Think the DMCA Is Outmoded? Complain to Congress, Not to the Courts
This is the second post on the recent Ninth Circuit opinion in the “Veoh case” (actually styled, UMG Recordings v. Shelter Capital). Last time, we focused on (marveled at, really) Universal’s surprising leading argument: that pretty much any website that makes user-uploaded content publicly accessible (with a probable exception for pure displayed text) is not covered by the DMCA safe harbor on grounds that the safe harbor covers only passive storage, not display, performance or internal copying. We also worried about the fate of user-created videos of cute kittens.
In this post, we’ll look at one of Universal’s two remaining arguments, which are much more conventional.
Recall that, to be eligible for the DMCA safe harbor governing user-provided content, the website must meet three general requirements: (1) a knowledge requirement (or, more accurately, an ignorance requirement); (2) a financial-benefit requirement (or, more accurately, a financial-non-benefit requirement); and (3) a notice-and-takedown requirement (which is intertwined with the first requirement*). Universal attacked both the knowledge and financial-benefit requirements.
* As we’ll see next time, it turns out all three requirements are intertwined with each other.
This Mess We’re In
The main policy point is: who should be primarily responsible for policing infringement on user-directed websites, the website operators or the rights holders? Based on the plain language of the DMCA safe-harbor provisions and Congressional intent, courts have consistently held that that burden falls mainly on the backs of the rights holders. The DMCA includes a provision sparing the operators from having to investigate or police their own systems, and Congressional reports reasoned that rights holders would be best positioned to realize whether a use of their works is infringing or not.
All this made sense when the DMCA was passed in 1998. Although compression technology and connection speeds made the sharing of audio files possible–with difficulty–the sharing of audiovisual works was pretty much unthinkable. It just didn’t seem like it would be that much work for rights holders.
That was then. And, as Universal complained at considerable length in its briefs, today, it’s very difficult for rights holders to keep up with all of the infringing uses of their works–even with advances in technology that search for infringing uses and automate the process of sending takedown notices. (To be fair, website operators also find it difficult to keep with the the volume of takedown notices, again, even with advances in technology, in part because so many takedown notices are defective or have nothing to do with copyright infringement.)
Alas for Universal, changed circumstances aren’t grounds for changing the way courts interpret the DMCA. It is up to Congress to re-evaluate the way the DMCA notice-and-takedown regime operates.
* And it’s not as though Congress has been ignoring the plight of rights holders in recent years.
Refusing to Foot the CCBill
Remember that there are two types of knowledge, as I’ve discussed previously. First, there is actual knowledge: you know for a fact that some content is infringing.* Weirdly, if you receive a defective take-down notice, you aren’t regarded as having actual knowledge.** Second, there is “red-flag” knowledge: you are aware of “facts and circumstances from which infringing activity is apparent.” At the same time, you have no affirmative duty to investigate, which seriously dents the scope of this type of knowledge.
* Sometimes infringement is obvious, but what if fair use is involved? Do you have to perform a legal analysis? And what if you guess wrong?
** This makes sense if you think about it, but it’s still very counter-intuitive.
Universal attacked along both prongs (without really distinguishing between the two) by arguing, essentially, that Veoh hosts a lot of music videos and everyone knows that music videos are copyrighted. Further, there was evidence that Veoh executives had a general sense that infringement was going on, and (most damning) even the press* knew that Veoh was hosting infringing content. At a minimum, Universal argued, there was so much infringement that Veoh had to investigate–and at that point, Veoh at least red-flag knowledge.
I know I promised to stop mocking Universal’s arguments, but this is some pretty weak sauce. There was already Ninth Circuit authority–the CCBill case–directly on point. Universal’s attempt to distinguish CCBill was breathtakingly superficial. Instead of attacking the controlling and relevant holding of CCBill, Universal argued that the lower court just didn’t do a very good job citing to CCBill. Even if that were so, that doesn’t make CCBill’s controlling holdings suddenly disappear–it’s still the law.
