The First AI-Designed Drug Enters Phase III
Rentosertib is the first drug with an AI-discovered target and an AI-generated molecule to reach the final stage of human testing. Here is why that matters, and why it is not yet proof.
In October 2026, Insilico Medicine dosed the first patient in GENESIS-IPF-3, a Phase III trial of rentosertib for idiopathic pulmonary fibrosis, a progressive and often fatal lung disease. Insilico describes it as the world's first Phase III trial of a generative AI-driven novel drug.
How it got here
Insilico's platform nominated TNIK as a fibrosis target, then generated the molecule to hit it. The company took the program from concept to a Phase I trial in under 30 months, a fraction of the typical timeline. Its Phase IIa results, published in Nature Medicine in June 2025, showed encouraging signals in lung function.
Why it is not proof yet
The Phase IIa study enrolled 71 patients for 12 weeks. That is a signal, not a verdict. Phase III is where most drugs fail, and AI has not changed that so far: AI-discovered drugs have failed Phase II at rates similar to traditionally discovered drugs, and none has completed Phase III.
The broader field is mixed. Generate:Biomedicines is advancing an AI-designed antibody toward Phase III in severe asthma, while some early AI pipelines have stumbled in Phase II readouts.
What leaders should take from it
AI is clearly compressing the front end of drug discovery: target selection and molecule design. Whether it improves the odds of success in the clinic is the question the next few years will answer. For health systems, investors, and regional life sciences hubs, the practical move is to build the data infrastructure and AI fluency that let you participate in either outcome.
Get the next one in your inbox.
Events, program news and the monthly roundups from ETA.
