The AI Act isn’t enough: closing the dangerous loopholes that enable rights violations

The EU's AI Act, designed to regulate high-risk AI systems, faces criticism for significant loopholes that could permit unchecked use of AI in areas like national security and law enforcement. These exemptions raise concerns about potential mass surveillance and discriminatory practices, underscoring the ongoing debate over the balance between technological advancement and fundamental rights. The central issue revolves around how strictly the AI Act will be enforced, particularly within sensitive domains. A new report from EDRi affiliate Danes je nov dan focuses on Slovenia and highlights the need for stricter national safeguards to mitigate the risks associated with the AI Act's loopholes. The recommendations emphasize the necessity of transparent oversight mechanisms to ensure accountability and prevent abuses of AI systems. This is especially crucial given the potential for these systems to be deployed in contexts where individual rights could be significantly impacted. These loopholes could affect various groups, including protesters and migrants, who may face increased surveillance or discriminatory treatment. Broader implications involve the erosion of trust in digital technologies and a potential chilling effect on freedom of expression. The European digital sovereignty angle is relevant here, since these risks undermine the EU's aim to be a leader in ethical AI development.