Insilico Medicine and Lilly published a framework envisioning a fully autonomous “prompt-to-drug” pipeline integrating generative AI across design, synthesis, and testing. Parallel advances are cropping up in lab‑in‑the‑loop platforms: researchers demonstrated MULTI‑evolve and similar workflows that compress directed evolution cycles using machine learning and compact experimental measurement sets. Together these reports outline converging approaches where AI proposes designs and closed‑loop experimental platforms validate multi‑mutant protein variants rapidly. Sponsors and CROs will need to define data‑quality standards, IP pathways, and validation strategies as these methods move from proof‑of‑concept to regulated discovery workflows.