The Digital Regulatory Framework through EU AI Act: The Regulatory Sandboxes’ Approach

In the past years, the EU political agenda has been marked with an accelerated digital transformation programme intended to provide for both a safety net for the implementation and use of technology and the regulation and democratisation of access to technology. In this context, the regulation of artificial intelligence (AI) has been flagged as a priority considering how it has contributed to a transformation on both business activities and legal practice. The EU AI Act currently under discussion has the advantage of marking a pioneer regulatory intervention laying down harmonised rules on AI, shaping both how AI is to be used as well as its risks. From a legal perspective, the EU AI Act also presents as an opportunity to launch a new digital regulatory paradigm which deals with the difficulties of the traditional regulatory framework in balancing the need to regulate a constantly changing reality alongside the creation of a regulation that is not excessively burdensome for developers and providers. On the one hand, new regulatory principles of function equivalence and technology neutrality have been implemented as application and interpretation criteria, while on the other hand, new supervisory and regulatory technologies are being experimented to align digital monitoring and reporting requirements. Accordingly, discussions around the so-called digital law philosophy, legal thinking and legal design mechanisms are to guide a new construction of how the legal field is to adapt to the digital era. Despite being a cross-cutting regulatory approach towards digital fields (as digital finance), the proposed AI regulation stands out in this domain for its assumed break with typical regulatory standards by adopting a risk-based approach (as opposed to a regulation based of rights) or implementing regulatory sandboxes aimed to test and develop new AI technologies. In this article, we propose to address the main features of the new digital regulatory paradigm having the EU AI Act in the horizon. In particular, we propose to discuss the underlying rationale of a sandbox approach and how it fits the development of emerging technologies as AI. Keywords: Digital regulatory framework; EU; EU AI Act; regulatory sandboxes; innovation.