LSTS researchers Nathan Genicot and Thiago Guimaraes Moraes have published a new article in the Cambridge Forum on AI: Law and Governance examining how AI regulatory sandboxes will function under the EU AI Act.
The AI Act introduces regulatory sandboxes and real-world testing as tools to support AI innovation, but their roles and limits remain unclear. This paper analyses the legal framework for sandboxes, focusing on how much flexibility they truly offer and how they interact with other experimentation mechanisms. It assesses whether sandboxes allow lighter regulatory requirements or mainly provide structured dialogue with authorities. The study also contrasts sandboxes with real-world testing: the former centres on legal guidance, while the latter removes regulatory barriers to ease market entry. Key challenges include how Articles 57 and 60 relate, what compliance each mechanism requires, and how regulators should oversee them. The authors conclude that clearer guidance from the European Commission is needed to resolve inconsistencies and ensure effective implementation.
Consult the article here: https://www.cambridge.org/core/journals/cambridge-forum-on-ai-law-and-governance/article/exploring-the-boundaries-of-ai-regulatory-sandboxes-under-the-ai-act-flexibility-and-realworld-testing/33039F16B76448F0EA86699385FD799E