Artificial Intelligence-Driven Design and Optimization of Advanced Drug Delivery Systems: From Formulation Prediction to Clinical Translation: A Comprehensive Review

Advanced drug delivery systems (DDS) — including polymeric and lipid nanoparticles, lipid nanoparticles (LNPs) for nucleic acid therapeutics, solid oral dosage forms, and three-dimensional (3D)-printed and stimuli-responsive platforms — require the simultaneous optimization of dozens of interacting formulation and process variables to achieve target size, drug loading, release kinetics, stability, and biological performance. Traditional design-of-experiments and trial-and-error approaches struggle to navigate this high-dimensional, nonlinear design space efficiently. Artificial intelligence (AI) and machine learning (ML) — spanning classical regression and ensemble methods, deep neural networks, graph neural networks, generative models, and Bayesian optimization — are increasingly used to predict critical quality attributes, accelerate excipient and lipid screening, optimize manufacturing processes, and guide the transition of novel delivery systems from bench to clinic. This review synthesizes recent literature on AI-driven formulation prediction across nanoparticulate, lipid-based, oral solid, and 3D-printed dosage forms; AI-guided process optimization and digital-twin-enabled manufacturing; and the evolving regulatory landscape for AI in pharmaceutical development, including the U.S. Food and Drug Administration's 2025 draft guidance on AI in regulatory decision-making. We also discuss persistent barriers to translation — data scarcity and quality, model interpretability, cross-laboratory generalizability, and validation standards — and outline emerging directions, including self-driving laboratories, federated learning, digital patient twins, and large language model-assisted formulation design, that may further shorten the path from AI-predicted formulation to clinically approved product.