Online Content Sharing
by Robert Parker
Part I: Safe Harbor
According to the 2015 publication of an annual study by Cisco, within three years, 85 percent of American internet consumption will be watching video, nearly all of which is protected by copyright. While most online video is licensed from the copyright holders, some of the most popular websites online are not content creators or licensees, but “online service providers” like YouTube, which allow users to post their own material and share with the world.
If that seems to you like a recipe for copyright infringement, the early years of the internet would indicate that you are correct. Although many service providers tried to take down infringing material as best as they could, as the process of posting and accessing online videos became more automated, it became very clear that the service providers were just not going to be able to keep up. At the same time, the anonymity of the internet made it difficult to stop people from posting online content, leaving copyright owners with little resource to keep their content off the internet.
This put the legal system at a crossroads: if there’s nothing we can do to stop people from posting and accessing infringing content online, how do we protect copyright owners’ interests without crippling the online sharing community? Keep in mind that the websites which hosted this plethora of infringing material were the same websites that the copyright owners themselves frequently used to proliferate their material, so most of them didn’t want to see sites like YouTube brought down any more than the rest of the online community.
Ultimately, the first step to finding a compromise was the Digital Millennium Copyright Act, or DMCA. Among other things, the DMCA and its subsequent additions created the service provider safe harbor clause, commonly simply called the “safe harbor” or “section 512” (because its location in the U.S. Code is 17 U.S.C. § 512). For the next few weeks, we will be posting some information on how the safe harbor works and how it affects various people. This week, we will provide some basic background on the safe harbor provision. In Part II, we will explain one of the most important elements of the safe harbor clause for content providers: takedown notices. Part III will address some of the drawbacks of the safe harbor and discuss a little about the future of safe harbor provisions.
What is the Safe Harbor Provision?
An important thing to remember about the safe harbor provision is that it does not grant service providers immunity from infringement claims. In the past, service providers argued that only the uploaders themselves should be held liable for any infringing content, particularly when they have automated upload processes. However, the law on infringement is that anyone who has control over the infringing material may be liable for any infringement that occurs, especially when that party is benefitting in some way from the infringement, such as receiving revenue for advertisements which appear next to the infringing content. That default rule still holds true today; if a company like YouTube knows, or has reason to know, that a video is posted on its website which infringes a copyright, the company has a part in the unlawful copying of that video.
The safe harbor provision creates a list of precautions that every service provider is expected to take in order to keep infringing material off its website. So long as the service provider adheres to all of those precautions, it cannot be held liable for any infringing material that the system fails to catch. The safe harbor does not protect the individuals who post infringing content (nor does a statement saying ‘no infringement intended’).
The required precautions are:
The service provider must expeditiously remove any content from the site that it knows or has reason to know is infringing on another person’s copyright
The service provider cannot ignore red flags that infringing content might have been uploaded
The service provider cannot receive any direct financial benefit from infringing material
The service provider must provide an easy way to provide it with takedown notices by owners of infringing copyrights (this goes hand-in-hand with the requirement to actually take that content down)
The service provider must have registered a designated copyright agent with the U.S. Coypright Office and post contact information for that designated agent on the service provider's website.
As you can see, taking advantage of the safe harbor provision requires more than simply signing up and sitting back—it requires constant vigilance by service providers, and as we will see in part III, service providers who drag their feet on any of the requirements or push back too much against claims of infringement take a risk of opening themselves to total liability for any infringement that they know or financially benefit from.
Part II: Takedown Notices
As we’ve noted, one of the challenges for copyright owners who find their material posted online is finding the individuals who posted it. To combat this problem, the safe harbor provision requires service providers to have a system in place by which copyright owners can request that allegedly infringing material be taken down. It is up to service providers to determine the best way to accept such notices, bearing in mind that removal of content must be sufficiently expeditious. As an example, YouTube allows individuals to submit a formal DMCA notice, fill out an online form, or sign up for Content ID, which automatically scours the YouTube library and tries to find claims for you.
For a takedown notice to trigger the service provider’s safe harbor requirement, it must meet several criteria:
the claimant (i.e. complaining party) must identify the copyrighted work the offending post allegedly infringes
the claimant must identify the online location of the offending material
the claimant must provide accurate contact information
the claimant must attest a good faith belief that the offending post is infringing
the claimant must confirm that it either actually owns the copyrighted work or is authorized to act on the owner’s behalf
the claimant must confirm that it has an exclusive right to the offending post
Although many of these requirements are primarily intended to ensure that the complaining party is not filing a false or illegitimate complaint, you will notice that it does not actually require the service provider to actually confirm that the work is, in fact, an infringement.
Safeguards for the Original Poster
Once the service provider has received a valid takedown notice, it is essentially on notice that it must take the material down or risk loss of safe harbor. It is then essentially the online poster’s burden to show that he is not infringing on the claimant’s work. However, there are safeguards in place which try to preserve balance in the law. Anyone who files a false or frivolous claim takes on substantial legal risk, and for non-meritorious claims generally, the poster has an opportunity to submit a counter-notice, which claims that his work is valid.
