The recently published National Data Resources Survey Report (2025) reveals that China’s annual data production volume soared to 52.26 zettabytes in 2025, reflecting a year-on-year surge of 27.28%. This growth rate marks a 2.28 percentage point increase compared to the previous year, positioning China as the source of approximately 27.44% of the world’s total data. Driven by this expansion, artificial intelligence (AI) applications have entered a new phase of large-scale deployment, with the total volume of data utilized for AI training and inference reaching 199.48 exabytes—a 42.86% year-on-year increase. Notably, for the first time, the volume of inference data has outstripped that of training data. Furthermore, the number of high-quality datasets has surpassed 110,000, with a total scale exceeding 908 petabytes, reflecting year-on-year growth of 61.13% and 142.58%, respectively. The annual volume of token calls reached approximately 21,100 trillion, with the daily average call volume skyrocketing from over a trillion at the beginning of the year to 100 trillion by the year’s end, underscoring exponential growth.