Under the dual carbon goals, the digital economy has become a crucial driver promoting the green transformation of agriculture. Using panel data from 30 provinces in China from 2015 to 2023, this study explores the impact mechanism and spatial spillover effects of digital economy development on agricultural carbon emission intensity through benchmark regression models, moderating effect models, and the spatial Durbin model. The findings are as follows: (1) From 2015 to 2023, the digital economy index of China’s 30 provinces showed overall rapid growth, with the growth rate decreasing from east to west. The gap between provinces with high and low levels of digitalization continues to widen. In terms of temporal evolution, agricultural carbon emission intensity exhibited a steady downward trend overall. From a regional distribution perspective, agricultural carbon emission intensity in central and western regions is significantly higher than in eastern provinces, and gradient differences in carbon reduction levels among provinces are quite pronounced. (2) In eastern regions, the scale characteristics of agriculture lead to a more prominent positive effect of the digital economy on agricultural carbon emission intensity. In central regions, agricultural carbon emission intensity shows a declining trend. In western regions, the pace of digitalization is relatively slower compared to the other two regions, and its impact on agricultural carbon emission intensity is not significant. (3) Moderating effect analysis reveals that increased financial support for agriculture and the advancement of urbanization can indirectly curb agricultural carbon emission intensity by promoting digital development. The decomposition results of spatial spillover effects indicate that rural residents’ disposable income significantly inhibits agricultural carbon emission intensity, and this effect is fully reflected in the total effect, direct effect, and indirect effect, demonstrating both local and cross-regional transmission characteristics in its impact on agricultural carbon reduction. (4) The inhibitory effect of digital economy development on agricultural carbon emission intensity is particularly significant in eastern regions, major grain-producing areas, and crop farming zones. Therefore, the state should enhance financial support for agriculture, increase farmers’ income, and fully leverage the positive role of the digital economy in technology penetration, industrial structure adjustment, and urbanization to reduce agricultural carbon emissions. Meanwhile, when formulating differentiated carbon reduction policies, the government should clarify the core direction to ensure that the digital economy can exert its low-carbon positive externality, thereby contributing to high-quality agricultural development.