We regret to inform you that the internet is on red alert once again. On Wednesday, the EU’s Legislative Committee voted to adopt sweeping measures that will upend the web in every way that we know it. Memes, news, Wikipedia, art, privacy, and the creative side of fandom are all at risk of being destroyed or kneecapped.
By the time Americans woke up on Wednesday, the Legislative Committee had voted on the final form of the EU Copyright Directive—the first major update to European copyright law since 2001. Much of what’s in the legislation has been met with approval, but Article 11 and Article 13 are considered disastrous by some of the foremost tech experts in the world.
Explaining what’s wrong with these two points of the legislation in detail is difficult because the articles themselves are so vague. That’s the primary issue for critics. Both articles make unprecedented demands on anyone operating a popular website to monitor copyrighted material and to pay fees to news organizations when linking out to their articles. Defenders of the plan say that critics are exaggerating because of assumptions they’re making about how the legislation will be implemented. Critics, like one of the “fathers of the internet,” Vint Cerf, and the inventor of the World Wide Web, Tim Berners-Lee, say the risks outweigh the benefits. Who are you going to believe?
Let’s take a look at what’s at stake with these new forthcoming regulations:
This section of the directive will completely reconfigure websites’ responsibilities when it comes to enforcing copyrights. Until now, the so-called Ecommerce Directive has given online platforms broad protection from being subject to copyright penalties when they simply acted as a conduit for user uploads. It’s very similar to the laws in the U.S. that exempt YouTube from penalties as long as its making its best effort to take down infringing material when it’s reported. YouTube uses an automated content recognition system combined with an army of human beings to review the material users’ upload. It costs the company millions of dollars to do this. Critics of Article 13 say that every popular platform—estimated to mean the top 20 percent—that allows users to post text, sounds, code, still or moving images will need one of these systems.
Last week, 70 of the most influential people in the field of technology signed a letter opposing Article 13. Pioneers like Cerf and Berners-Lee were joined by experts in virtually every facet of the online world to say that the legislation would harm freedom of speech, education, expression, and small businesses while giving major platforms that already heavily monitor their service a distinct advantage.
Activist, author, and special advisor for the Electronic Freedom Foundation, Cory Doctorow has written extensively about the potential implications of Article 13 since it became a crisis in the last few months. He told Gizmodo over the phone that as it’s written, the legislation will cost “hundreds of millions” of dollars in penalties for platforms that can’t handle monitoring, and he’s confident that companies like Google and Facebook will be the only ones that can survive.
The big question around Article 13 is its vague requirement that websites use “appropriate” measures to prevent copyrighted material from ever appearing on their service. It suggests “effective content recognition technologies” be used several times without explaining what that means, how it would work, how claims would be filed, or anything practical. For critics like Doctorow, the natural conclusion is that big platforms will use their own system and some sort of centralized system will be required for the rest. Because there’s no outline of how such a system would work, there are no penalties for people who falsely claim ownership over the content. In the event that someone uploads a claim over the complete works of Shakespeare—which is in the public domain—a platform would have to individually decide if that claim is worth taking a risk and allowing someone else to quote a sonnet by the bard. If the platform doesn’t want to take the risk, someone fighting a copyright claim would have to go to court.
As we see all the time, algorithms by the richest companies in the world are terrible at doing their jobs. This week, we saw YouTube blocking educational videos from MIT and the Blender Foundation because they were erroneously flagged by its piracy filters. In the past, we’ve seen bullshit piracy claims over white noise and birds chirping.
Also, what’s possibly the most important problem with Article 13 is that it makes no exceptions for fair use, a foundation of the internet an essential caveat in the law that allows people to remix copyrighted works.
Article 11 has been variously called the link tax or the snippet tax. Designed to mitigate the power over publishers that Google and Facebook have amassed in the last decade, it codifies a new copyright rule for linking to news organizations and quoting text from their stories. Online platforms will have to pay for a license to link out to news publishers, and this will theoretically help support organizations that are vital for public information and drive users to their homepages.
That all sounds decent in principle, but Article 11 doesn’t bother to even define what constitutes a link. Details will be left to the 28 individual countries in the EU to figure that out. That opens the door for political abuse of how news is spread in each country, and it will likely have the opposite of its intended effect.
