Regulating artificial intelligence in the EU

Abstract

The EU AI Act marks a significant milestone as the world’s first comprehensive legal framework for artificial intelligence. This pioneering legislation aims to regulate the development, deployment, and use of AI systems to foster innovation while ensuring safety, transparency, and adherence to fundamental rights. The chapter will analyse the AI Act in light of the underlying goal of ‘regulation for innovation’, which seeks to enhance the adoption of AI within the Internal Market by bolstering trust in technological advancements. The narrative provides an overview of the AI Act, outlining its framework and key components. It delves into the Act’s adoption of a risk-based regulatory approach, which precludes AI systems that pose an unacceptable risk, heavily regulates AI systems that pose high risks, and lightly regulates systems posing limited or minimal risks. The chapter further explores other essential aspects of the Act, including the regulation of general-purpose AI models, requirements and obligations like conducting Fundamental Rights Impact Assessments, the role of standards and benchmarks, regulatory sandboxes, and the functions of the European AI Office and AI Board. These elements underscore the Act's potential to harmonise AI regulation within the Internal Market and influence the future landscape of AI in Europe and beyond. The discussion refers to the achievements and challenges faced by the EU legislator in creating a unified regulatory environment, in light of the potential impacts on the digital single market and the EU’s competitiveness on the global stage. It assesses how the Act navigates the balance between fostering innovation and safeguarding fundamental rights. In summary, this chapter offers an overview of the AI Act, focusing on selected aspects that are pivotal in shaping the Internal Market and influencing global AI regulation frameworks. It also discusses the protection of fundamental rights within these regulatory structures.