This paper examines the conflicting objectives within the European Union’s AI regulatory sandboxes established under the AI Act. Although these sandboxes are intended to promote both innovation and legal certainty, they must balance the needs of two very different stakeholders, regulators and companies. The purpose of the paper was to analyse these conflicting goals and propose step-by-step procedures that could be adopted at the European level to address them. Based on this analysis, the paper proposes two tailored procedures depending on the primary stakeholder: one designed for private companies and another for public authorities. However, the authors argue that private companies should remain the main beneficiaries of regulatory sandboxes, as they are the ultimate drivers of innovation. Therefore, it is strongly recommended to structure sandbox procedures in a way that mostly supports companies. The paper also includes graphical representations of the proposed procedures and concludes with practical recommendations for improving the sandbox model to achieve stronger technological and regulatory outcomes.
European Business Law Review