Regulatory sandboxes: innovation under EU AI Act supervision
Regulatory sandboxes under Art. 57-61 are controlled environments, supervised by the national authority, in which organizations can develop and test innovative AI systems with guidance and temporary relief from certain administrative requirements, without suspending the material safeguards or incident-reporting duties.
Updated: June 2026
Introduction: the sandbox as innovation policy tool
The EU AI ActEU AI ActRegulation (EU) 2024/1689, the European Union's law on artificial intelligence. It takes a risk-based approach: prohibited practices, requirements for high-risk AI systems, transparency obligations for specific uses, and a separate regime for general-purpose AI models. Obligations are divided between providers and deployers. See general-purpose AI, conformity assessment.Open full entry →'s regulatory sandboxregulatory sandboxA supervised scheme run by a regulator that lets organizations test an AI system under controlled conditions before full deployment.Open full entry → framework (Art. 57–61) acknowledges a tension inherent in technology regulation: rules designed to address today's risksriskIn the EU AI Act's terms, the combination of the likelihood that a harm occurs and the severity of it if it does. The link between a principle (via the harm that would breach it) and a control (the measure that reduces it). Naming the harm and assessing its risk is required by Art. 9 before any mitigation measure is chosen. See harm, control, residual risk.Open full entry → can inadvertently prevent tomorrow's innovation. Regulatory sandboxes are the legislative response, a structured environment in which AI systemsAI systemA machine-based system that, for explicit or implicit objectives, infers from input how to generate outputs (predictions, content, recommendations or decisions) that can influence physical or virtual environments. The OECD-style definition followed by the EU AI Act.Open full entry → can be developed, tested, and validated under regulatory supervision, with reduced compliance obligations during the development phase.
What is a regulatory sandbox?
A regulatory sandbox under the EU AI Act is a controlled testing environment established by national competent authorities. Within the sandbox, participants may develop and test AI systems, including potentially high-risk systems, without the full compliance obligations that would normally apply to market deployment. The sandbox is time-limited and operates under a supervised framework: the authority monitors the testing and participants must comply with the terms of their sandbox agreement.
Who can apply?
Art. 57 provides that national authorities shall establish at least one regulatory sandbox within each member state. Priority access is given to:
- Startups and SMEs
- Micro-enterprises
- Innovative organizations developing novel AI approaches
Larger organizations may access sandboxes, but the selection criteria explicitly prioritize smaller innovators with limited resources to navigate full compliance frameworks.
Benefits of sandbox participation
- Reduced compliance burden during testing: Sandbox participants are not required to meet all Art. 9–27 obligations during the testing phase
- Regulatory guidance: Direct access to supervisory authority expertise, invaluable for navigating grey areas in the regulation
- Liability protection: Art. 57(12) provides that good-faith sandbox participants are not fined for infringements identified during testing, subject to conditions
- Regulatory certainty post-sandbox: Testing under supervisory oversight builds the evidentiary record that supports full market deployment approval
The dutch sandbox
In the Netherlands the AI regulatory sandbox is being set up jointly by the AP and RDI from 2026 and is not yet operational. No market surveillance authoritymarket surveillance authorityThe national body that enforces the AI Act in a member state, with powers to investigate, order corrective action and apply penalties.Open full entry → has been formally designated yet; the draft Uitvoeringswet AI-verordening (UAIV) proposes the AP and RDI as coordinating supervisors. Organizations interested in participation should watch for the AP and RDI's opening announcements.
Sandbox vs market deployment
The sandbox does not provide a permanent exemption. Once testing concludes, organizations must achieve full compliance before market deployment. The sandbox is best understood as a fast-track pathway to compliance, with regulatory guidance that accelerates the journey, rather than an alternative to compliance.
Compliance checklist
- Are any of your AI development projects at a stage where sandbox participation would be beneficial?
- Have you assessed whether your organization qualifies for priority sandbox access (SME/startup)?
- Have you reviewed the AP's sandbox program terms?
- Is there an internal process for evaluating sandbox applications?