Learning from Practice: Designing Effective Regulatory Sandboxes

This paper examines sandbox initiatives across multiple jurisdictions, including in AI, data protection and financial services, to identify the core success factors that make these models effective. It distils these insights into practical recommendations for regulators designing and operating sandboxes and for participating organisations.

Building on this analysis, the paper considers the specific context of the EU AI Act, which mandates that each Member State establish an AI sandbox by August 2026, and offers targeted recommendations to support their effective implementation.

More broadly, the paper argues that the sandbox model has the potential to advance responsible innovation across the wider digital ecosystem, and recommends its adoption in areas such as online safety, cybersecurity, and data interoperability.

Learning from Practice: Designing Effective Regulatory Sandboxes

Download Now