[Note: This analysis is published in the context of the 2019 #CopyrightWeek, under the theme of 14 January on ‘Copyright as a Tool of Censorship’, which revolves around the idea that: “Freedom of expression is a fundamental human right essential to a functioning democracy. Copyright should encourage more speech, not act as a legal cudgel to silence it”.]

On 14 January, the #CopyrightWeek theme is ‘Copyright as a Tool of Censorship’. The starting point is that: “Freedom of expression is a fundamental human right essential to a functioning democracy. Copyright should encourage more speech, not act as a legal cudgel to silence it”.

Long before the introduction of the Copyright Act in the early 1700’s with the so-called ‘Statute of Anne’, the earliest interpretations of copyright did not revolve around protecting the fruits of creators, but rather about the desire of the Crown to exercise control over the book publishers. Copyright was thus initially a tool of censorship, instead of an enabler of creativity and freedom of expression, as Law Professor Lydia Loren explains.

Whilst the impetus to encourage creativity also made its way in the copyright legal world over the decades, the tension with censorship has never fully disappeared and, unfortunately, copyright is slowly starting to revert back to its original role: being the weapon of choice for targeted censorship, under the label ‘protecting creators’ (well, it’s actually rightholders, but…). Everything ranging from political commentary to reviews of construction companies, movies or games seems to be prone these days to being taken down based on copyright infringement grounds (Techdirt has been keeping track of these ‘anomalies’ to the copyright system).

The Electronic Frontier Foundation highlighted this ‘dual-use’ of copyright in a 2017 Copyright Week contribution, as they observed that: “Carefully balanced and limited, copyright can reward artists and promote creativity. But today’s copyright law, and the mechanisms created to enforce it, enables censorship that harms artists, voices of political dissent, and all who speak against the powerful.”

The Gratuitous Nature of Copyright Infringement Claims: Censorship ‘at a Click’ and Bogus Claims Remain Unpunished

The fact that copyright can too easily be abused today for targeted censorship stems from the gratuitous nature of copyright infringement claims, as such claims can be filed with a simple click and abusive copyright claims (often) remain unpunished.

Today rightholders, or anyone pretending to claim rights over some form of content, can just file notice-and-take down requests and easily get content online removed without any form of due process for the user who uploaded it. In Europe, contrary to other geographies such as the US and Canada, this take down happens without any opportunity for the users to file a proper counter notice, hence given no ‘right to defend oneself’. This is obviously the preferred approach by rightholders as it is much quicker than going after people through court proceedings, or actually just accepting that they are exercising their legitimate right to freedom of expression.

These practices form of moral hazard, resulting in illegitimate acts of censorship from false copyright claims done either in full knowledge or inadvertently. The fact that such claims can be filed gratuitous at the click of a button (or even through automated tools) and that in the majority of cases such bogus claims are never punished in any way has helped to transform the copyright enforcement system into the perfect tool for target censorship.

False copyright claims sometimes even target content that is no longer protected by copyright, as it is in the public domain, a phenomena that was coined as ‘copyfraud’, back in 2005, by Professor Jason Mazzone.

Recent examples show that even Members of the European Parliament, although enjoying political immunity, are definitely not immune to false copyright claims: mid-2018, German Pirate Party MEP Julia Reda reported about how ‘Topple Track’, a service that was offered by ‘Symphonic Distribution’, and which described itself as “one of the leading Google Trusted Copyright Program member” and boosting a 99% removal rate, had issued a notice to Google, revolving around the album and song ‘This Time’ from ‘Real Housewives of Melbourne’ TV personality ‘Gamble Breaux’, wrongfully claiming that a blog post from her, and a number of other websites, where infringing Breaux’s copyright. Result: thanks to the automated mechanisms operated by Google, MEP Reda’s page was instantly delisted from Google’s search results, without any questions or human intervention.

MEP Reda noted that thanks to research from the Electronic Frontier Foundation (EFF) and coverage by TorrentFreak, which shed some light on more bogus removal requests by ‘Topple Track’, Google apparently decided to terminate its cooperation with the firm. The latter could explain why the service suddenly vanished from the earth (and Internet) since then. Along with this experience of being ‘shadow banned’, MEP Reda also witnessed how long it took before the webpage concerned was reinstated into the Google search results.

