This is an old revision of the document!
Governance, Innovation & Change
Regulatory sandbox
Regulatory sandboxes provide a controlled, time-limited space in which innovators can test new products, services, or approaches with reduced regulatory requirements, while regulators observe and gather evidence to inform future frameworks.
Sandboxes let innovators test in real markets before rules exist — and let regulators learn what rules are actually needed.
Why this matters
The energy system is undergoing rapid transformation driven by digitalisation, decarbonisation, and the proliferation of distributed resources. Regulatory frameworks, designed for a different system architecture, often constitute barriers to new technologies and business models before sufficient evidence exists to regulate them fully. Regulatory experimentation tools, of which sandboxes are the most widely used, offer a structured way to generate that evidence in real-world conditions while maintaining consumer and system protection.
To keep pace with innovation, regulation needs to learn from experimentation. Sandboxes are one answer to this challenge.
Shared definitions
Three related but distinct concepts underpin this topic. They are often conflated but operate at different levels.
Regulatory experimentation is the umbrella term for any structured test or trial of a new product, service, approach, or process designed to generate evidence that can inform the design or administration of a regulatory regime.1) It includes sandboxes, pilot regulations, testbeds, and living labs. The shared logic is that evidence precedes rule-making rather than following it.
A regulatory sandbox is the most structured form of regulatory experiment. It provides a temporary, limited exemption from specific regulatory requirements — or a streamlined regulatory process — within a supervised environment. Regulators actively monitor the sandbox, set conditions to uphold consumer protections, and can modify or close it if risks emerge. Crucially, the sandbox starts from an innovation seeking regulatory accommodation: the question being tested is whether the innovation can work safely within a modified regulatory environment.2)
Regulatory learning is the outcome that experimentation tools are designed to produce: the process by which regulators update frameworks, knowledge, and practices on the basis of evidence generated through experimentation. A sandbox that yields positive results still requires a formal regulatory process to translate findings into permanent change. The quality of that translation determines whether experimentation produces durable impact. Regulatory learning is therefore not automatic — it requires structured mechanisms to capture sandbox outcomes and feed them into rule-making.3)
The relationship between the three: regulatory experimentation is the activity, the regulatory sandbox is one specific tool for conducting it, and regulatory learning is the intended result.
Experimentation tools
Regulatory sandboxes are one of several tools available for regulatory experimentation. The EC Staff Working Document (2023) groups these by their primary focus:4)
- Testbeds focus on technical development and testing in controlled near-real-world conditions, with primary motivation to develop and upscale innovations.
- Living labs operate in uncontrolled real-world or virtual environments, revealing hidden user needs and potential social impacts, and providing foresight about future socio-technical systems.
- Regulatory sandboxes test innovations and regulations in controlled real-world market conditions to improve legal certainty, focusing on technologies that are mature enough for market deployment.
While these are distinct tools, they can be combined. Synergies between them reinforce both innovation and regulatory learning.5)
Logics of experimentation
The term “experiment” covers fundamentally different approaches to knowledge generation. Ansell and Bartenberger (2016) identify three distinct experimental logics:6)
Table 1. Three logics of regulatory experimentation.
Source: Ansell & Bartenberger (2016).
| Logic | Primary aim | Approach | Allowance for failure |
|---|---|---|---|
| Controlled | Isolate causality | Deductive; settings controlled as much as possible | High (researcher neutral to outcome) |
| Darwinian | Enhance systemic innovation | Inductive; variation more important than control | Very high (few variations will succeed) |
| Generative | Generate new solution concepts | Abductive; iterative refinement toward success | Low (researchers strive for success) |
Regulatory sandboxes in the energy sector tend toward the generative logic: the aim is not to test a falsifiable hypothesis but to iteratively develop a viable product, service, or business model within a regulatory environment. Controlled logic applies where pilot regulations test specific policy measures. Darwinian logic characterises innovation tender systems that run multiple parallel experiments and select successful approaches.
