Intelligent Molecules: A Transformative Roadmap of AI in Drug Discovery

ABSTRACT Pharmaceutical research is undergoing rapid transformation through the integration of artificial intelligence (AI), enabling data‐driven strategies for molecular design and therapeutic development. Central to this progression is the emergence of intelligent molecules–drug candidates generated and optimized through closed‐loop AI systems that simultaneously balance efficacy, safety, and pharmacokinetic properties within defined chemical, biological, and clinical constraints. These approaches address persistent limitations of conventional drug discovery, including prolonged development timelines, escalating costs, and high clinical attrition rates. Recent advances in machine learning, particularly deep learning, have facilitated the integration of multi‐omics datasets to improve target identification, explore previously intractable biological systems, enable de novo molecular design, and support systematic drug repurposing. Large‐scale biological foundation models such as AlphaFold3 and AlphaGenome further accelerate structure‐guided drug discovery by integrating structural prediction with functional genomic interpretation. Beyond molecular design, AI contributes significantly to precision medicine by integrating genomic profiles with real‐world clinical data to improve patient stratification and adaptive dosing. Emerging explainable AI (XAI) frameworks strengthen model transparency and regulatory confidence by improving interpretability in molecular prediction and therapeutic decision‐making. Despite substantial progress, challenges related to data quality, reproducibility, regulatory validation, and ethical governance remain. Continued integration of AI with experimental validation and domain expertise will be essential to ensure reliable clinical translation and sustainable innovation in pharmaceutical development.