The EU AI Act: a proactive framework for comprehensive AI regulation

Abstract

The rapid advancement of artificial intelligence (AI), enabling autonomous decision-making and self-learning, is transforming legal systems and societal structures, while presenting significant governance challenges. Regulators grapple with the Collingridge Dilemma, uncertainty hinders early intervention, and entrenchment complicates later action, exacerbated by the pacing problem, where AI’s exponential growth outpaces static laws, rendering traditional command-and-control regulation inadequate. The European Union’s AI Act (Regulation (EU) 2024/1689) introduces proactive tools like regulatory sandboxes and data governance, yet struggles to balance innovation with ethical oversight. This paper explores proactive law, an anticipatory and collaborative approach emphasizing stakeholder engagement and adaptive governance, as a viable paradigm for AI regulation. Through doctrinal, comparative, and socio-legal analysis of the EU AI Act alongside USA and Chinese strategies, focusing on transparency mandates and codes of practice, the study discusses a refined proactive model to mitigate risks, foster trust, and sustain innovation, advancing global responsible AI governance.