We developed a new global dataset that provides daily seamless observations of XCO₂ over land from 2003 to 2022. Using artificial intelligence to integrate multiple satellite missions, atmospheric reanalysis, and environmental data, we filled data gaps and ensured continuity across satellite records. The dataset captures both long-term increases in XCO₂ and short-term enhancements associated with events such as wildfires, supporting carbon-emission monitoring and climate-change studies.