Governance Conditions and Local Barriers to Artificial Intelligence in Agricultural Extension and Rural Public Services in Yongding District, China

Abstract

Aims: This study examines how grassroots and local actors perceive AI-enabled agricultural extension and rural public services in Yongding District, Longyan City, Fujian Province, China, with particular attention to perceived opportunities, implementation barriers, and governance conditions. Study Design: A qualitative case-study design was used, incorporating thematic analysis with descriptive frequency coding. Place and Duration of Study: The primary study area was Yongding District, Longyan City, Fujian Province, China. Interview materials were collected between December 2025 and March 2026. Methodology: The analysis included 30 anonymised interviews, comprising 25 Yongding-focused core interviews and 5 contextual reference interviews from other Chinese localities. A multi-label thematic coding procedure was used to identify recurring themes related to AI use, agricultural extension, rural public services, and grassroots governance. Theme frequencies were calculated descriptively and were not used for statistical inference. Results: Policy diffusion, feedback channels, and institutional support formed the most frequent theme, appearing in 25 of 30 interviews (83.3%). Efficiency and administrative workload reduction appeared in 21 interviews (70.0%), followed by digital literacy, training, and human-capacity gaps in 15 interviews (50.0%), data security, privacy, ethics, and accountability risks in 14 interviews (46.7%), and local fit and practical implementation barriers in 13 interviews (43.3%). Agricultural extension and production support appeared in 8 interviews (26.7%). Participants generally viewed AI as useful for document preparation, information retrieval, data handling, agricultural advice, and routine service support, while emphasising constraints related to policy communication, training, cost, infrastructure, privacy, accountability, and local adaptation. Conclusion: AI-enabled agricultural extension in Yongding District should be understood as a local governance and implementation challenge rather than solely as a technical issue. Effective adoption requires locally appropriate applications supported by training, feedback mechanisms, infrastructure, responsible data governance, and clear accountability arrangements.