There are also several safeguards in the safe harbor statute itself to protect the poster: limitations on the claimant, legal remedies for the poster against frivolous claims, and a counter-notice process.
First, as noted above, the claimant must affirm his or her statement under penalty of perjury. Most of us notice the “under penalty of perjury” wording on almost every contract we sign and may not give it a second thought, but it is a powerful deterrent against knowingly false claims. It is an important safeguard against certain kinds of frivolous or fraudulent takedown notices.
Second, in addition to possible criminal liability for perjury, by submitting a takedown notice, the claimant exposes himself to a lawsuit by the poster. Here’s an example: say the poster puts up a variety of memes, which feature pictures of a prominent public figure and critical statements about that politician. The public figure then files takedown notices, claiming that the pictures used are protected by copyright. His claim isn’t perjury, because the images really are protected by copyright, but the postings may be protected by fair use or freedom of press. If the poster suffers any financial damage as a result of his legally permissible posts being taken down—say, because the memes function as advertisements to the poster’s website—the poster can now sue the claimant for damages.
This is why it’s so important that the takedown notice includes the claimant’s accurate contact information; it allows for the government (in case of perjury) or the poster (in case of a lawsuit) to actually find the claimant and litigate the issue.
Counter-Notices
The third and most important safeguard against false or frivolous takedown notices is that the DCMA provides posters with the opportunity to file a counter-notice. The counter-notice must include:
the location of the offending material prior to its removal
the poster’s accurate contact information
a statement that the material is not infringing the claimant’s copyright
consent by the poster to be sued in a U.S. jurisdiction
As with the notice, the information in the counter-notice is signed and affirmed under penalty of perjury. It is then up to the service provider to decide whether to allow the material to stay online. If the service provider decides that the material can be restored, it must give the claimant 10 days to respond, and if the claimant does not respond, the material can go back online.
It may seem at first blush that this counter-notice completely negates the takedown process, but in practice, it is very rarely used. For one thing, in cases where the material is actually infringing, the poster is unlikely to try to keep it up (and many service providers will include negative consequences or even banishment for a poster who frivolously defends his infringing posts). Counter-notices contain the same deterrents against frivolous use as takedown notices: the poster must provide his contact information, attest under penalty of perjury, and not only must he consent to be sued, but must do so in a U.S. jurisdiction, which is not required for the claimant. Unless the poster is particularly invested in that particular post and service provider, the poster is much more likely to either write off the loss or go to another service provider and post the offending material there.
Part III: Is It Working?
The Digital Millennium Copyright Act (DMCA), in particular the section 512 safe harbor provision, fundamentally changed how content is shared on the internet. As we discussed in Part I, the safe harbor attempts to incentivize service providers places where content can freely be shared online without destroying the copyright holder’s hopes of protecting his intellectual property. The takedown notice and counter-notice procedures we discussed in Part II attempt to provide an effective but balanced way to enforce copyright protections online without subjecting the process to abuse. This week, we’ll address the question: is it actually working?
The short answer is that a change as major as the DCMA is always going to be a work in progress for several years after its passage. The DCMA itself is evidence that the Copyright Act, over a century old, continues to adjust, improve, and adapt to this day. However, there are still plenty of challenges to address. This week, we will briefly address a few of the problems with the current system that have been discovered.
Online “Mini-Courts” and Service Provider Innovations
The DCMA has provided copyright holders with powerful and affordable tools to protect their interests while still allowing service providers to provide a medium for people to share their content online. However, by making it cheaper and easier for rightful holders to enforce their copyrights, the DMCA has also provided channels for “strategic” takedown notices as well. (A “strategic” claim is one where the claimant doesn’t necessarily think there’s a valid claim, but simply the accusation will chill others from posting online).
Additionally, whereas lawsuits are handled by licensed attorneys and judges, the safe harbor provision puts the burden on the content providers themselves, who may not be as well-versed in the law, to make judgment calls as to whether claims are legitimate. And because there is no penalty for a content provider to take down online content even if it does not violate copyright law (it is, after all, the service provider’s right to decide what goes on its own website), the penalty for allowing actually violative content to remain online is a total loss of safe harbor. Some argue that this puts a thumb on the scale in favor of allowing all infringement claims, whether legitimate or frivolous, to be enforced.
As discussed in Part II, several safeguards are put in place to counter this danger and ensure that the poster of legal material has legal recourse to protect her content. However, one of the main outcomes of the DMCA is the creation of a system by which disputes are resolved without court involvement. Although the safe harbor provision sets out the procedural requirements of such a process, it is the service provider, not the courts, that enforces takedown notices. Larger sites like YouTube have put fairly robust procedures in place to handle takedown notices, creating a sort of “mini-court” within the confines of the service provider and the website itself. Within these mini-courts, these service providers provide some innovative solutions beyond what is contemplated by the safe harbor rules themselves.
