European policymakers have recently reached a provisional deal on the highly anticipated AI Act, a regulation set to govern the use of artificial intelligence across the European Union. While this development marks a significant step forward in regulating AI, concerns are surfacing about potential loopholes that could leave migrants and asylum seekers without adequate protection from AI biometric systems.
Migrants’ agency at risk: Critics weigh in
A research article published on the eve of the AI Act’s passage has raised a red flag, highlighting how mandatory biometric data collection practices within the EU may undermine the agency of migrants residing in the Eurozone. Critics argue that the Act, as it stands, may fail to bring about meaningful change for this vulnerable group.
Rights organization Algorithm Watch issued a statement, pointing out that the Act contains exceptions for high-risk systems, particularly those used in national security, law enforcement, and migration. These exceptions potentially allow authorities to sidestep the core provisions of the Act. Among the systems that raise concern are live facial recognition used in public spaces to identify crime suspects and technology employed to assess the emotions of asylum seekers.
The level of protection these systems will offer remains uncertain until the final text of the AI Act is released, adding to the anxiety of rights advocates.
Migrants excluded from key protections
Advocacy groups have long voiced their concerns about the exclusion of migrants from the safeguards provided by the AI Act. The European Council’s exclusion of border control areas from the definition of “publicly accessible spaces” is a particular point of contention. This exclusion opens the door to the use of remote biometric surveillance in these areas, further raising questions about the rights and privacy of migrants.
In a report published in September, the Border Violence Monitoring Network (BVMN), a coalition dedicated to promoting the rights of migrants and asylum seekers, warns that the growing convergence of biometric migration and criminal databases could contribute to the criminalization of migrants. BVMN also criticizes the lack of substantial safeguards and the limited focus on the rights of people on the move within the AI Act.
Outsourcing responsibility: A new concern
European countries appear to have found an alternative route to handling migrants’ biometric data and the AI systems that rely on it – outsourcing. According to BVNM, the EU may be compromising the rights of asylum seekers by shifting responsibility for managing applications to third states, particularly those located along the Western Balkan migrant route, a major pathway into Europe.
Balkan countries have been establishing their own “Balkandac” system, inspired by the EU’s fingerprint database, Eurodac, which is used to identify asylum seekers and irregular border-crossers. Cooperation between countries in the region, including non-EU members such as Serbia, Bosnia, and Herzegovina, along with EU member states and agencies like Frontex, has propelled the development of these data systems. Funding from the EU’s Instruments for Pre-Accession Assistance and bilateral agreements with EU countries has further supported these initiatives in the Western Balkans.
Despite these developments, concerns abound. The biometric data systems in the Western Balkans are seen as contributing to heightened securitization in managing migrant flows into the EU. Besides potentially blurring the lines between immigration and criminal law, the Balkandac systems lack anti-discrimination safeguards and may sidestep critical data protection principles.
Balancing act needed
As the EU advances towards ratifying the AI Act, the concerns surrounding its impact on migrants and asylum seekers remain prominent. Critics argue that without stringent safeguards and a clear focus on the rights of these vulnerable populations, the Act may inadvertently perpetuate discrimination and surveillance.
Balancing the need for effective AI regulation to protect individual rights, particularly in the context of migration and asylum, is a complex task. As the final text of the AI Act is unveiled, it remains to be seen whether European policymakers can address these concerns and strike that delicate balance between technological progress and human rights. The clock is ticking, and the world is watching closely.