Google can afford a license, there’s no guarantee smaller organizations can. Member of European Parliament Julia Reda is firmly opposed to Article 11 and 13. She recently wrote on her website: “Instead of one Europe-wide law, we’d have 28, with the most extreme becoming the de-facto standard: To avoid being sued, international internet platforms would be motivated to comply with the strictest version implemented by any member state.” In response to her MEP counterpart Alex Voss’s defense of Article 11, Reda gave The Next Web an illustration of how the differences between countries could play out:
The sentence ‘Angela Merkel meets Theresa May,’ which could be a headline of a news article, cannot be protected by copyright, because it is a mere statement of fact and not an original creation. Mr. Voss said repeatedly that he wants these purely factual statements to be covered by Article 11, that the protection granted to press publishers will therefore be much broader than even what the journalists themselves get.
Reda also pointed out that egregious sampling or wholesale theft of news content is already illegal under current copyright law. There’s no reason to believe that Facebook with its fancy link license will ever face penalties for users posting an entire article on their wall. But when Facebook decides it doesn’t like your particular political point of view, it’ll be a lot harder for you to start a small platform and express it.
The consequences of Article 11 and Article 13 remain a matter of speculation, but the nature of the legislation—both its design and its vagueness that makes it ripe for abuse—make it all but inevitable that they will leave the internet torn and tattered in its wake. Here are some likely victims.
Even if you think that people who pirate music should be executed and all news organizations are the devil, you probably like memes. Well, whoever took a picture of that one guy looking at that one girl instead of the other girl, will be having a field day running around filing complaints against any platform that uses it without permission. Just kidding, that guy sold the photo to iStock, a subsidiary of photo-licensing giant Getty Images. No fair use means you’ll have to go shoot your own photo to caption and make it clear that anyone is allowed to further caption it in the pursuit of creating a meme.
Remixes and mashups are fucked. Any artist that relies on fair use to make transformative works is fucked. And the Metallica’s of the world who love running around policing where their work will have platforms, and their grunts making sure to pull down that birthday party video of you and your friends just trying to have a good time while some song was on the radio. Are you wearing a Rick and Morty shirt in that perfect profile pic? Sorry, the stupid algorithm flagged it, and now it’s gone.
Aside from the potential of individual countries in the EU to decide what is and isn’t news, copyright claims could be used to suppress material for political purposes. Doctorow gave us the example of a politically sensitive video uploaded to a platform just days before an election. Let’s call it the pee tape. If the target of the pee tape were to know it was about to be released, it could be uploaded to a content monitoring platform with a copyright claim lying in wait. The censorship filters would catch it before it was seen by the public, and the election could come and go while a legal fight plays out behind the scenes.
There’s also the issue of surveillance. We already accept that companies like Facebook hire people to comb through our shit while trying to identify infringing content. The EU is trying to force many more companies to deputize a bunch of sleuths, human and algorithmic, expand this shadow surveillance state that monitors everything we post on these platforms. As Doctorow put it to us, “Any kind of censorship in the modern age is surveillance.”
There are numerous other bad implications that have been flagged by activists, academics, human rights groups, and online businesses. We didn’t even discuss Article 3, which has artificial intelligence startups sweating bullets. I’d urge you to call your representatives, but we mostly serve an audience in the U.S, where we largely lack the power to force lawmakers’ hands. We don’t know if this will have the ripple effects we’ve seen with the GDPR privacy rules that are slowly being picked up as the global standard. But Doctorow told us that implementation of the copyright directive adopted on Wednesday will be just like GDPR’s chaotic rollout last month. “Everyone is going to forget about it” during the waiting period before the law is implemented, “and in two years they’re going to wake up and say holy shit!”
Update 8:50am, June 20: On Wednesday, the European Parliament’s Committee on Legal Affairs voted to move forward with its new Copyright Directive. Article 11 narrowly passed with a vote of 13 to 12 and Article 13 passed with a wider margin of 15 to 10. The legislation will now have to undergo a plenary vote by the full Parliament. There’s no set timeline on when that will occur but its expected sometime between December of this year and the beginning of 2019. While it was considered unlikely to be a subject of much debate after reaching a plenary vote, Joe McNamee, executive director of digital rights association EDRi, told The Verge that public outcry may have changed that. “I was told that the volume of calls, emails, and texts everyone in the Parliament has been getting has led people not in the [JURI] committee to start getting worried,” he said.