Last week, the Council of the European Union — the EU body that represents the executive governments of its member states — signed off on the new EU Directive on Copyright in the Digital Single Market — the first significant update to EU copyright rules in almost two decades. That was the last Europe-level hurdle for a process that lasted nearly three years. (The European Parliament had given its final approval to the directive late last month.) In the weeks leading up to the final votes, artists like film producer and director Pedro Almodovar, singer Debbie Harry, and ABBA’s Benny Andersson, as well as independent music labels and European publisher associations, had endorsed the reform, citing the need to adapt copyright rules to the contemporary digital world. The landmark changes are designed to safeguard copyright-protected content on the internet to ensure that copyright holders like publishers, broadcasters, and artists are paid their fair share when their work appears online.

At least, that’s one way to look at it. 
Some of the world’s top human rights, technical, and copyright experts and organizations — and some members of European Parliament themselves — have fiercely criticized the reform, calling it a “censorship machine” that would restrict free speech, put new burdens on platforms that would be virtually impossible to comply with, and eventually mark the end of the internet as we know it. A Change.org petition opposing the reform gathered more signatures than any other in the site’s history (currently more than 5.2 million).

Normally, this might be a technical affair watched only by a bunch of copyright geeks and industry insiders. This time, however, the overhaul has drawn huge amounts of attention and controversy, with some seeing it as pitting the creative industries against internet activists.

Why the outcry? Most of the outrage has been over two provisions of the directive: Article 11 and Article 13. (Confusingly, the directive’s final version changed Article 11 into Article 15 and Article 13 into Article 17. Since the months of debate over these measures uses the old numbers, we will here too.) The first, called the “link tax” by opponents, gives publishers the right to be paid when online platforms and aggregators like Facebook and Google News share even small portions of their stories.

The second, dubbed by critics the “upload filter,” is a substantial reworking of how copyright works on the internet. It compels content-sharing services “storing and giving access to large amounts of works and other subject-matter uploaded by their users” to license copyrighted material from the rights holders. Currently, in cases of copyright violation, right holders have to submit a “takedown notice” in order for the services to pull the concerned content or else face the threat of litigation. The path is meant to go: copyright infringement goes online → copyright holder sends a takedown notice → infringing content is removed or platform faces potential lawsuit.

Article 13 abolishes that middle step in the process, eliminating the need for rights holders to search the web for infractions. Instead, it makes platforms liable for copyrighted material uploaded by users — meaning they can be sued without even a takedown notice. In other words: copyright infringement goes online → platform faces potential lawsuit right away.

For proponents, both measures are aimed at leveling the main disparity at the heart of today’s internet: the modern, advertising-based web has let big tech companies earn massive amounts of money providing access to content generated by other people, while all that remains for those who created the material (music, literature, journalism) is crumbs. (The Council’s vote last week included 19 member countries voting in favor, 6 against, and 3 abstaining.)

Critics of the reform, however, say it’s filled with blind spots. Julia Reda, a German member of the European Parliament and one of the reform’s most outspoken critics, argued the link tax might drastically chill the average person’s ability to post and share news stories. (In fact, the final text adopted exempts individuals, saying that the new rights granted to publishers don’t apply to “private or non-commercial uses of press publications by individual users.”) Google argued during the debate over the changes that they could force the shutdown of Google News in Europe and leave blank results for news searches. (Google had previously shut down Google News in Spain when that country passed a similar law. The final version of the EU directive states that the new protections “shall not apply to acts of hyperlinking” — though it’s unclear if that means the presence of a link alone eliminates the protection for excerpts or only that the link itself doesn’t count as an excerpt.)

It’s also still unclear who fits in the “information society service provider” category — as well as what sort of behavior. For instance: Are blogs or RSS feeds that aggregate newspaper headlines, much in the way Google News does, covered by the new rules? What about individually operated Facebook pages with large followings? Does a GIF or a meme built off a news video count? The EU says “the ‘meme,’ the ‘gif,’ the ‘snippet’ are now more protected than ever before” and that excerpting “for purposes of quotation, criticism, review, caricature, parody or pastiche” is protected — but the reform leaves it up to member countries to determine many of these boundaries, and some aren’t building a sterling record on press freedom and free speech issues.

There’s also ambiguity about what counts as too much of a story to excerpt. Under the scheme, a text containing more than a “very short extract” is covered by the overhauled copyright rules, and must be licensed and by the platforms that use it. But the EU has also left it to each country to decide what counts as a “very short extract.” The directive’s final wording states that individual words and “very short extracts” can’t be taxed, which would seem to keep headlines out of that category. In any event, the publisher protection expires on the first January 1 to come two years after a story’s publication.

Article 13 raises separate issues. One example: If YouTube wishes to let a user upload a video of themselves featuring songs by Drake or Beyoncé, it must ensure that that user has received permission from the rights holders of those songs. The implementation of such a measure could prove extremely challenging. YouTube uses a system called Content ID to enforce copyright, but it certainly doesn’t catch everything. (YouTube would look very different if it did.) Google argues that tracking down rights holders is complex; it says it’s spent more than $100 million on ContentID, and argues that tighter copyright rules would drastically increase that cost.

The reform’s final version, however, says that content-sharing platforms can host content for which they were unable to obtain permission — so long as they have:

— made their “best efforts” to secure permission
— made, “in accordance with high industry standards of professional diligence” and “expeditiously,” their best efforts to take down any copyrighted material rights holder inform them about
— worked to prevent future uploads of the copyrighted material they know about.

When judgments are made about whether an “online content-sharing service provider” has followed these rules, countries are supposed to consider the size and audience of the online service, and certain kinds of companies (those less than 3 years old, with annual turnover under 10 million euros, or with under 5 million monthly uniques) can sometimes get an easier set of standards to meet.

Still, opponents are wary of the provision because they contend that enforcing such a scheme would oblige sites like YouTube and Facebook to review any material users post or share and cross-check it with a database of protected material. Such a mechanism, they argue, would be ripe for abuse by copyright trolls, potentially leading to millions of both deliberate and erroneous blocks. (We saw one example of overzealous actions by rights holders last week, when the TV network Starz managed to get the tweets of journalists and others — even a law professor! — about a news story taken down improperly.)

One obvious parallel to these changes is the GDPR, the European data protection standards that went into effect last May, leaving the web with many more “I accept cookies” buttons to click. The first fines have begun to be issued for GDPR violations, including a $57 million fine for Google over problems with its Android onboarding process. But in general, the most extreme predictions of negative impact haven’t turned true, at least not yet.

Reda, the European MP most vocally opposed to the copyright changes, said the battle against them was “not in vain,” noting that public opposition had led to a number of important changes in the text of the directive. But she also argues that publishers “will without doubt advocate for the strictest possible interpretations of the Directive — civil society must resist, and we must do this in all member states.”

The activity indeed now moves to the member states. Each of the 28 countries in the EU now has two years to transpose it into its own national laws. Until we see how those laws shake out, especially in countries with struggles over press and internet freedom, both sides of the debate will likely have plenty of room to continue arguing their sides — that it marks a groundbreaking step toward a more balanced, fair internet, or that it will result in a set of legal ambiguities that threaten the freedom of the web.