Maxime Lambrecht has created a YouTube channel to popularise Internet law, intellectual property and theories of justice, which has been awarded the Wernaers Prize for popular science. Let’s talk about what rankles: copyright on online platforms.
Freedom on the Internet is the subject of heated debate. Fake news, cyberbullying, incitement to hatred, apologia for terrorism, filter bubbles that tell you what they think you want to read...plenty of grist for the mill.
This year it’s the European directive on copyrightthat has crystallised attention, particularly Article 17 (formerly Article 13) on the ‘use of content protected by providers of online content sharing services’. The best known of these platforms are YouTube, Facebook, Twitter and Instagram.
When platforms play cops
Many users acquire or distribute, sometimes illegally, excerpts from musical, photographic, cinematographic or even literary works. Most of these works1 are covered by copyright. The European Union therefore wants platforms to try to negotiate agreements with all rights holders. Failing agreement, they should control content more. However, the only way to exercise this control on a large scale is to do it via algorithms that filter content users want to post online and block those likely to violate the law.
According to Maxime Lambrecht, a visiting professor at UCLouvain, that’s the rub. ‘Turning platforms—which, mind you, are private companies—into cops is giving them too much power. This is problematic on many levels. First, accountability to society. There’s a great deal of opacity around filtering algorithms. How are they designed? Based on what criteria? How do they work? We don’t know anything about it.’
Censorship and self-censorship
What is known, on the other hand, is that the algorithmic filtering that is supposed to preserve copyrights operates before uploading occurs. If a platform blocks content, it's simple: it's not published. ‘This prior censorship is very problematic’, Prof. Lambrecht says. ‘Certainly, the user has recourse, but it’s rarely used. Moreover, censors have the power of inertia. That is, they can decide not to follow up on recourse or within a time frame they determine.’ Which leads to another perverse effect: users aware of the platform’s limits tend to self-censor. For better or worse.
Algorithms have no sense of humour
Today, almost no one disputes the right of creators to control the marketing, distribution or (re)use of their works. But even outside the Internet, various laws provide exceptions to copyright. The right to parody and satire, for example. ‘The Internet is full of transformative content’, Prof. Lambrecht says. ‘That’s content that hijacks or transforms existing content for humorous or artistic purposes. This is the case of remixes, mashups, animated gifs, etc. Myself, on my YouTube channel, I frequently use references to popular culture. Unfortunately, algorithms have no sense of humour! They don’t differentiate between fraudulent use (plagiarism, for example) and parodic use. Blocking the latter is an infringement of freedom of expression.’
Admittedly, total freedom of expression is at the root of many Internet abuses. For Prof. Lambrecht, these challenges must be met, but not at the expense of the most basic democratic principles. ‘The Internet still has great emancipating potential. It’s a tool that increases the real freedom of individuals. Freedom to speak, learn, grow, start up ... These freedoms deserve to be protected.’
Freedom of expression: default value ?
The EU copyright directive still needs to be implemented in national legislation. Prof. Lambrecht sees this as an opportunity to interpret the articles of the directive wisely and to put the necessary safeguards in place. ‘I support “free speech by design”. In other words, freedom of expression by default. Platforms must demonstrate that they respect freedom of expression. Their algorithms must be designed to identify and account for exceptions to copyright. The goal is to prevent legitimate content from being removed out of expediency over overzealousness. Finally, platforms must be able to explain their algorithmic decisions.’ One challenge: at a time when computers are learning by themselves thanks to artificial intelligence, designers are finding it harder to understand the evolution of their own algorithms.
Notes (1) In Belgium, copyright runs until 70 years after the author’s death, when the work enters the public domain.
A glance at Maxime Lambrecht's bio
Maxime Lambrecht has been visiting professor at UCLouvain since 2014, a Hoover Chair research associate, and an FNRS scientific collaborator at ULB’s JurisLab. He is also a lecturer at the School of Graphic Research, in Brussels. He holds a law degree and a university certificate in economic and social ethics, obtained in 2008 from UCLouvain, and a 2009 bachelor's degree in philosophy from the Université Saint-Louis Bruxelles. He earned a PhD in legal sciences from UCLouvain in 2015. He specialises in ethics and Internet law, theories of justice and intellectual property.