
For a decade, "AI will design new medicines" has been a promise. This week it turned into a peer-reviewed result — with an unexpected side finding.
The drug is rentosertib, from Insilico Medicine. It treats idiopathic pulmonary fibrosis, a scarring lung disease that usually strikes around age 65 and has few good options. What makes it unusual is its origin: AI picked the biological target, generative AI drew the molecule, and the whole thing went from idea to drug candidate in about 18 months.
Then the researchers measured something else. The results, published in Nature Biotechnology this month, are what got attention.
Here's the breakdown:
Why it matters: this is the first time an AI-originated drug has been followed this publicly from target to Phase III, with the peer-reviewed paperwork to match. It's evidence that AI isn't only speeding up existing chemistry — it's pointing at biology humans skipped.
The catch: 42 patients. Four weeks. And the researchers say it plainly — the study cannot yet separate slower aging from a treated lung. Sick lungs getting better may be enough on its own to make the blood look younger. Aging clocks are research instruments, not diagnoses, and nobody should read this as a longevity pill.
What it is, is a real drug for a brutal disease, found by a machine, now being tested in 320 people. That is remarkable enough without the anti-aging headline.