Antiterrorists in a bike shed – policy and politics of the Terrorist Content Regulation

The short story: an ill-fated law with dubious evidence base, targeting an important modern problem with poorly chosen measures, goes through an exhausting legislative process to be adopted without proper democratic scrutiny due to a procedural peculiarity. How did we manage to end up in this mess? And what does it tell us about the power of agenda setting the name of the “do something” doctrine?

By EDRi · May 25, 2021

How it started – how it’s going

A lot of bafflement accompanied the release of the Terrorist content regulation proposal. The European Commission published it a few days after the September 2018 deadline to implement the Directive on Combating Terrorism (2015/0625). It is still unclear what the rush was with the regulation if the preceding directi ve hadn’t got much traction. At that time, only a handful of Member States met the deadline for its implementation (and we don’t see a massive improvement in implementation across the EU to this day). Did it have to do with the bike-shed effect pervading modern policy-making in the EU? Is it easier to agree on sanitation of the internet done mostly by private corporate powers, than to meaningfully improve actions and processes addressing terrorist violence in the Member States?

TERREG’s problematic rationale – doubtful evidence of how acute the problem of exposure to terrorist content is[1] combined with blatant lack of evidence of causation between exposure to content and radicalisation[2] – were followed by ill-designed solutions we wrote so many times about [3]. It was immediately clear that the best possible scenario for this file would be to shelf it into legislative oblivion for eternity. With a general overhaul of responsibilities of platforms looming, it made little sense to tackle controversial political communication before addressing the fundamental ills of the platform ecosystem (what we now know as the Digital Services Package) and see how these work first.

We knew that once the EC gives life to a legislative proposal, no matter how bizarre, killing it is rather out of the question. Still, if it had to be proceeded through for political reasons, it should have been narrowed down to illegal content. This was the second-best scenario, and surprisingly, it seemed to get some traction with the first rapporteur for the file in the EP, Belgian conservative, Helga Stevens. Her successor, Daniel Dalton (ECR, UK) was not hell-bent on shelving the file but did a considerably decent job gutting what became the final EP report from the most toxic provisions.

The Regulation that the EU legislator adopted in the end is not the worst scenario out of all options discussed during the legislative process. It is not the original version that contained no exceptions, and introduced a catastrophic concept of referrals, as well as state-sanctioned content filtering[4]. It is also not the German Presidency’s “compromise” with common sense, proposing that speech is in fact illegal until determined otherwise. Still, TERREG is a very troublesome piece of legislation, with dysfunctional judiciary oversight and cross-border content takedowns that will surely backfire and complicate the future of politicised speech in the EU.

Chances we lost, lessons we learned

There was almost no popular mobilisation around the dangers of filtering content to eliminate terrorist propaganda. One of the reasons was the fatigue of those potentially affected soon after the copyright reform debacle. It is clear that there was no one with a combination of means and outreach to effect an outrage in popular culture with the help of youtubers and tiktokers – the mighty powers that drove youth to the streets of European cities in the final stretch before the plenary vote on the copyright Directive (CDSM). The civil society seemed to have a limited ability to awaken the society to the danger of content filtering when it was labeled as a solution to stop terrorists radicalising the children.

The popular mobilisation around CDSM was in fact fleeting capital and not a solid foundation to build awareness and resistance to surveillance capitalism upon. Digital rights do not have, in fact, its grassroot EU-wide organisational backbone and perhaps the opportunity to capture the energy that erupted during the “no filter” campaign is lost forever. The question is also who would be in a position to convincingly call for a digital-rights equivalent of extinction rebellion?

We need to come clear here: EDRi is a network of organisations that is mandated to pursue a highly specialised agenda setting for digital policies with modest financial means and staffing. The Wikimedia movement can be considered pretty close to an example of a grassroot activism towards better access to information online, but policy and advocacy are not the primary focus of its institutional nodes. The movement itself is torn on the issue whether they should at all support advocacy on general digital issues that don’t affect their core projects[5].

Harnessing popular outrage against TERREG was perhaps the only thing that could either make the outcome better in its substance or even retire it to a legislative shelf. But our organisations did not make it a tactical priority. It may have been a mistake only evident in the future. We cannot speak for others who worked on the file and took their own strategic decisions.

Who is a terrorist?

What didn’t help, was of course the nominal objective of the law: the so-called “war on terror”. In the example of copyright laws, a growing number of people in the EU, especially young, do see the influence of it on what they can see and communicate online. A relationship between anti-terrorist laws and freedom of expression is not half as evident. Terrorists are bad, everyone agrees with that.

There is a perception that speaking against anti-terrorist laws is difficult to defend and justify – we knew this as we also grappled with doubts if we can meaningfully challenge the presumptions of the legislators without being labeled allies to terrorists. Some stakeholders, such as consumer rights organisations who had been vocal against content filtering during the copyright reform, were nowhere to be seen in the discussions around TERREG, and it may be that the “radioactivity” of the topic was the cause of that withdrawal. Similar trends are evident now in the child sex abuse material (CSAM) discussions, where few groups want to be vocal about a file that may be a PR-disaster if it is poorly communicated.

Convincing the politicians

In order to successfully advocate for better legislation, it was crucial to convince the legislators that TERREG is in fact about curbing political speech and that it misses the mark of meaningfully tackling terrorism. An outline of threats to fundamental rights, and political speech in particular gained understanding on the progressive left side of the political spectrum and some interest among representatives of other political groups.

