Artificial Intelligence (AI) is increasingly embedded in China’s environmental supervision, reshaping how environmental risks are detected and regulated. Existing research mainly focuses on technical performance or isolated policy initiatives, while paying limited attention to the institutional sustainability of regulatory systems integrating AI, understood as the capacity to operate consistently, transparently, and accountably over time. This article examines how AI becomes institutionally embedded in environmental supervision. It focuses on three dimensions: the formation of regulatory evidence, the allocation of responsibility, and the exercise of administrative capacity. Drawing on qualitative analysis of legal and policy documents issued between 2017 and 2025 and two case studies (AI-enabled air-quality governance in Beijing and AI-enabled water-quality monitoring in the Yangtze River Basin), the study shows that AI deployment generates recurring governance tensions, including opacity in algorithmic evidence formation, fragmented accountability chains, and uneven administrative capacity. The article argues that sustainable AI-enabled supervision depends less on technological intensification than on institutional and governance conditions ensuring transparency, reviewability, and responsibility in routine regulatory practice, thereby contributing to debates on algorithmic regulation and providing policy-relevant insights for maintaining sustainable environmental governance in rapidly digitalizing regulatory contexts.
Zuo et al. (Fri,) studied this question.