Against the backdrop of comprehensively advancing the “AI+” initiative and accelerating the achievement of high-level scientific and technological self-reliance, AI for science is becoming a key force in leading the transformation of research paradigms, fostering new productive forces, and reshaping the global landscape of scientific and technological competition. In the second half of 2025, major economies such as the United States, the United Kingdom, the European Union, and Japan intensively released numerous policy documents on AI for science. Although each country chose different development paths due to differences in resources and institutions, they universally elevated AI for science to a national strategic project, exhibiting common trends such as the automation of research processes, the assetization of scientific data, and the construction of sovereign research infrastructure. The United States emphasizes consolidating its technological dominance through a national mobilization system; the United Kingdom focuses on agile breakthrough strategies in areas of strength; the European Union maintains technological sovereignty through pan-European integration; and Japan builds competitive advantages through theory-driven approaches and standards leadership. International experience offers important insights for promoting AI for Science in my country during the “15th Five-Year Plan” period. AI for Science should be approached as a systemic public project, integrating computing power, energy, and research network infrastructure. China's efforts should focus on developing domestically controlled automated experimental platforms and interface standards, improving the institutional framework for the assetization of scientific data, and providing sustained momentum for the development of new productive forces through organized research and the cultivation of interdisciplinary talent.