Abstract: Artificial Intelligence for Science (AI4S) is reshaping global scientific research paradigms. Analyzing the United States’ policy framework and implementation pathways holds significant intelligence value for China, which is currently in a critical window of opportunity for AI4S. To deeply dissect the underlying mechanisms of AI4S, this paper selects relevant U.S. policy texts and employs the WSR methodology to construct a collaborative analysis framework. Conceptual Dimension: Utilizes concept maps to depict the overarching operational logic of U.S. AI4S policies. Physical Dimension: Focuses on three core AI technological elements—algorithms, computing power, and data—and analyzes U.S.AI4S technology governance pathways through thematic clustering. On the human dimension, content analysis extracts policy provisions into categories to distill regulatory measures for key actors. Findings reveal that U.S.AI4S policies coordinate technological breakthroughs with risk prevention through dynamic equilibrium mechanisms, forming a strategic framework centered on “innovation-first, security-backed” principles, layered technical governance strategies, and chain-based talent management pathways. Finally, drawing from U.S. practices and China’s existing policies, this paper proposes optimization recommendations for AI4S development across governance frameworks, open collaboration, and strategic deployment.

Key words: AI4S; Science and Technology Policy; Strategic Layout; WSR Methodology; US