Researchers from Insilico Medicine and Eli Lilly published a framework in ACS Central Science proposing a fully autonomous “prompt‑to‑drug” pipeline that links generative AI, automated synthesis, and iterative biological testing. The paper outlines integrated agentic systems that convert human prompts into candidate molecules and experimental plans, with feedback loops to refine designs. The authors present architecture, component examples, and potential workflows rather than clinical results. They note technical and validation gaps—data quality, assay automation, and safety oversight—that must be addressed before deployment. The publication signals escalating industry ambition to move discovery decisions from human-led cycles to machine‑led loops, with implications for R&D timelines, capital allocation, and regulatory expectations.