Prohibited AI surveillance practices in the Artificial Intelligence Act: promises and pitfalls in protecting fundamental rights

In Europe and all around the world, in the last years there has been an intense debate about regulating and limiting artificial intelligence (AI) surveillance technologies increasingly recognising the need to introduce prohibitions on mass surveillance practices. On 13 March 2024, the European Parliament adopted the Artificial Intelligence Act (AI Act), the first comprehensive AI legal framework in the world that addresses the risks of AI systems and introduces prohibitions on certain AI practices that violate fundamental rights and European values. In particular, AI-based surveillance and predictive policing have been a critical contention point in the negotiations of the prohibitions to be included in the AI Act. The chapter aims to evaluate to what extend the AI Act endorses fundamental-rights based approach by prohibiting unacceptable AI practices. It will reveal that the AI Act sets only partial prohibitions on AI-based surveillance, predictive policing and other harmful practices as they contain wide exceptions. Regarding remote biometric identification the new provisions of the AI Act, rather than introducing clear prohibitions, seems to reflect the essential guarantees set by the European Court of Human Rights (the ECtHR) for assessment whether interference with the right to privacy and data protection, entailed by mass surveillance measures, including facial recognition, can be justified. The chapter argues, that effective enforcement of fundamental rights and data protection law will continue to play a crucial role in putting the AI Act in practice to prevent mass surveillance and other unacceptable AI practices.