Rice is a staple food for a significant portion of the global population. Arsenic (As) accumulated in rice grains influences rice quality which threatens human health. In this study, we used three machine learning models to predict arsenic accumulation in rice based on over 300 surveys. The prediction results of soil arsenic indicate that high-arsenic soil areas are mainly distributed in South and Southeast Asia such as India, China, and Thailand. In addition, higher bioaccumulation factors (BAF), associated with higher temperature, are predominantly observed in eastern India and southern Myanmar. However, arsenic content in soil is relatively lower in these areas. About 5.5 billion population may be threatened by the consumption of high-arsenic rice. It can be concluded that temperatures may influence the BAF except for soil arsenic, and soil physicochemical properties. Further research on the relationship between climate parameters and BAF should be conducted to address and adapt to future climate change. Additionally, understanding the mechanism of arsenic accumulation under different climatic conditions is crucial for developing agricultural technologies to reduce arsenic accumulation in rice.