A cross-border regulatory learning platform for novel foods.

The EU’s pre-market authorisation process for novel foods guarantees high food-safety standards, but the approval process itself lacks a deeper understanding of the applicant as the user of the regulatory process.

Because the approval process is centralised at EU level, the understanding of scientific expectations and regulatory practice varies significantly across Member States. This makes it even more challenging for Member States to offer applicants appropriate guidance.

Furthermore, there is no institutionalised mechanism for sharing what Member States and applicants learn at a national level with the European regulator. As a result, understanding stays local and fragmented, rather than improving the system as a whole.

EFSA pre-submission advice, expanded under the proposed Biotech Act I, assists individual applicants but operates on an application-by-application basis; it cannot build shared understanding across the system or improve the efficiency of the regulatory process.

Not pre-market authorisation, but cross-border regulatory learning in order to increase regulatory efficiency

Although they do not constitute pre-market authorisation procedures, regulatory learning platforms such as the New Foods RegLab can address the lack of knowledge transfer and cross-border learning at a system-wide level.

Inspired by the sandboxes already used by the AI and fintech sectors, the New Foods RegLab aims to showcase how cross-border learning platforms can help to achieve regulatory efficiency for a key strategic technology while continuing to guarantee high food-safety standards.

The New Foods RegLab has been designed to attract industry by supporting the sharing of frameworks, guidance, and best practices, while feeding those learnings back into EFSA advice. As a learning platform, it is not an authorisation procedure and does not lower safety standards – it simply changes how the various parties learn from one another and helps to facilitate a more efficient transfer of knowledge.

A targeted enabling provision could establish this cross-border regulatory learning within the Biotech Act I proposal, while Biotech Act II could give it a durable legal basis and EU-level support, allowing willing Member States to begin without obliging others to join, and without changing the authorisation procedure itself.

Europe does not have an innovation problem; it has an implementation problem. Its regulatory systems are still too often fragmented, and therefore slow and inefficient.

Although a lack of regulatory speed is not exclusively a European problem, Europe can no longer afford for the commercialisation and scale-up of strategic technologies to be held back by unpredictable time to market and fragmented regulatory frameworks. As a result, Europe keeps investing in R&D, only to keep losing key technologies and their innovators mostly to the US market.

From input to food processing – a dangerous level of dependency from feed to food

At the same time, China is establishing technological leadership also in emerging food-and-feed biotechnologies. The strategic consequence is that Europe’s dependencies are continuing to increase, while industry continues to leave – putting the Union’s security, economic capabilities, and autonomous food-and-feed production at risk. Europe still treats biotechnologies for food and feed production as though it could afford to cede them to China, and that stance is driving dangerous levels of dependency.

The scale of the challenge

As more and more citizens become aware of these critical competitiveness and security issues, the urgent need for quickly applicable solutions is becoming more widely apparent. As Mario Draghi rightly pointed out:

This Future Prototype therefore focuses exactly on this: a new mechanism of enhanced collaboration. A cross-border regulatory learning platform based on shared frameworks to make Regulation (EU) 2015/2283 on novel foods work within the regulated timeline.

Making the novel-foods regulation work through enhanced collaboration is a security obligation

Solving the novel-food deadlock is clearly possible and failing to do so is a serious security liability for Europe. This is not about a conflict between technology and agriculture – it is about the importance of domestic food production as a strategic asset for Europe.

The prototype from the New Foods RegLab was born out of the conviction that Europe cannot afford to become further polarised and divided by proxy wars initiated by small groups, given what is at stake. Making the regulatory approval process for novel foods work within its regulated timeline is a security obligation and needs to be recognised as a political priority in this new geopolitical era.

Beyond novel foods: no European investment in strategic technologies without faster time to market

This Future Prototype focuses on learning platforms for novel foods. However, in our experience, such platforms are needed for all strategic technologies in which Europe invests.

The successful deployment and commercialisation of many key technologies developed in Europe – technologies that are essential for decarbonisation, strategic autonomy, and security – depend on the implementation of more efficient regulatory strategies that significantly reduce time to market. The design principles behind RegLabs as enhanced collaboration mechanisms can contribute to this significantly.

Europe excels in food innovation and research but delivers poor results in terms of time to market and commercialisation. In the case of novel foods, applicants, consultants, and national assessors each learn the same lessons separately, with little of that understanding captured, shared, or channelled back to the regulator at the European level.

‘Efficiency’ here does not mean lower standards or fewer requirements: it means that the parties to the process understand what the rules require. The current lack of mutual understanding leads to a familiar deadlock in which promising work stalls, not for want of scientific progress, but due to a lack of shared understanding. This fragmentation is not limited to the market approval of novel foods and extends into many other sectors.

Comparison of the EU’s novel-foods approval process to other jurisdictions: timelines and costs

When comparing the EU’s novel-food approval process to other jurisdictions, the regulated timeline is comparable.² In practice, reaching an EFSA opinion takes an average of two-and-a-half years,³ about three times what the rules allow.⁴ The timing is also highly unpredictable. In some cases, the opinion alone has taken more than five years, and the Commission’s authorisation procedure still has to follow.

That gap is why, to date, companies have targeted the United States as their first market. The US framework exempts substances that are generally recognised as safe from pre-market approval, and time to market through the GRAS route is much shorter than in the EU.

Similarly, the costs of generating data to support a novel-food application in the EU are significantly less predictable than in other regions, due to the cost of additional data requests that occur during the risk-assessment period. One of the main reasons for the increased costs associated with an EU dossier is that EFSA requires data from at least five independent batches of the novel-food ingredient, as well as stability testing covering at least the proposed shelf life of the ingredient in relevant food matrices, whereas other regions are less prescriptive.⁵ This adds cost and complexity to the initial submission.

Pre-submission advice and the applicant as a regulatory user

The Biotech Act I proposal expands EFSA’s role by expanding the scope of pre-submission advice. This is valuable, but it only engages one applicant at a time and does not move knowledge between applicants or build a shared understanding between Member States.

Effective pre-submission advice requires more than case-by-case guidance; it depends on a deeper structural understanding of the applicant’s user journey through the regulatory system.

The recurring difficulties in novel-foods dossiers are largely systemic: practices are uneven, and understanding of scientific expectations and regulatory knowledge about novel foods varies from one country to the next, and between consultants. An independent analysis of 292 applications found that reaching an EFSA opinion takes two and a half years on average, with much of that time spent on administrative steps and on answering requests for further data. Its authors recommend clearer guidance and better pre-submission support.⁶ These are problems of knowledge transfer, not of the authorisation rules.

Beyond fragmentation: tear down those walls

As things currently stand, applicants, investors, the regulators, and Member States all work in separate silos, each with only a partial view of the process. Interviews we have conducted with industry and selected national authorities across Member States indicate that these silos, combined with the loss of highly specialised expertise in novel foods regulation at Member State level, have led to a major drop in regulatory knowledge of novel foods, and consequently to a loss of regulatory efficiency. Member States can no longer properly support applicants, who then increasingly rely on consultants who often lack hands-on experience in the preparation of dossiers.

How to design for regulatory efficiency? EFSA’s expanded role

Biotech Act I proposes an expanded EFSA role, including broader pre-submission advice, but this is still under negotiation between the Council and Parliament, and how it would work in practice remains to be seen. In the meantime, with no regulatory cross-border peer learning possible across Europe, and no sandbox available for novel foods under the current proposal, European food-tech innovators continue to scale up mainly in the United States.

Prototyping the future: the New Foods RegLab as a cross-border regulatory learning platform

An open cross-border regulatory learning platform brings these stakeholders into a single structured learning space, so that regulatory requirements are understood collectively, rather than discovered in a fragmented and ad hoc manner. The New Foods RegLab fills this gap and addresses these issues from a systems perspective, while Biotech Act I is being negotiated and before Biotech Act II provides a durable legal basis.

Learning from AI and fintech: cross-border learning models in EU law

Cross-border learning models are not new in EU law. National AI regulatory sandboxes, under the AI Act, as well as the European Forum for Innovation Facilitators in financial services, both rely on a permissive structure: willing Member States take part and share learning centrally without requiring all Member States to join. Under the AI Act, a Member State can even meet its obligation by joining a sandbox run by another Member State.⁷ Once the framework is built, others can opt into it.

Not a regulatory sandbox, but essential for improving the quality of novel-food dossiers

Inspired by the European regulatory sandboxes, such platforms, although not themselves sandboxes, can facilitate a cross-border regulatory-learning infrastructure that willing Member States could adopt within existing joint frameworks.

This Future Prototype can help to improve the quality of dossiers and to accelerate approval within the regulated timeline. Furthermore, it can generate the learnings needed to set up a food-and-biotech regulatory sandbox in the future. It is applied here to novel foods as an emerging technology but can be extended to other critical areas.

As a practical way forward, it aims to support the development of future pre-submission advice by providing an urgently needed user-centred and structured learning space in which to compare approaches, test how the rules are understood, and build common ground before, and independently of, any individual authorisation.

Using design methods for policymaking is not a new idea. However, it is not yet common practice to embed design and innovation methods in policymaking: identifying where a system fails its users, prototyping and testing solutions, and opening new routes to change.⁸

Central to this approach is understanding the problem space before moving to solutions, and working with the people inside the process.

While design methods have been used mostly to build public services around the citizen, the New Foods RegLab uses them to build regulatory learning around the applicant as a regulatory user. The cross-border models in the AI Act and in financial services already show that willing Member States can act together. What we propose is to start working and learning with the users in order to improve the efficiency of the novel-foods approval process, ensuring higher dossier quality and saving resources on both sides.

Designing the pre-submission advice with its users

A learning platform, such as the New Foods RegLab, works if it is built around the people who use the regulatory system, not just the process and its guidance. The New Foods RegLab concept is based on six design principles. While some of these principles have been developed specifically for novel-food applications, others are applicable far beyond the scope of food innovation: