Artificial intelligence-driven design of inorganic materials for biomedical applications

Inorganic non-metallic materials have garnered significant attention in biomedicine due to their structural tunability and multifunctional properties. The emergence of next-generation artificial intelligence (AI) technologies is transforming the conventional trial-and-error paradigm of materials research, supporting a more data-driven and predictive approach. This review highlights the frontier applications and key breakthroughs of AI in the design and development of inorganic non-metallic biomaterials. We first introduce the fundamental paradigm and workflow of AI4S, which integrates data acquisition, model construction, and material optimization. Based on this framework, we summarize recent advances in two major directions: forward prediction and inverse design. Forward prediction focuses on critical performance indicators such as drug release profiles, biological interactions, material stability, toxicity and biosafety, and biocatalytic activity. Simultaneously, AI-driven inverse design is accelerating the development of targeted materials, including drug delivery carriers, biomaterials for inflammatory disease treatment, antitumor, and tissue engineering. Finally, we discuss key breakthroughs in AI-assisted inorganic biomaterial design, emphasizing the emerging role of generative models in enabling inverse design. Future perspectives on integrating domain knowledge, high-throughput experiments, and interpretable models are also outlined, which may provide guidance for the intelligent development of next-generation biomaterials.