This illustrates that political actors can be silenced through the use of copyright, even when no infringement has taken place. Whilst the motivation of Topple Track seems to be that of a copyright troll, other cases have been more clearly driven by political motives. In 2014,  Techdirt reported that Turkish Prime Minister, Recep Tayyip Erdogan, had “taken out a copyright” in his phone calls, to request their removal from social media platforms. There is also the bizarre case of China blocking Winnie the Pooh images due to the analogy between the cartoon character and China’s president, but then again that was not based purely on copyright. In the US, the removal of ‘inconvenient’ political ads through DMCA takedowns has been reported multiple times: ads related to the ongoing adoption of specific legislation, or ads related to specific elections have been known to be hit by bogus claims to shut down political speech during timely moments.

The Slippery Slope: From Bogus Takedown Notices to Automated Censorship Filters

In the meantime, European Union legislators have been discussing the EU copyright reform since September 2016, and the outcome that we are bound to get looks far from pretty for citizens. Instead of focusing on enforcement policies in the context of the e-Commerce Directive, by improving the notice-and-takedown regime, including through better user safeguards – for example through the introduction of counter notices, the European Commission, and subsequently the EU policymakers, have fallen into the rhetoric of the rightholders (mainly from the music industry, and please do not confuse rightholders with the actual creators).

The rightholders successfully convinced a number of policymakers that the notice-and-action approach does not have enough teeth and that it is too burdensome for them (read all about the intense lobbying on this file in the report from Corporate European Observatory). Instead, according to them, and supported by firms such as Audible Magic,  the silver bullet is to force all online platforms to adopt YouTube ‘Content ID’ like automated content recognition technology, but on steroids, that would ensure no copyright infringement ever sees the daylight again. The result: the Article 13 #CensorshipMachine upload filter was born.

Since this proposal was unleashed to the world, little has improved, and the final language is set to create a true monstrosity (read-up on Article 13 here and here). In order to pretend to avail the numerous complaints, be it from Internet luminaries, civil and digital rights organisations, the UN Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression, to startups, some cosmetic changes have been introduced.

These changes include some ill-defined user-safeguards that will be swept under the carpet, as platforms will find it easier to ensure compliance with this legislative mess by blocking content based on their service’s terms and conditions, hence by-passing this entire ‘redress mechanism’ charade to please some rightholders and reduce the cost of compliance.

The fact that a number of Member States, including the current Romanian Council Presidency, are pushing eagerly to ensure that content that is being taken down stays down forever is especially worrying in terms of exercising our fundamental right to freedom of expression. This facilitates the task of those wanting to censor critical voices, as killing them softly once through a copyright claim is enough to mute their voice for good.

A stay-down obligation also equates to a form of general monitoring, as CDT already remarked in 2012 that: “No matter how specific the content being targeted, an obligation to prevent re-uploads means that a service provider must monitor all uploads on an ongoing basis to be sure to catch re-uploads”. Looking at the stay-down approach, Dr Felipe Romero-Moreno, from the University of Hertfordshire, found that Article 13 would be violating platforms and users’ rights under Articles 6, 8 and 10 of the European Convention on Human Rights.

Conclusion: True Safeguards Would be Good, No Article 13 #CensorshipMachine Would be Better

Legislators have thus brushed aside the numerous concerns raised around the impact Article 13 #CensorshipMachine will have on freedom of expression. Instead, they are set to open the flood gates to enable targeted copyright censorship on an unprecedented level.

The current experience (and the worst is yet to come) already shows the need to set higher standards of evidence for copyright claims in order to guarantee citizens fundamental rights. In order to better protect users, redress mechanisms should ensure that content at least remains online until any dispute around copyright claims is resolved. This will incentivise rightholders acting legitimately to defend their claims, and will drastically reduce the powers of those with less good intentions to magically make content disappear.

In order to make it even less appealing for these boogie-men of the Internet to abuse the system, real-life implications should be linked to putting forward false copyright claims, for example by imposing damages in cases of recurring or obviously inaccurate claims.

In this context, more public scrutiny around copyright claims should become the norm. The trend towards open data could be beneficial in this area, making copyright claims, or at a minimum statistics about them, a matter of public record would enable more transparency and make abuses easier to spot (in the US there’s already an initiative of this kind: the Lumen Database).

The best solution would actually be for EU legislators to stop this whole Article 13 #CensorshipMachine in order to #SaveYourInternet, and instead rethink the whole approach towards copyright enforcement from a creator and user-centric approach. Instead of trying to tick the boxes of the wish list of some rightholders, policy makers should focus on ensuring that proper safeguards are put in place for users of those platforms that today already ‘voluntarily’ put in place some form of content recognition technology, in order to ensure the many filtering mishaps that take place today are no longer allowed to take place without some form of scrutiny.