Operationalising AI Regulatory Sandboxes under the EU AI Act: The Triple Challenge of Capacity, Coordination and Attractiveness to Providers

2025-09-07

AI Summary

The EU AI Act mandates Member States to establish national AI regulatory sandboxes, yet faces three interrelated challenges: insufficient regulatory capacity, heterogeneous national designs (risking “sandbox arbitrage”), and limited innovator uptake under voluntary participation—constrained by weak confidentiality safeguards, absence of legal exemptions, and no presumption of compliance. This study employs legal-policy analysis, comparative institutional design assessment, and compliance pathway evaluation to systematically identify structural tensions impeding sandbox effectiveness. It proposes an innovative, Commission- and AI Board-led collaborative governance framework, advocating for harmonized operational guidelines, cross-border mutual recognition mechanisms, and enhanced functional design to strengthen legal flexibility and compliance incentives. The findings offer actionable institutional pathways to reconcile AI innovation promotion with regulatory coherence and enforceability across the EU.

Technology Category

• Philosophy and Ethics of AI: AI & Law, Justice, Regulation & Governance
• Humans and AI: AI for Accessibility
• Intelligent Robots: Embodied AI

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Responsible Web: Technology governance, policy, and regulations
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The EU AI Act provides a rulebook for all AI systems being put on the market or into service in the European Union. This article investigates the requirement under the AI Act that Member States establish national AI regulatory sandboxes for testing and validation of innovative AI systems under regulatory supervision to assist with fostering innovation and complying with regulatory requirements. Against the backdrop of the EU objective that AI regulatory sandboxes would both foster innovation and assist with compliance, considerable challenges are identified for Member States around capacity-building and design of regulatory sandboxes. While Member States are early movers in laying the ground for national AI regulatory sandboxes, the article contends that there is a risk that differing approaches being taken by individual national sandboxes could jeopardise a uniform interpretation of the AI Act and its application in practice. This could motivate innovators to play sandbox arbitrage. The article therefore argues that the European Commission and the AI Board need to act decisively in developing rules and guidance to ensure a cohesive, coordinated approach in national AI regulatory sandboxes. With sandbox participation being voluntary, the possibility that AI regulatory sandboxes may prove unattractive to innovators on their compliance journey is also explored. Confidentiality concerns, the inability to relax legal rules during the sandbox, and the inability of sandboxes to deliver a presumption of conformity with the AI Act are identified as pertinent concerns for innovators contemplating applying to AI regulatory sandboxes as compared with other direct compliance routes provided to them through application of harmonised standards and conformity assessment procedures.

Problem

• Addressing capacity-building challenges in national AI regulatory sandboxes
• Ensuring coordinated approaches to prevent inconsistent AI Act interpretation
• Resolving attractiveness concerns for innovators in sandbox participation

Innovation

• National AI regulatory sandboxes for testing validation
• European Commission guidance ensuring cohesive coordinated approach
• Addressing confidentiality legal rigidity conformity presumption concerns