it look like it was infringing. Combined with this is the fact that Viacom allowed plenty of clips it found on YouTube to remain... even if they were infringing. As we pointed out last time, Viacom had a huge and ever-changing book of "rules" as to what got left up and what got taken down (i.e., what was authorized and what was not). YouTube did not have this book. Suggesting that YouTube should know what to take down here makes little sense.

As for the supposed "smoking gun" emails that supporters of Viacom's position always point to, beyond the fact that many were taken totally out of context, none of them show that YouTube knew of any Viacom content that was infringing.

When Viacom filed its appeal, one of the Viacom supporters in our comments highlighted a part of Viacom's filing that focused on the Congressional record, suggesting that Congress meant to codify common law vicarious liability rules into the DMCA. However, YouTube's response points out what Viacom supporters would like to ignore: that the Congressional record being quoted was for a preliminary version of the DMCA, not the final one, and that Congress specifically decided not to follow through on those suggestions. Oops.

There's a lot more in the filing, but it systematically destroys each of Viacom's questionable claims. It won't come as a surprise to anyone that I find YouTube's arguments here persuasive, and Viacom's ridiculous. The district court saw it as well, and hopefully the Appeals Court will agree. That this is happening in the same circuit that decided the Tiffany/eBay case, which noted that eBay was protected from liability in trademark law -- where there aren't even the safe harbors that copyright law has -- would seem to bode well for YouTube. It suggests that YouTube's actions taking down content it knew was infringing clearly meet the standards necessary to be free from liability for videos uploaded by users. Either way, the filing is a fun read if you want to see a rather systematic takedown of Viacom's weak arguments.

Filed Under: copyright, dmca, red flags, youtube
Companies: google, viacom, youtube

Google Finally Gets Involved In Torrent Search Engine Lawsuit... But Just To Reject 'Red Flag' DMCA Violations

from the but-of-course dept

TorrentFreak is noting that Google has, perhaps for the first time, waded into any of the lawsuits concerning torrent search engines, filing an amicus brief in the ongoing IsoHunt appeal. In the past, other torrent search engines have been somewhat upset that Google has stayed quiet, noting that many of the arguments used against them could equally apply to Google. Google, of course, has stayed away because it goes to great lengths these days to avoid any appearance of "supporting piracy."

While TorrentFreak (and IsoHunt) seems surprised or disappointed by Google's actual amicus brief in the IsoHunt case, it's really not surprising. Google's participation here is entirely about the YouTube/Viacom lawsuit and appeal. What was really notable about the original IsoHunt ruling, was that it was the only real legal ruling that said that you could violate the DMCA even in absence of takedown notices. That is, it highlighted the idea of "red flag" awareness of infringement. This was the key issue in the YouTube/Viacom lawsuit. Google argues that as long as YouTube took down any content it received a takedown notice on, it was in compliance and protected by safe harbors. Viacom leaned heavily on the IsoHunt ruling, to claim that the DMCA doesn't just cover takedown notice responses, but also requires a response to "red flag" infringement.

However, Google knows that the IsoHunt ruling is basically the only legal precedent out there that reads the DMCA in this manner. So, from Google's perspective, dumping that reasoning is key. So its amicus brief still argues that IsoHunt is guilty of contributory infringement, a la the Grokster standard, but not because of red flag infringement. So, I don't find it that "unexpected" that Google got involved, or filed the brief that it did. It's focused entirely on its own lawsuit here, and hoping that the appeals court will take away the one serious case that Viacom has in its pocket for the YouTube appeal.

Filed Under: copyright, dmca, red flags, torrents
Companies: google, isohunt, mpaa, viacom, youtube

Even IP Lawyer Trade Group Thinks Viacom Is Wrong About Its DMCA Interpretation

from the expanding-the-law dept

As you might imagine, I rarely agree with the American Intellectual Property Law Association (AIPLA) on its positions. While there are plenty of "IP" lawyers who I know well and talk to frequently -- many of whom seem to agree with my position on things -- it's no secret that the belief that "IP is all good" and "more IP is better" tends to be a bit more common among such practitioners than the views in the other direction. In fact, if you asked me, I would have just assumed that the AIPLA was 100% behind Viacom in its lawsuit against Youtube/Google. So, consider me quite surprised that the main part of the AIPLA's amicus brief in Viacom's appeal of the YouTube case is actually siding with YouTube and saying that Viacom's argument (as we've said) goes way too far.

