The European Parliament is contemplating imposing strict conditions on the utilisation of real-time remote biometric identification technologies as part of the AI Act, a proposed EU legislation aimed at governing the potential risks associated with AI systems. The use of remote biometric identification (RBI) has been a point of contention, as the Commission initially proposed permitting real-time RBI in specific situations, while the Parliament favoured an outright ban to prevent widespread surveillance. A compromise text has now been introduced, reintroducing the possibility of real-time RBI in exceptional cases, primarily for certain criminal offences carrying a minimum punishment of five years, such as terrorism and human trafficking.

Law enforcement agencies are required to register the system in the EU database, conduct a thorough assessment of its impact on fundamental rights, and obtain validation from a judicial authority. Additionally, national authorities would oversee its usage. In return for these concessions, the Parliament has expanded the list of prohibited AI applications, including systems that create facial recognition databases through indiscriminate data scraping, technologies inferring sensitive information through biometric categorization, and emotion recognition systems in specific contexts. Nonetheless, differences persist concerning the use of AI in predictive policing and exemptions for law enforcement and national security purposes.

Why does this matter?

This move comes after heated debates and negotiations amongst EU institutions. Furthermore, the EU aims to strike a balance between maintaining security and preventing potential abuses of AI technologies for mass surveillance or other invasive purposes. This balance is crucial for maintaining public trust and addressing security concerns.