Artificial intelligence applied to formulation design, process optimization, and quality assessment of 3D-printed dosage forms and drug-delivery systems: a scoping review

Three-dimensional (3D) printing is a computer-aided, layer-by-layer manufacturing technology with strong potential for personalized medicines. Combined with artificial intelligence (AI), it enables predictive, data-driven development, improving formulation design, process robustness, and quality control. However, evidence remains fragmented across technologies, dosage forms, and computational approaches. This scoping review aimed to map and describe the application of AI to formulation design, manufacturing-process optimization, and quality assessment of pharmaceutical dosage forms and drug-delivery systems produced by 3D printing, including device-assisted delivery systems such as microneedles. The review followed Joanna Briggs Institute methodology and PRISMA-ScR guidelines. Searches were conducted in MEDLINE/PubMed, Scopus, and Web of Science until November 18, 2025. Original studies applying AI to the design, manufacturing, or quality assessment of 3D-printed drug products or drug–device combinations were included. Two reviewers independently performed study selection, data extraction, and analysis. Findings were synthesized through descriptive tables and narrative analysis. Twenty studies published between 2020 and 2025 met the inclusion criteria. Technologies included Fused Deposition Modeling, Selective Laser Sintering, Semi-Solid Extrusion, Vat Photopolymerization, and Inkjet Printing. Oral dosage forms and associated formulation or printing workflows predominated; two studies specifically investigated microneedle-based delivery systems. Reported approaches included random forests, support vector machines, artificial neural networks, deep learning, chemometric regression, generative models, Bayesian optimization, and evolutionary algorithms. Applications addressed formulation and printability prediction, processing-condition selection, drug-release modeling, and image-based quality assessment. Individual studies reported potential benefits related to process robustness, experimental effort, and development time; however, differences in study designs, datasets, and outcome definitions limit comparisons and generalization. The review maps applications of AI in formulation design, process optimization, and quality assessment in pharmaceutical 3D printing, without establishing their comparative effectiveness. Future research should prioritize methodological harmonization, standardized validation, and interdisciplinary collaboration to facilitate translation and regulatory acceptance.