Specifically, the AIPLA points out that Viacom's belief that "general knowledge" of infringing content should disqualify safe harbors makes little sense, and is clearly not Congress' intention, as seen from the DMCA itself and the Congressional record:

AIPLA urges this Court to affirm the district court's holding that more than a generalized knowledge of infringement is required to deprive an Internet service provider ("ISP") of the protection of the safe harbor provisions of the Digital Millennium Copyright Act ("DMCA"). The district court correctly held that the DMCA requires either "actual knowledge" of specific instances of infringement, or awareness of "facts or circumstances" from which specific instances of infringing activity are apparent. Whether based on actual knowledge or awareness of facts or circumstances, the level of knowledge that is sufficient to strip the ISP of its protection under Section 512 of the Copyright Act, as amended by the DMCA is knowledge of specific instances of infringement. The district court's holding is consistent with the legislative history of the DMCA and relevant case law. AIPLA urges this Court to reject Viacom’s broad attempt to deprive Internet service providers of the benefits of the safe harbor provisions of the DMCA based on generalized knowledge that infringing activity is occurring on a site.

The AIPLA brief highlights how the Congressional record clearly shows that both the House and Senate said that "defective" DMCA notices need not be followed, and there should be no liability for not following such a notice. Yet, if Viacom's argument is correct, then a defective DMCA notice would still serve as "general knowledge" of infringement or a "red flag" that would require further investigation. Yet, both houses of Congress specifically rejected that position -- and with it Viacom's main argument.

There are two other parts of the AIPLA's brief which actually suggest points that ask the court to push back on two elements of the district court's summary judgment ruling, but those are smaller points compared to the "main event" question of whether or not general knowledge or specific knowledge are needed to trigger takedowns. It's surprising, but nice to see the AIPLA come out in favor of the only interpretation that makes sense to many of us: you can't be expected to police content if you haven't actually learned that it is, in fact, infringing.

Filed Under: actual knowledge, copyright, dmca, red flags
Companies: aipla, google, viacom, youtube

Perfect 10 Loses Again, As Court Says DMCA Notices Need To Be Properly Filed

from the red-flags? dept

Just last week we were talking about Perfect 10's lawsuit against Google in Canada, where we noted that in Perfect 10's own bragging press release, it effectively admits that its takedown filings were not properly filed. They admit that they just sent images to Google saying that it owned the images, without telling Google where they were actually located to take down. This was the same charge that Rapidshare recently made against Perfect 10, noting that the company seemed to purposely not want companies to take down their images, so that it could sue.

Thankfully, in the US version of the lawsuit (in which Perfect 10 seems to lose over and over and over again), the judge noted this failure by Perfect 10 to properly file DMCA takedown notices and dismissed large parts of the lawsuit. Similar to what Perfect 10 bragged about and what Rapidshare claimed, it appears that Perfect 10's "notices" were hardly informative. It also seems to have gone out of its way to make it difficult for Google to quickly respond -- including sending the notices to the wrong email address. As EFF notes:

For example, many of its "notices" consisted of a cover letter, a spreadsheet with URLs (many of which linked only to a top-level URL for a website, as opposed to a specific infringing URL) and a hard drive or DVD containing Perfect 10's electronic files of its photos. Not good enough, said the court -- the information required by the DMCA must be contained in a single written communication; forcing a service provider to cobble together adequate notice from a variety of sources is just too burdensome.

While this is entertaining in that it's the latest in a long line of legal smackdowns against Perfect 10 and its questionable litigation strategy, this ruling could be important in a variety of other cases as well. One of the key issues being fought about in a series of cases is what constitutes "knowledge" for a service provider, requiring it to take action under the DMCA. In both the Veoh/Universal Music case and the YouTube/Viacom case, judges found that the knowledge had to come from specific DMCA takedown notices, that indicated where the specific infringing works were. However, in the IsoHunt case, a judge went in a different direction, claiming that "red flag" knowledge was enough. That is, if there was enough information out there to raise a "red flag," then the service provider needed to take action.

Now, we've long argued that such "red flag" knowledge is somewhat meaningless. If I know that lots of people are using a tool for infringement, but don't know which specific works are infringing, how can I be expected to do anything specific? Since there's no way for the service provider to pinpoint which works are infringing -- even if they know that many works likely are infringing -- then how can the service providers act in a way that doesn't create massive collateral damage for legitimate communication?

But this ruling, again, effectively is a vote against the concept of "red flag knowledge," since you could make the argument (and, Perfect 10 did) that even in the absence of a complying DMCA takedown notice, sending over its mess of information could constitute a red flag. But, as the judge properly notes, that makes little sense. The ruling goes through the ridiculous hoops that Google would need to jump through in order to find and take down specific works, and notes that the DMCA clearly did not intend for that to happen.

Of course, this isn't the first time that Perfect 10 has lost on this exact argument. The CCbill case involved more or less the same questions about "red flag" knowledge, and Perfect 10 lost there. This ruling relies heavily on that one. But, we seem to keep racking up rulings that say that any "red flag" knowledge still requires specific notification of what is infringing -- with the IsoHunt ruling being the one exception. It makes you wonder if the IsoHunt ruling is on a collision course with all of these others.

Filed Under: dmca, red flags
Companies: google, perfect 10