Benign-by-design chemistry: Reinventing ligand-based drug design at the edge of AI

Ligand-based drug design (LBDD) has long driven therapeutic innovation; however, its traditional potency-centered paradigm often oversimplifies biological complexity. Advances in artificial intelligence (AI) now enable multiobjective strategies that integrate polypharmacology, safety and environmental sustainability. Despite this progress, a unified framework that systematically incorporates these dimensions within AI-augmented LBDD is lacking. Here, we propose that embedding benign-by-design principles into multiobjective optimization can enable the proactive mitigation of toxicity, off-target effects and ecological impact from the early design and discovery stages. This shift redefines LBDD as a complexity-aware and ethically grounded discipline capable of delivering safer and more sustainable therapeutics. Accordingly, the primary objective of this review is to outline a unifying framework for AI-enabled LBDD that moves beyond potency-centered optimization by integrating efficacy, safety, sustainability and societal impact.