Universal spends most of its time complaining about the current state of affairs: how difficult it is to keep up with all of the infringing activity, the uselessness of the notice-and-takedown regime (even using the now-familiar refrain, “Whack-a-Mole”). These complaints are not without persuasive power, but unattached to anything like powerful legal authority*–or a compelling argument for overturning the controlling legal authority (i.e., CCBill)–it just comes across as whinging.
* Except, of course, for the Fung decision, the only decision finding the existence of red-flag knowledge. But as I discussed here, Fung’s behavior was so egregious that it’s not much use as authority. However bad you might think Veoh is, it’s nothing like Fung’s websites.
When Knowledge Isn’t Knowledge
But what about those instances of actual knowledge? Veoh’s CEO told a journalist that he wasn’t surprised that infringing activity occurred. Didn’t that put Veoh on notice of infringing activity?* Well, no. The court reasoned that Congress assumed that a certain amount of infringing activity would be going on–that’s why it created the safe harbors in the first place. By Universal’s logic, any website that permitted the uploading of content would have red-flag knowledge because everyone knows that infringement is going to happen. Again, that may not be fair, but that’s the balance that Congress decided to strike back in 1998.
* Universal’s reliance on news reports of infringing activity really seems like grasping at straws. Even assuming you could prove Veoh’s decision-makers read the reports, why in the world should Veoh believe them enough to launch an investigation? I think journalists would be genuinely surprised to learn they have power to put corporations on notice of their alleged misdeeds. (I’m also not convinced that hearsay doesn’t enter in this. Yes, the news reports are being offered to show that Veoh was on notice of what was written therein, but isn’t there an underlying assumption that the news reports were, in fact, true? Doesn’t that mean the reports are being offered for the truth of the matter asserted?)
More tantalizing was an email to Veoh from a disgruntled user, who was angry that Veoh had rejected his content because Veoh thought it was infringing.* He said that lots of other users were getting away with uploading infringing content, so why not him, too? More important, he actually identified one such user.
* Based on Veoh’s filtering technology. Take a moment to savor the irony here.
The court’s discussion here is easily overlooked, but it’s the closest Veoh comes to getting into trouble. First, the court held that the email doesn’t constitute actual knowledge because there’s no reason to believe the disgruntled user. But, second, the court held that it could be powerful evidence of red-flag knowledge. The problem is that Universal completely forgot about the second half of the knowledge requirement. You see, it’s not just actual or red-flag knowledge that matters. It’s that you have sufficient knowledge of infringement, and you do nothing about it (i.e., you fail to expeditiously remove the infringing content that you know about).
Thus, it wasn’t enough for Universal to simply point to the email. It also had to show that Veoh didn’t act on that email–which it didn’t do. It would not have been all that difficult to establish such a failure during discovery, so it’s reasonable to assume that there was no such evidence–and Universal was either overlooking this element of proof, or was trying to pull a fast one.*
* Another interesting email was from Disney CEO Michael Eisner (who was also a Veoh investor), commenting that certain Disney-owned properties were available on Veoh without authorization. The court held that this email was not evidence of red-flag knowledge (the opposite conclusion as from the disgruntled user’s email) because it came from the rights holder. As such, it constituted a defective take-down notice and was, therefore, a nullity. (It was too defective even to warrant a corrective response from Veoh.)
There may an important lesson for operators, though. Somewhat perversely, a communication from a non-rights-holder (e.g., an unsolicited email from a random dude) is much more dangerous than one from a rights holder, for two reasons. First, the random email might come through an unexpected channel–it might not have to come in through the DMCA agent. Second, the random email doesn’t have to comply with the notice-and-takedown requirements. Having said all that, it would still be a factual question whether the email rose to the level of red-flag knowledge. While we now have a better idea of what that level is, we still aren’t quite sure where exactly it is, or even how we are to determine it.
The bottom line, though, is that Universal’s arguments regarding the knowledge requirement have only made it harder for rights holders to attack this requirement.
Next time, we’ll look at the court’s rejection of Universal’s arguments about the financial-benefit requirement–and we’ll actually answer (pretty much definitively) one of the great unanswered questions about the DMCA safe harbor. I’ll also have some thoughts about how and whether the Veoh decision will affect the imminent decision from the Second Circuit in the YouTube case.
Thanks for reading!