AI applications are rapidly expanding, both in terms of their prevalence and economic value, reflecting significant investments by the European Union in this field and related research. While AI-driven innovations hold the promise of substantial benefits, they also introduce notable and concerning risks for data protection, many of which are challenging to anticipate and are only partially addressed by the General Data Protection Regulation (GDPR). This situation creates an urgent need to determine whether and how regulators should intervene to balance the advancements in AI with the imperative of data protection.
Regulatory sandboxes offer a promising solution to this challenge. A regulatory sandbox can be defined as a controlled environment where innovative products, services, or processes are tested by selected private entities under the oversight of a regulatory authority. The insights gained from these sandboxes can help refine future regulations and support compliance with existing ones.
The project REgulatory sandBOx fOr a Trustworthy Artificial Intelligence (REBOOT.AI) aims to develop a legal and theoretical framework to assist Member States and their data protection authorities in establishing regulatory sandboxes specifically designed to promote the development of trustworthy AI. This framework will be made readily accessible through a set of recommendation guidelines, grounded in both theoretical and empirical research.
To achieve this, the project examines two key regulatory sandboxes within the European Economic Area that focus explicitly on AI: those overseen by the Norwegian Data Protection Authority and the French Data Protection Authority. By establishing a standardized framework for AI-focused regulatory sandboxes, REBOOT.AI aims to facilitate the broader adoption of trustworthy AI applications across Europe, thereby making a significant contribution to the Horizon Europe Action (HORIZON-CL4-2021-HUMAN-01-02).