The Copyright Directive Is SOPA Part 2, and It’s Coming for Your Internet

Last week, the European Parliament voted 348 to 274 to pass the Copyright Directive. Unless something truly extraordinary happens during the upcoming meeting of the European Council — think of it as the Senate of the EU, where the governments of Member States are represented — draconian and highly disruptive new rules on content licensing and monitoring will become EU law.

In this blog post, I explain the two key elements of the the Copyright Directive, the link tax and upload filters requirements. I also explain why they are poised to change the internet as we know it, why you should be worried about similar policies coming to the United States, and what you can do to help keep the internet open for all.

The “Link Tax”

Article 15 (previously known as article 11) of the Copyright Directive mandates that virtually all internet platforms and websites, regardless of their size, that reproduce more than “individual words or very short extracts of a press publication” must secure a license from the linked media organization. As a baseline, this means that any website that offers snippets next to a link would require a license to do so. It’s unclear what “individual words or very short extracts” means in practice, and setting the definition will require costly, prolonged litigation.

It’s not even clear that the provision solves the problem it was designed to address. If European legislators intended to ensure that tech companies pay news organizations for providing snippets of their work, it is unclear that the “link tax” is the right remedy. A similar approach was already tried in Spain and Germany; Google News is no longer available in Spain, and in Germany, publishers eventually exempted Google from paying any fee. It may turn out that news organizations might find that tech platforms can live without them: After all, only 4 percent of Facebook’s content is news.

Upload filters

Article 17 (previously article 13) establishes that commercial online platforms “perform an act of communication to the public” when they give access to copyright-protected works uploaded by their users. Those sites and apps will have to make “best efforts” to preemptively purchase licenses for all copyrighted content that users might potentially upload. In practice, this could mean they must attempt to license all copyrighted content in the world.

In addition, almost all sites and apps will need to show that they are doing all they can to prevent users from uploading unauthorized copies. In practice, this means that virtually all companies that allow users to upload content will have to deploy upload filters: Otherwise, they risk being directly liable for their users’ copyright infringement.

The category of sites and services exempted from this mandate is absurdly narrow: It excludes only those that are both (a) less than three years old, and (b) have annual revenues of less than ten million euros, and (c) have fewer than 5 million monthly unique visitors. For example, an online web forum for a local community that’s more than three years old will face the same licensing and filtering obligations as Google and Facebook, even if it makes no money and has less than a thousand monthly visitors.

Here, European policymakers are being enormously and unrealistically techno-optimistic. Content ID — an upload filtering technology that took all the money and know-how of Google to develop, and one that smaller players will most likely have to license — is still incapable of understanding the context in which content is presented. For example, a German professor failed to upload public domain classical music recordings to YouTube because the system doesn’t understand the intricacies of copyright law.

Upload filters create a chilling effect. Most users will not challenge upload filters when they are blocked from uploading, if the system tells them that the reason for denial is possible copyright infringement. The average person is fortunate enough to not have to know the intricacies of copyright law and their rights to use copyrighted content legally under some circumstances. For good reason, our European allies refer to the upload filter provisions of the Directive as a “censorship machine.”

A natural experiment that might be coming home

It will take years before we see and understand the full extent and implications of the Copyright Directive. Unlike the General Data Protection Regulation (GDPR, Europe’s privacy law), which by virtue of being a Regulation is of binding and uniform application across the EU, the Copyright Directive must be transposed into national law by each EU Member State by 2021. In practice, this means that there are going to be 27 different flavors of the Directive across Europe. In addition to this geographical and cultural complexity, given that the new law is unclear in some key provisions as exposed above, we can almost certainly expect years of litigation before the European Court of Justice settles on the material meaning of the Directive.

In an ideal world, the U.S. should take the European experience with the Copyright Directive as a massive natural experiment with content licensing and filtering technologies; there’s going to be much to learn from the European experience in what to do, and what not to do. We can’t draw any meaningful conclusions when the experiment has just been announced. That’s why we urge the Copyright Office to delay the Section 512 Study, a public study and report to evaluate the impact and effectiveness of the safe harbor provisions contained in section 512 of the Digital Millennium Copyright Act (DMCA), until we know more about the implementation and litigation of the Directive.

However, the same actors that pushed for the Copyright Directive in Europe are advocating for similar policies in the U.S.. In a recent op-ed published in the New York Times that sounds the call for a “link tax,” the president of the trade association of news publishers wonders why “Facebook and Google flatly refuse to pay for news even though they license many other types of content” and asks “[w]hy are the platforms so unwilling to pay news publishers for access to the quality journalism that users need and value?” Regarding upload filters, don’t forget that rightshodlers’ associations such as the Motion Picture Association of America (MPAA) have been long asking for “automated detection and removal of infringing content” mechanisms in online platforms.

With the fight against the Stop Online Piracy Act (SOPA), the digital rights community demonstrated that it could stop legacy entertainment and media industries power grabs over the open internet. That fight is coming back. In the EU, the legacy entertainment industry has won, and the open internet will suffer. Are you going to allow them to take your internet too? Tell your members of Congress that you care about internet freedom and that you expect them to fight back against all proposals that mimic SOPA or the Copyright Directive.