The core objective of condensed matter structural chemistry lies in unraveling the intrinsic relationship between the microscopic structure and macroscopic properties of materials. With the rapid development of artificial intelligence (AI) technology, condensed matter chemistry is undergoing a profound transformation in its research paradigm. The traditional experience- and trial-and-error-based material design mode is gradually being replaced by an AI-driven closed-loop system of “data-algorithm-experiment.” This paper systematically reviews the application progress of AI in condensed matter chemistry, focusing on its breakthroughs in potential energy surface construction, electronic structure calculation, phase structure prediction, and material inverse design. Through the introduction of deep neural network potential functions, quantum machine learning, and generative models, AI has significantly improved the simulation accuracy and efficiency of complex systems. Meanwhile, strategies based on multi-scale modeling and active learning further realize cross-scale correlations from the atomic scale to macroscopic performance. Finally, this paper envisions the trend of autonomous material discovery and theoretical-computational integration empowered by AI, and points out that future research should focus on the construction of high-quality datasets, the development of interpretable models, and multi-modal collaborative optimization to promote the intelligent and sustainable development of condensed matter chemistry.