LUDWIG VAN BEETHOVEN has been dead for nearly 200 years. The copyright on his music is long expired. But when Ulrich Kaiser, an academic at the University of Music and Performing Arts Munich, recently tried to upload a public-domain recording of his Fifth Symphony to YouTube, he was thwarted by Content ID, an automated copyright filter. Mr Kaiser tried again with recordings of music by Schubert, Puccini and Wagner. Despite being in the public domain, all were flagged for copyright violations by the algorithm.
YouTube built Content ID a decade ago, under pressure from copyright-holders worried that users were uploading commercial music and videos without permission. Ever since users have complained that the algorithm is too aggressive. Now YouTube and other big internet firms may be obliged by European law to employ similar methods there. On September 12th members of the European Parliament approved, by 438 votes to 226, a draft of a new copyright law designed to update the EU’s copyright legislation, which predates the rise of big internet gatekeepers such as Google and Facebook. The rules sparked death threats against MEPs and a million-signature petition against the proposals.
Upgrade your inbox and get our Daily Dispatch and Editor’s Picks.
Two provisions are particularly contentious. The first is Article 13, which compels internet firms, whose users upload large quantities of video, music, text and the like, to work with copyright-holders to ensure that anything that breaches copyright can be detected as soon as it is posted. That probably means they will have to deploy many more content filters like Content ID, which are worryingly imp`recise.
Technology companies, and those who advocate an open internet, say the effect will be dire. In the quest to give more protection to copyrighted work, everything from political-protest videos to citizen journalism and viral memes, they argue, risks being squashed by overzealous enforcers. The freewheeling nature of the internet could change, they warn. That is a prospect often wheeled out by the internet lobby and is probably an exaggeration. But some collateral damage seems likely. The effects could reach into unexpected places. GitHub, for instance, is an online code repository. It worries that open-source computer code hosted on its site might fall foul of the new filters.
Whose internet is it anyway?
The second fight was over Article 11, which pits tech firms against publishers. It requires social networks and aggregators such as Google’s “News” search engine to obtain a licence from publishers before displaying snippets of news reports to their users. Firms such as Google and Twitter profit from the attention generated by news that is gathered by others, note Article 11’s advocates, and should therefore share the revenues that result. But critics decry it as a “link tax” that would also radically limit the freedom of internet users.
Article 11 is a Europe-wide version of similar rules introduced in Germany and Spain in 2013 and 2014 respectively. Google’s response in Spain was to pull the plug on its news service, to the detriment of publishers that relied on it for traffic. By making a similar law apply across the entire European market, the hope is that Google (and other companies) will be forced to keep services running and share some of their revenues.
Predicting the exact consequences of all these new rules is difficult, says Jim Killock of the Open Rights Group, a British organisation that opposed the changes. They must be approved by both the European Commission and the EU’s 28 member states before they can be finalised. But the planned legislation is another example of rising European assertiveness when it comes to regulating the internet—in May the EU brought into force the General Data Protection Regulation (GDPR), a far-reaching privacy law. One result could be yet more “geo-fencing”, whereby the internet becomes fragmented along geographical lines. After the GDPR came into force, some American websites decided to block Europe-based visitors rather than comply.
More regulation may be in the offing. The day before the parliamentary vote Google was at the European Court of Justice, in Luxembourg, to do legal battle with the CNIL, France’s data-protection authority. The dispute concerned the “right to be forgotten”, under which the EU requires search engines, in certain circumstances, to remove links to webpages with personal information about European citizens. Google’s approach has been to remove links to the offending pages only for EU users. The CNIL says that, because the tech used to determine where a user is based can be circumvented, links should be removed for all users, anywhere in the world. A ruling is expected next year. Europe may not have its own internet giants, but it is having plenty of impact on America’s.