Cleaner by Design? Can artificial intelligence invent less-polluting chemistry—and how could we prove it?

This student review explores whether artificial intelligence can help invent less-polluting chemistry by predicting molecular properties, generating candidates and optimizing chemical processes. It distinguishes computational predictions from experimental and environmental proof, while examining toxicity, persistence, waste, energy consumption and life-cycle impacts. The article proposes an original framework—the Green Chemistry Evidence Gate (GCEG)—to evaluate function, hazard, environmental fate, process efficiency, life-cycle burden and evidence quality before laboratory testing. It also presents a safe open-data research project suitable for a young scientist.