Table 2. Key terms in regulatory experimentation.
| Term | Definition |
|---|---|
| Regulatory experimentation | The umbrella category for structured tests or trials designed to generate evidence that can inform regulatory design or administration. Includes sandboxes, testbeds, living labs, and pilot regulations. |
| Regulatory sandbox | A supervised, time-limited framework granting partial exemption from regulatory requirements to allow real-world testing of innovations. Starts from an innovation seeking accommodation; generates evidence for regulatory design.7) |
| Regulatory learning | The process by which regulators update frameworks, knowledge, and practices on the basis of evidence generated through experimentation. The intended result of regulatory experimentation — but not automatic; requires structured translation mechanisms.8) |
| Pilot regulation | A limited trial of a new regulatory measure with a specific group or area. Distinct from a sandbox: the regulation itself is being tested, not an exemption from it. |
| Regulatory innovation zone | A broader, geographically defined framework adapting regulatory conditions across multiple sectors to support innovation ecosystems, rather than testing a single product or service. |
| Testbed | A controlled technical environment for developing and testing innovations, with primary focus on technical rather than regulatory learning. |
| Living lab | An open, real-world or virtual environment for testing innovations with users, with primary focus on revealing social needs and socio-technical dynamics. |
Perspectives
Regulatory sandboxes sit at the intersection of actor strategies, technical infrastructure, and institutional design. Who initiates, who monitors, and who translates findings into rules all shape whether a sandbox produces durable regulatory learning.
Actors and stakeholders
Sandboxes involve at minimum the regulator (who sets conditions and monitors), the innovator (who tests the product or business model), and affected consumers or users (who are protected by sandbox conditions). In energy sector sandboxes, system operators and network companies are often also involved, as the innovation typically requires grid interaction. Effective sandboxes require clear roles for each actor and mechanisms for consumer redress if the experiment causes harm.
EU – Horizon Europe Cluster 5
Research consortia are required to integrate societal readiness assessment alongside regulatory considerations, reflecting a broadened view of who bears the costs and benefits of regulatory experimentation.9)
Technologies and infrastructure
In the energy sector, regulatory sandboxes are frequently triggered by technologies that cross the boundary between generation, storage, and demand — distributed resources that existing rules do not cleanly accommodate. Smart metering, peer-to-peer trading platforms, vehicle-to-grid services, and community energy sharing schemes have all been the subject of sandbox experiments because they require both technical integration and regulatory accommodation simultaneously.10)
Italy – regulatory sandbox for distributed energy
Among the earliest EU energy sector sandbox initiatives, the Italian regulatory authority ARERA developed structured experimental arrangements for testing distributed generation and storage configurations that did not fit existing network connection rules.11)
Institutional structures
Institutional readiness is as important as technical maturity for sandboxes to function. IR/TRL alignment at risk gates determines whether a technically ready innovation can actually be tested: missing grid codes, unclear liability rules, or absent data governance frameworks can make a sandbox inoperable even when the technology works. The Austrian Energie.Frei.Raum sandbox addressed this by co-designing the sandbox framework with regulatory authorities from the outset, building the institutional conditions for experimentation rather than assuming they existed.12)
Regulatory experimentation is unevenly distributed across EU member states. Based on data collected through 2023, initiatives have been adopted or are under development in twelve member states, with a further three considering adoption.13)
Austria – Energie.Frei.Raum
A regulatory sandbox framework designed to bridge the gap between technology and institutional readiness. It allows for testing tariff models and market rules under controlled experimental conditions before permanent legislation, and was co-designed with regulatory authorities to ensure institutional workability from the outset.14)
Distinctions and overlaps
Regulatory sandbox vs. pilot regulation
A regulatory sandbox suspends or streamlines existing rules to allow an innovation to be tested; a pilot regulation tests a proposed new rule on a limited group. The sandbox starts from an innovation seeking regulatory accommodation; the pilot regulation starts from a regulatory proposal seeking empirical validation. The two can be combined but address different knowledge gaps.
Regulatory sandbox vs. innovation zone
A regulatory sandbox is narrow in scope and time-limited, focused on a specific product, service, or model. A regulatory innovation zone is a broader, place-based arrangement that creates favourable conditions across multiple sectors and actors. In practice the two are often combined, with zones hosting multiple concurrent sandboxes.
Regulatory experimentation vs. regulatory reform
Experimentation tools generate evidence; they do not themselves constitute regulatory reform. A sandbox that yields positive results still requires a formal regulatory process to translate findings into permanent change. Regulatory learning — the outcome experimentation aims to produce — requires structured mechanisms to capture findings and feed them into rule-making.