Resilience Governance in Nuclear Safety: Insights from China’s AI-Driven Innovative Practices

Artificial intelligence (AI) is increasingly deployed in nuclear power operations to enhance safety and resilience, yet systematic empirical assessments of its impact on risk governance remain limited. This study addresses this gap by developing and applying a resilience governance framework based on the triad of robustness, redundancy, and rapidity to evaluate AI applications through an analysis of case studies from Chinese nuclear power plants (2020–2025). The findings demonstrate that AI enhances system robustness via proactive policy formulation and emergency planning, enables redundancy in operational systems, and significantly improves response speed during emergencies through AI-driven inspection robotics. However, the integration of AI is neither universally effective nor without risk; limitations include context-specific performance variability and emerging technical and ethical challenges. The study concludes with recommendations for adaptive regulatory frameworks, rigorous validation of safety-critical AI systems, and inclusive global governance mechanisms to ensure equitable participation in the AI-driven evolution of nuclear safety. This work advances the discourse from conceptual promise to empirically grounded, risk-informed governance.