Regulatory Sandboxes in Artificial Intelligence: bridging the gap between technology and law

Abstract

As technical developments of Artificial Intelligence grow in speed and complexity, ensuring the legal protection of related products and services become increasingly difficult. With legislation struggling to keep up with innovation, more flexible and technology-friendly experimental regulatory tools have been recently introduced in the European context, such as regulatory sandboxes. After discussing main legal issues arising from the integration of AI in daily life, the paper outlines the framework of AI regulatory sandboxes set out in the AI Act, as well as some potential benefits and risks associated to their employment to test the impact and legal compliance of innovative AI technologies. The paper also delves into a practical example of regulatory sandbox combining AI in the healthcare and data protection and security.