Skip to content

Protecting refugees in the Mediterranean area. The role and challenges of AI

On 5 April 2022, the Jean Monnet Chair in Mediterranean Digital Societies and Law-Digital Mare Nostrum (DIGIMED) held its inaugural seminar with a focus on the use of digital technologies and AI in the protection of refugees in the Mediterranean area.

Experts from academia, international organisations and NGOs provided different insights and complementary perspectives, combining a theoretical approach with fieldwork experiences on migration and border management.

The Seminar started with a keynote given by Eleftherios Chelioudakis, co-founder of Homo Digitalis (a Greek NGO aiming to ensure fundamental rights and freedoms in the digital world), who highlighted the role of civil society organisations in the field of migration and border management in the European Union.

He considered both the digital rights community and the social justice community, pointing out that they pursue the same goals, namely the protection of the fundamental rights of migrants and asylum seekers. Though these two communities look at these issues from different perspectives, they focus on the same objectives: to monitor critical situations and human rights violations, and to influence policy reforms with the aim of addressing the main challenges in this field. Eleftherios stressed the importance for these two communities to work together, combining their efforts, sharing expertise, and pursuing cross-cutting actions. Focusing on concrete case scenarios, Eleftherios discussed five successful actions by civil society organisations in addressing critical applications relating to the use of social media monitoring, facial recognition, and behavioural analysis by public authorities.

The keynote was followed by an expert panel discussion with Massimo Marelli (Head of the Data Protection Office, International Committee of the Red Cross), Clementina Barbaro (Head, Youth Policy Division – Secretary to the Joint Council on Youth and the European Steering Committee for Youth, Council of Europe) and Aaron Martin (Postdoctoral Research Fellow at the Tilburg Institute for Law, Technology and Society and Assistant Professor in Maastricht University’s Humanitarian Action Programme). These three speakers focused mainly on the international perspective and the interaction between the different actors involved in humanitarian crisis management.  

In discussing the role of humanitarian organisations, Massimo Marelli emphasised the importance of digital technologies in providing better assistance in crisis scenarios. He noted the role that technologies such as AI and ML can play in better understanding the contexts in which humanitarian organisations operate and in identifying missing persons and those who need assistance. Digital platforms are also crucial in facilitating interaction with people affected by conflicts, enabling them to access humanitarian aid, as well as providing access to social media and making it possible for these people to contact their families and loved ones. From a broader perspective, Massimo then pointed out that, in this digital environment, international data protection principles and their technical implementation play a key role in safeguarding human dignity in the context of humanitarian actions.

This international perspective focused on human rights and human dignity was further elaborated by Clementina Barbaro, based on the long-standing experience of the Council of Europe in safeguarding human rights, democracy, and the rule of law. She highlighted the important contribution the Council of Europe is giving to human-centred regulation of AI applications, which are the most challenging but also promising technologies, including in the field of refugee protection and humanitarian actions. This is the case, for example, with technologies such as facial recognition tools and drones used to predict and control migration flows. In this regard, Clementina highlighted the various challenges arising in the regulation of AI-based technologies, involving both the impact on human rights and economic interests. She therefore stressed the importance of establishing a strong, predictable, and clear legal framework for AI, which should include both common principles and sector specific provisions.

Finally, Aaron Martin shifted the discussion from technology regulation to the crucial aspect of partnerships between the technology sector and the humanitarian sector. He drew attention to the opportunities and risks of such partnerships, where humanitarian organisations are starting to become overly dependent on the private sector. In addressing these issues, Aaron provided a very concrete and interesting experience by discussing some case studies that illustrate the main goals and issues of technology partnerships for humanitarian aid, such as the recent partnership announced during the Covid-19 pandemic between the World Health Organisation and Whatsapp to provide official information on various topics.

The last session of the seminar was a roundtable discussion where Chloé Berthélémy (European Digital Rights, Policy Advisor), Theodora Gazi (University of Athens, Lawyer), and Bayram Selvi (The Turkish Red Crescent, Director of Migration Services) focused on the use of personal data in the context of humanitarian actions, combining legal analysis of concrete cases and fieldwork experience.

The use of data-intensive technologies at EU borders and in refugee camps (such as biometric tools used by humanitarian organisations to regulate access to food) was discussed by Chloé Berthélémy, also examining several tools and applications used by law enforcement to control migration flows and pointing out the risks of prejudices to human rights.

The impact that these technologies, including AI, have on data protection in the various humanitarian aid operations (e.g. camp management, nutrition and education needs) was then discussed by Theodora Gazi pointing out the resulting risks to human rights. Theodora suggested the adoption of specific data protection policies for the humanitarian sector to mitigate these risks and emphasised that, even in an emergency context, there should not be a dilemma between using data in an effective way and using data in a lawful manner, since the former does not preclude the latter.

Finally, Bayram Selvi shared his experience in developing humanitarian aid projects, providing several examples of technology applications used in this context. Despite some critical issues concerning these data-intensives applications, he pointed out that facial recognition and data-driven solutions can facilitate various operations, for example in the field of missing person identification, making it possible to reunite displaced family members of refugees and migrants. In the same way, chatbots can facilitate faster communication in different languages and fuzzy search tools for names can help to better manage people from different countries whose name and surname are spelt differently.

The different voices and experiences that contributed to this seminar clearly showed the existing tension in our Mediterranean societies between the use of digital technology and the focus on human rights.

While there is no doubt that data-intensive technology and AI can provide significant advances in a number of fields, including refugee protection and humanitarian action, several relevant aspects require close attention. This is the case with the role of state actors, the interaction between public and private actors, the role of NGOs and civils society, the need for a clear international legal framework, the proper implementation of existing data protection rules, and the need for a human rights by-design approach in the development and use of technologies.

These considerations suggest avoiding a techno-solutionist approach and building on the experience of Mediterranean countries to lead the regulatory and policy debate in the direction of common best practices and regulatory solutions that address the challenges posed by the dynamics faced in humanitarian scenarios.