Donate to EDRi to build a people-centered, democratic digital future. Donate Now

A number of MEPs showed understanding to our positions without any consistent follow-up on our recommendations. These were mostly representatives of well-established political groups with somewhat moderate and pro-free market digital policies on agenda. We can only speculate why those liberals and social democrats who often champion personal freedoms and privacy decided to support the final outcome in the final LIBE Committee vote[6]. Perhaps they did not want to reject an anti-terrorist law knowing how hard it was to arrive at the mediocre but palatable compromise. Pragmatically speaking, if the final compromise is rejected, there won’t be a better text next time around.

MEPs and representatives of Member States to the EU are not as deaf to nuance as some would expect, which is good news. All that is a proof that we can present our narratives around “radioactive” topics to such as terrorism decision-makers if we come prepared, do it confidently and coordinate. Our homework is to become even better at this, to push people from interest to action across the political spectrum.

Representatives of the European Commission were an exception though. They often simply repeated talking points about how well designed this legislation is, not stooping to engaging with our arguments in any way. That, however, is not incapacity, it is political tactics of repeating falsehoods in hope that they will eventually ring true.

What the lawmakers get away with

Some may argue that the lack of the final vote did not change much – everyone expected the Regulation to be adopted, anyways. But if democracy is about betting on the expected, then we don’t need either the Brussels or the Strasbourg seat of the European Parliament. Democracy is putting both the expected and unexpected outcomes through the public scrutiny: through a vote of elected representatives. Missing the deadline to request the TERREG’s final Plenary vote deprived the citizens of that scrutiny.

It may be argued that the procedure of the Second Reading, which does not default to a final vote, is in principle undemocratic; and this assumption seems fair. In the case of TERREG, the murkiness of the final act is exacerbated by another problem that seriously undermines the faith in the democratic nature of the EU – secrecy of the trilogues. Any trilogue negotiations between Member States and the European Parliament are behind closed doors, and unless a MEP or a State sees it as their duty to inform citizens, no one will know what was the nature of the compromise, who had the winning strategy and why. TERREG is one of the many victims here. It is impossible to plan for a better regulation within the EU or cultivate trust in its institutions without considering a reform that has transparency throughout the whole legislative process as its key objective.

The looming shadow of justice

The secrecy enables Member States to ignore their national legislation and even lobby for the solutions that go against court decisions in their own jurisdiction. Let’s not forget that during the negotiations the French Constitutional Court struck down the so-called Avia Law, questioning the provisions that are also present in TERREG: 1 hour content takedown deadline for removal orders issued by administrative power[7]. The result is that TERREG itself may be challenged in one or more Member State’s jurisdiction. Such problems may potentially take disputes over TERREG as far as to the Court of Justice of the EU.

One convoluting circumstance that may cause problems is the label that the European Commission stuck on the law – making it, nominally, an instrument of the Digital Single Market. Yet it was not the DG CNECT but DG HOME who piloted it on behalf of the EU, and it is evident that TERREG creates an alternative system of criminal law enforcement by both states and private actors. As it opens avenues to police what people say online with little transparency and no proper judicial process, it may be weaponised by governments in those EU countries where authorities already organise crackdowns on citizen dissent.

The sum of all fears

TERREG is not unique in the way that the law was drafted and proceeded under political pressure to respond to a badly defined problem, with a potentially destructive effect on perfectly legal actions. For laws regulating online communications, however, it established a firm precedent of reinforcing private surveillance and uncountable governmental control over political speech through the backdoor of market regulation and under a pretext of an issue people are genuinely concerned about. It remains to be seen how this experience will deteriorate the quality of other legislation impacting fundamental freedoms online.

The sum of all mishaps, procedural problems, political calculations and lack of climate for meaningful campaigning (last chapter of this fight happened during the second Europe-wide COVID-19 lockdown) brought the usual effect of lackluster lesser evil. And as lesser evils tend to, in the long term, this one will undermine both people’s freedom and their convictions of how effectively they can hold their representatives accountable.

Image credit: Book Catalog/Flickr

(Contribution by: Diego Naranjo, Head of Policy, EDRi & Anna Mazgal, EU Polizy Advisor, Wikimedia Deutschland)

[1]According to the Eurobarometer poll cited in the Impact assessment to the proposal, only 6% of people in the EU reported being exposed to online terrorist content. Statistical margin of error (3%) aside, self-reporting does not reveal what the exposed saw exactly, so there is no way to evaluate whether the content they have in mind would be considered “terrorist” according to the new law.

[2]During an event in the European Parliament on the issue we participated in, in February 2019, a representative of DG Home casually admitted that the European Commission did not see any evidence supporting this claim and they “had to make an educated guess”. If that is not policy-based evidence, we don’t know what is.

[3]Check, or l for details

[4]EDRi analysed the dangers in detail here: and Wikimedia here:

[5]Wikimedia communities debated the role of policy work and its alignment to the mission of the movement for example during the 2030 Movement Strategy work; within a working group on advocacy

[6]One can think of Sophie in ’t Velt, the liberal powerhouse if it comes to privacy protection; or Marina Kaljurand, the S&D Shadow Rapporteur in LIBE, who contributed to improvements of the original proposal during the process in the Committee – both voted in favour of the adoption of the final trilogue outcome. Only Greens and the Left (formerly GUE) voted against it.

[7]Wikimedia analysed the interplay between the French decision on the Avia Law and TERREG: