The International RegLab Project has published the first report of the first RegLab cycle, marking an important milestone in the joint effort to explore how innovative technologies, particularly artificial intelligence (AI), can be effectively, safely and transparently integrated into the nuclear sector.
RegLab is a “sandbox” activity that brings together technologists, operators, and regulators to collaboratively explore how emerging technologies progress from concept to deployment. The NEA RegLab project is tailored to address the challenges and regulatory expectations of the nuclear sector, building on experience with regulatory sandboxes in other sectors such as finance, aviation, and healthcare, as well as earlier international pilot work supported by the Electric Power Research Institute (EPRI), the International Atomic Energy Agency’s ISOP Initiative, and the Canadian Nuclear Association in 2024.
A newly released report brings together insights from the first full RegLab cycle (RegLab #1) and focuses on the use of AI technology in real-time monitoring of nuclear power plant data to identify operational inconsistencies.
Main findings
AI use case: Real-time monitoring for anomaly detection
In RegLab #1, we explored representative AI applications designed to detect anomalies in real-time operational data. Regulatory, industry, and technology community participants highlighted the potential benefits of such systems, including improved safety margins, earlier detection of deviations, and the potential to reduce operating costs.
The following two major issues became clear through the discussion:
- Explainability of AI
While explainable AI is essential to support safety justifications, participants concluded that explainability alone is not sufficient for applications with high safety implications. The report emphasizes that defense-in-depth measures remain fundamental. AI systems must be able to “show it works” with quantifiable and auditable legitimacy to support regulatory confidence. - data guarantee
Robust data assurance was identified as a key enabler for any AI-supported safety case. Participants emphasized the need for high-quality, well-maintained, and representative datasets, rather than mere data availability, to establish a reliable technical basis.
Evaluation of the RegLab approach
Participants reported that RegLab’s structure was logical, comprehensive, and effective in fostering constructive dialogue. The iterative development of hypothetical cases encouraged stakeholders to consider the problem from multiple perspectives, including regulators, operators, and technology developers.
Recommendations for future work
The report recommends establishing a working group to develop good practice guidance for an AI nuclear assurance framework based on the identified gaps. Priority areas include:
- AI verification and validation (V&V) standards.
- Clarify the practical boundaries of AI applications.
- An approach that uses defense-in-depth measures to manage residual risk.
- Enhanced training and development for both AI developers and nuclear end users. and
- Harmonize metadata structures and taxonomies to support consistency across industries.
These measures aim to support both innovation and regulatory robustness, ensuring that AI technologies can be safely and effectively deployed across a range of nuclear applications.
international cooperation
The initiative is being conducted under the auspices of the Nuclear Energy Agency (NEA) in collaboration with the Electric Power Research Institute (EPRI) and the International Atomic Energy Agency (IAEA)’s International Network on Innovations Supporting the Operation of Nuclear Power Plants (ISOP), and is supported by regulatory bodies and technical support organizations including:
- Canadian Nuclear Safety Commission (CNSC)
- Radiological Protection Agency (ASNR), France
- Nuclear Regulation Authority
- Korea Atomic Energy Research Institute (KAERI)
- Consejo de Seguridad Nucuria (CSN), Spain
- UK Office of Nuclear Regulation (ONR)
- U.S. Nuclear Regulatory Commission (NRC)
Looking to the future
As AI technology advances rapidly, the RegLab initiative is expected to play an increasing role in helping the international nuclear community understand both the opportunities and challenges associated with the application of AI technology. Insights from RegLab #1 provide the initial foundation for a harmonized approach and open a new chapter in internationally coordinated regulatory innovation.
The full report is available on Ennuvo’s website.
