The First AI-Designed Drug Now Shows an Anti-Aging Signal

A drug that artificial intelligence both found and drew has left a mark researchers did not go looking for. In a paper published on 7 September in Nature Biotechnology, scientists at Insilico Medicine and a group of academic collaborators report that patients who took rentosertib, an experimental treatment for a scarring lung disease, came out of the trial looking biologically younger by the measure of six independent aging clocks.

Rentosertib, known in development as ISM001-055, is unusual in its origins. Insilico used its own machine-learning platform to pick out the drug's biological target, an enzyme called TNIK tied to both fibrosis and the wear of aging, and then used generative models to design the molecule that acts on it. It is one of the first medicines with both an AI-chosen target and an AI-generated structure to reach human trials, which is why every result it produces gets read closely.

What the clocks showed

The finding comes from the Phase IIa trial called GENESIS-IPF, a double-blind, placebo-controlled study of 71 patients with idiopathic pulmonary fibrosis across 22 sites in China. The trial's job was to test safety and look for early signs the drug worked on the lungs. The aging result is a separate analysis layered on top of it.

Researchers took blood drawn during that trial and ran it through six proteomic aging clocks, tools that estimate a person's biological age from the mix of proteins circulating in their blood. Each clock was built by a different team using different methods. All six pointed the same way: patients on rentosertib looked younger by these measures than those on placebo. The authors describe it as the first head-to-head comparison of several proteomic clocks inside a single drug trial.

Reasons to stay measured

A biological age reading is a biomarker, not a lifespan. A protein signature moving in a promising direction is not the same as a person living longer or staying healthier, and the trial was small and built to answer a different question. The result is a signal worth chasing, not a proof of anything about aging itself.

What makes it interesting is where it came from. AI drug discovery has spent years promising to compress the slow work of finding targets and designing molecules. Rentosertib is now in Phase III testing in China for its original lung indication, and this paper adds a second reason to watch it. The same computational approach that mapped the wiring of a fruit fly's brain is now being turned on the biology of aging, a field long crowded with bold claims and thin evidence.

Insilico's founder, Alex Zhavoronkov, has argued for years that aging and disease are worth treating as one problem. This is the first time a drug from that thesis has produced a clinical hint that the two might move together. The next trials will show whether the hint holds.