The article is available in iMagazine 7/2018.
Despite widespread opposition, on 21 June 2018 the European Parliament’s Legal Affairs Committee approved a draft directive on copyright in the Digital Single Market. Critics say Article 13 will end the internet as we know it.
What is Article 13 about?
The proposed provision imposes a new duty on major online service providers: taking “measures” to protect copyright in works published on their platforms. It seeks proper performance of agreements with rights holders and prevention of unlawful publication, particularly through “effective content recognition technologies”. In simpler terms, providers would have to filter and monitor user uploads to prevent copyright infringement.
This significantly changes current law. Providers presently need not use preventive filters. A duty to remove unlawful content arises only after notification by a person harmed by infringement, requesting removal, for example. Independently, that person may pursue claims against the infringing uploader.
Criticism
Critics condemn the proposals. They argue that automated filtering ends the free internet, turning an open platform into a tool for tracking and controlling users. Ordinary uploaders, contributors to open platforms such as Wikipedia and open-source repositories such as GitHub would be affected. Smaller European providers, including start-ups unable to invest in filtering technology, may suffer, with some forced to close. Other objections concern unclear, undefined concepts, potential conflicts with earlier directives and abuses threatening free expression and information flows.
Is internet freedom threatened?
The draft raises doubts. It is unclear exactly who must filter. It refers to providers storing and providing access to a “large amount of works”, without defining that phrase. A relatively broad group is likely covered: Article 13 could encompass not only social networks but virtually any open publishing platform.
Nor does it explain the duty itself. Content recognition technologies must be “appropriate and proportionate”, without guidance on those standards. Crucially, responsibility for content legality shifts to providers, modifying the principle that the uploader bears primary liability. Responding to reports would no longer suffice: providers must prevent infringement actively. They may block anything carrying even a slight liability risk. Doubts would default against users, leaving them to appeal.
The draft also requires Member States to ensure complaint and redress mechanisms for disputes arising from these technologies, but does not explain how to make them effective and quick enough to avoid censorship or blocked access.
Free-expression risks demand particular scepticism. The proposal introduces prior scrutiny of what we intend to upload, creating a barrier between users and the internet. Although designed to protect copyright, less noble uses are easy to imagine. Those controlling the barrier could decide what reaches the internet. It is a short step to delaying or blocking inconvenient, politically incorrect content or material contrary to their interests.
Content ID for everyone?
Content filtering and recognition are not entirely new. YouTube creators already encounter Content ID, developed in 2007, which lets rights holders scan for copies of their works. Unauthorised publications may be blocked or monetised. The system uses machine learning and dispute tools. It cost Google tens of millions of dollars and, critics note, remains imperfect.
It is hard to escape the impression that EU lawmakers liked Google’s system so much they decided to impose it on everyone. Yet they did so without deeper thought about implementation or consequences. The directive’s fragmentary, imprecise rules merely sketch the idea, raising more questions than answers. Above all, they create fears of waking up separated from the internet by a cordon of filters. Further legislative work therefore deserves attention.
