Ray Corrigan (previously), a campaigning computer scientist at the UK's Open University, has an excellent explainer on the EU's disastrous copyright directive on the progressive academic group blog Crooked Timber (previously).
Corrigan's work is, as always, careful, detailed and measured — and incredibly worrying. He describes the technical problems with the proposals — a mandatory copyright filter and a requirement to license links to news stories — and delves into the reasons that policy-makers are so willing to create policies with such obvious deficiencies.
The first press outlets to notice these proposals were from the far right, and this has led some progressives to assume that this is a stalking horse for far right causes, but nothing could be farther from the truth. These proposals will crush all but the largest platforms — Google, Apple, Facebook, Twitter, etc — and the smarter elements of the far right understand that a world with fewer outlets for free expression will be difficult for them, because they're increasingly unwelcome on the mainstream platforms.
But they aren't the the only ones who need to fear an internet where the right to speak is contingent on the approval of a few giant corporations: as states take action against all forms of insurgent politics, including the struggles for racial equality and against wealth inequality, creating permanent advantages for Google and its rivals is also creating bottlenecks where increasingly authoritarian states can decide who gets to speak and who can be heard.
Creators, authors, ordinary internet users will be guilty of copyright infringement and automatically censored, until or unless they can prove their innocence, through an as yet unspecified process, run and administered by "content sharing service providers" in turn monitored by "rightholders." The result will a chilling effect as authors will attempt to adapt their writings to the vagaries of the filter just to make them available to otehrs.
There are broader problems. The internet is already a giant surveillance machine. The first step in blanket filtering for copyright infringement is blanket surveillance for copyright infringement. Google already have a major head start in the copyright robot filter stakes with their YouTube Content ID system. Article 13 amounts to doubling down on the surveillance architecture of the internet and handing the keys to it to Google and other large technology companies. While Google may not want this – since it would come with responsibilities (towards copyright holders) as well as power, it will surely use these tools to cement its own market advantage if it has to. More generally, network effects will mean that is only those large companies whose filters will be approved for use in the longer term.
When copyright goes wrong [Ray Corrigan/Crooked Timber]