For instance, YouTube recognizes the “whack-a-mole” problem that many copyright holders have to deal with, where an offending post is taken down only to be put right back up. YouTube attempts to address this problem in several ways. First, it has a “copyright strike” system, so a repeat offender’s entire account is taken down. Although this does not prevent the infringer from creating a new account, doing so can be enough of a hassle that the infringer will simply stop trying. Second, YouTube offers a ContentID service, where YouTube scours its own videos and detects suspected infringing material, more effectively and easily than if the copyright holder attempted to do so herself. Both of these processes have ways for the poster of allegedly infringing content to protect himself, including a “Copyright School,” which a poster can take to rehabilitate his record after receiving strikes, and of course, the counter-notice process discussed in Part II.
The Monetization Problem
Another issue that YouTube attempts to address is the fact that, by the time an infringing post is discovered, the damage may have already been done. In some cases, the infringing post may actually have become more popular than a claimant’s original work; in “scalping” the claimant’s work, the infringer may have to take down the material, but he still keeps the profits from whatever views (and accompanying ad revenue) he has received, and while the infringer loses any popularity the infringing post has received once it is taken down, the copyright holder has also lost that popularity. In some cases, the copyright holder might not want the post to be taken down; he just doesn’t want an infringer to be profiting off of it. YouTube attempts to address this problem by giving the claimant a choice: you can demand that the entire video be taken down, you can demand that only your copyrighted material be removed (for instance, if you own the rights to the music, you can demand that the video be played without audio), or you can allow that the video remain online, but that any monetization be given to you, the claimant, rather than the poster. If the poster did not monetize the video (by not including ads), the claimant can demand that monetization be added to the video, with the proceeds going to the claimant.
The monetization-shifting solution is a win-win in certain situations: the audience does not lose access to the copyrighted content, the poster does not (necessarily) get a strike, and the claimant profits from her copyrighted material. On the other hand, it provides a financial incentive for claimants to file notices that may not be valid. This has led to the rise of a handful of “ContentID Trolls,” in the same category as patent and other IP trolls. ContentID trolls falsely claim ownership of popularly-used material, or file claims of material they own but that others use legally, so that they can claim the ad revenues from ads which use that material. The poster can still file a counter-notice, but the stakes are high: YouTube will ask the claimant whether your defense is valid, and if YouTube decides that the claim is still valid, the poster who files the counter-notice will get a copyright strike, and while the claim is pending, YouTube may withhold all revenue from that poster, even on other videos.
YouTube has made some adjustments to its process to reduce the incidence of trolls, including holding monetization in escrow during copyright disputes rather than diverting it to the claimant, imposing punishments for repeated bad faith claims, and even offering to pay the legal bills of posters who have valid fair use claims for their content. However, all of this requires action by the poster to fight back, and most are still likely to simply allow the claimant to monetize the one claimed video. It is fair to point out that, by allowing videos for which there are claims to remain online, YouTube continues to receive its cut of the ad revenue as well, so there is a financial incentive for YouTube to continue its monetization-shifting model despite its problems.
The Doxxing Problem
One of the most celebrated features of the internet is also one of copyright holders’ biggest challenges: its anonymity. The internet provides a place where people can safely share content while remaining anonymous if they so choose. There are, without question, some downsides to anonymity, but it has also enabled people from all around the world to have their voices heard when they might otherwise fear retaliation or other consequences. As we noted in Part I, one of the consequences they avoid online, however, is liability for copyright infringement. The DCMA’s solution to this problem is to require the service providers, who are not anonymous, to take down infringing content, and if the poster wants to avoid this, she must provide the service provider and the claimant with her personal contact information.
A certain group of claimants have learned how to leverage this requirement. If an individual dislikes content online, he can file a takedown notice, even if there is no actual infringement. The poster then must make a choice: she can take the material down, or she can file a counter-notice which includes her personal information, which the service provider must give to the claimant. The ostensible purpose of allowing the claimant to sue the poster for infringement in court, but some claimants have used this process in order to “dox” the poster (a respelling of “docs,” as in revealing personal documents)—break the poster’s anonymity and publicize her personal information for others to see.
The Extrajudicial Problem
There is one other drawback to the DMCA safe harbor rules that has inadvertently exposed it to abuse. One of the problems with online content is that much of it comes from abroad. Thus, even if a copyright holder knows who is posting infringing content, if the infringer is outside of U.S. jurisdiction, there may be little that the copyright holder can do. The safe harbor provision attempts to solve this problem by putting the onus on the service provider, who is subject to U.S. jurisdiction (if it weren’t, it wouldn’t need the safe harbor in the first place), to take down any infringing material. Then, if the poster wants to fight the claim, in filing the counter-notice he must agree to subject himself to U.S. jurisdiction.
However, there is no equivalent requirement for claimants. A party can file a takedown notice from anywhere in the world, and the service provider still has to oblige. If the poster files a counter-notice and the claimant chooses to keep the claim alive, the service provider is still likely to keep the material offline (or shift the revenue to the claimant) rather than risk loss of safe harbor. If the poster still wants to pursue her counterclaim, her next step is to sue for a declaratory judgment that the material is not infringing; at this point, the claimant is unlikely to oppose, and the service provider is likely to restore the material. However, getting a declaratory judgment is expensive. And because the claimant is not subject to U.S. jurisdiction, the threat of perjury or liability for damages no longer has any deterrent effect to discourage frivolous claims.
