Abstract
Chitosan nanoparticles (CNPs) are promising anticancer drug-delivery carriers because of their biodegradability, mucoadhesive behavior, drug-loading capacity, and surface modifiability. However, previous studies have often examined green synthesis, CNP formulation, artificial intel-ligence (AI), and anticancer delivery as separate research domains, leaving limited synthesis of how AI-assisted optimization can improve environmentally sustainable CNP systems for cancer therapy. This systematic literature review evaluated studies published from 2016 to 2025 on green-synthesized CNPs optimized through AI, machine learning, or statistical optimization approaches for anticancer applications. A Scopus search using the Boolean string "chitosan AND anticancer AND optimization" identified 95 records. After eligibility screening, 42 records were excluded because they were older than 2016, non-article publications, non-English records, or out-side the oncology/green-CNP scope, leaving 53 studies for review. The evidence was synthesized into four domains: optimization and predictive modeling, green synthesis and material innova-tion, targeted and multifunctional nanocarriers, and preclinical efficacy and translational readiness. Across the included studies, optimized CNPs showed particle sizes ranging from approxi-mately 5 to 473 nm, encapsulation efficiency up to 98%, and zeta potentials from -31 to +98 mV. Reported therapeutic improvements included enhanced cytotoxicity, reduced IC50 values, sus-tained release up to 72 h, and inhibition rates up to 82% in selected cancer models. Nevertheless, cross-study comparison was limited by inconsistent model-validation metrics, incomplete toxicity reporting, limited in vivo validation, and insufficient scalability assessment. Integrating AI-guided optimization with green CNP synthesis can accelerate sustainable nanomedicine design, but future studies should prioritize standardized reporting, risk-of-bias control, life-cycle assess-ment, and regulatory translation.