In the midst of a global pandemic, the European Union has been implementing various measures aimed at combating terrorist threats in all their forms and declinations. Even though EU countries appear to have overcome the most critical phase in terms of the spread and frequency of bloody terrorist attacks, since 2015, the EU has not stopped keeping abreast of major terrorist adaptations to the latest technologies. Accordingly, in the aftermath of the attacks that occurred in Paris, Dresden, Conflans-Sainte-Honorine, Nice, and Vienna in November 2020, EU home affairs ministers agreed to further strengthen their joint efforts to fight terrorism. The shared political will resulted in a European Council conclusion reached one month later.
In recent years, the EU strategy for combatting terrorism has focused on the following:
- enhancing the exchange of information
- reinforcing checks at external borders
- preventing online radicalization
- criminalizing terrorist financing
- strengthening cooperation with non-EU countries
Despite the overall success in addressing terrorist threats, several steps still must be taken to prevent online radicalization. However, as stressed by human rights and journalist associations, the prevention of online radicalization must be reconciled with the need to ensure the respect of fundamental rights, such as freedom of expression and beliefs. Based on these premises, this article seeks to shed light on the regulation on addressing the dissemination of terrorist content online adopted by the European Council on 16 March 2021 and analyze its pros and cons.
The 2020 Counter-Terrorism Agenda
In order to understand the scope of the latest European Council regulation, it is important to briefly refer to the overall legal framework in which it came to light. Indeed, the European Union has been emphasizing over the last year the need to enhance cooperation among states when it comes to efficiently tackling the terrorist threat. By and large, the formal process of developing general counter-terrorism policies and laws lies in the hands of heads of Member States which, through the European Council, set up guidelines. On the other hand, the EU Commission produces proposals for laws and policies which are discussed by the Council and Parliament. Concurrently, the EU Commission was tasked to ensure enforcement and compliance with regulations.
Given the spike in terrorist threats, the Commission presented a new Counter-Terrorism Agenda within the EU’s Security Union Strategy in December 2020. The agenda aimed at enhancing wider coordination and cooperation between Member States, European institutions and civil society in fighting terrorism. With a pragmatic approach, it sets out a series of concrete actions to be taken at the national and EU level, by identifying four major action pillars:
Anticipation remains crucial to predicting and assessing the risk of potential threats. Indeed, the first pillar stresses the need to reinforce the early detection capacity through constant cooperation between national intelligence services and the EU Intelligence and Situation Center. Regarding the prevention of terrorist threats, the EU Commission proposed the adoption of a Digital Services Act to speed up the removal of illegal online content alongside the implementation of the EU Crisis Protocol. Contextually, the Commission highlighted the urgency to intervene in the areas of rehabilitation and reintegration, especially concerning foreign terrorist fighters and their family members. All these processes must be supported by constant research conducted by the Knowledge Hub on the prevention of radicalization.
From an operational standpoint, the Commission called upon Member States to ensure the protection of public spaces and critical infrastructures, pinpointing sensitive targets. Conclusively, the European Agenda aims at improving operational police cooperation at the European level, further urging to better protect the victims of terrorist acts through reparative and compensative measures. In this context, the Council adopted a regulation on addressing the dissemination of terrorist content online in March 2021, subject to the approval of the European Parliament at the end of April.
Combating Radical Content Online
According to the 2020 EU Terrorism Situation and Trend Report, social networks and messaging applications played a key role in the planning and execution of terrorist attacks that occurred in the last few years. Encryption technologies made it practically impossible for authorities to trace online communications, giving cover to terrorists to plan freely. Furthermore, it is an established fact that the internet facilitates and sometimes accelerates the radicalization process, acting as an “echo chamber”. However, radicalization is a complex process involving social, psychological and opportunistic factors.
For the time being, the European Parliament must decide whether they will approve the text, amid concerns expressed by human rights and journalist associations. The regulation calls upon Member States and any competent authority to issue removal orders to any service providers which host radical content. Internet platforms are required to disable or remove the content within one hour of the order. Accordingly, the service providers will be the ones responsible for evaluating the extent and the nature of the threat, thus deciding whether the lexicon or the offending expression falls under legal or illegal speech.
More specifically, any competent authority in a Member State will have the power to enforce its jurisdiction in any other Member State, ordering the deletion of content within one hour. The Member State of the hosting service provider will check whether the order “seriously or manifestly” infringes on fundamental rights. This clause could allow broad interpretations; however, the EU regulation mentions a generic duty for providers to respect freedom of expression and information alongside offering judicial remedies to users and service providers.
The decision to entrust algorithms and automated content moderation tools to quickly delete content may result in the violation of freedom of expression. Indeed, it is not completely clear whether those automated systems can distinguish between propaganda and satirical or educational content. International non-governmental organizations such as Human Rights Watch (HRW) have noted that the lack of a preliminary judicial oversight on the censuring measure is a concrete risk to freedom of expression, assembly and religion, and thus, contrary to the Charter of Fundamental Rights.
Machine Learning Algorithms
The use of machine learning algorithms to track and detect terrorist content has become extremely widespread nowadays. This is due to the fact that it would be almost impossible for judicial authorities to manually discover all terrorist digital footprints. Therefore, governmental agencies, as well as scholars and researchers, have been studying how to use machine learning algorithms to detect extreme language and measure propaganda thresholds. On the other hand, concerns related to the breach of personal privacy and freedom of expression are increasingly being raised around the world.
Studies have shown how search engines — by steering and refining inquiries — help users navigate content and eventually find the information they are looking for. Moreover, the algorithms behind recommendation systems create a spiral of related content, which can reinforce extremist attitudes. Social media platforms such as Facebook have the ability to filter out around 98% of malicious content on their platform thanks to machine learning algorithms. However, shortcomings remain, particularly when it comes to complex contexts and the translation of languages. For this reason, the latest machine learning-based innovations like natural language processing (NLP) are becoming very useful in managing user‑created content and improving content moderation. NLP could also analyze language in its contextual and syntactical structure, making it potentially capable of distinguishing free speech from terrorist propaganda.
The online world is permeating our entire existence in many aspects. Philosophers, AI scientists and sociologists are researching solutions and questioning the main ontological and ethical challenges of our time, underlining that the line between the online and offline realm is becoming blurry. The Italian philosopher of information Luciano Floridi, for this purpose, coined the term “onlife” to describe a new hybrid dimension where real-life encounters the digital sphere. Within this interrelated and intricate overlapping of communicative layers, it is not easy to find a concerted solution that may reconcile all instances.
The new EU regulation on addressing the dissemination of terrorist online content is certainly a valuable attempt to combat online radicalization, although it has been fairly criticized by human rights activists. Certainly, machine learning and AI may assist humans in their assessing processes; but, at present, they cannot completely replace them. Meanwhile, EU institutions must protect the safety of their citizens while ensuring full respect for fundamental rights. Within this subtle balance of rights, each actor must pursue its mission with the hope that this participatory dialectical process, otherwise called democracy, will be able to find a shared and satisfactory solution in the end.
European Eye on Radicalization aims to publish a diversity of perspectives and as such does not endorse the opinions expressed by contributors. The views expressed in this article represent the author alone.
 Floridi L., The Onlife Manifesto, Springer, November 2014.
 Schroeter M., Artificial Intelligence and Countering Violent Extremism: A Primer, in Global Network on Extremism and Technology, King’s College, October 2020.
 Valentini D.; Lorusso A.; Stephan A., Onlife Extremism: Dynamic Integration of Digital and Physical Spaces in Radicalization, in Frontiers in Psychology, March 2020.