Early Data Indicates an A.I.-Generated Drug Could Slow Aging - The New York Times
Insilico Medicine used artificial intelligence to develop rentosertib, a drug for chronic lung disease, and early trial data suggests it may also reduce biological markers of aging. The findings come with major caveats: the sample size was small, the drug hasn't been tested in healthy people, and the "aging clocks" used to measure results are scientifically debated.
- Rentosertib reduced biological aging markers across six different AI-based "aging clocks" in a clinical trial, according to results published in Nature Biotechnology.
- The drug was originally developed for idiopathic pulmonary fibrosis (IPF) and has only been tested in sick patients, not healthy ones, so anti-aging claims remain speculative.
- The study has significant limitations: small sample size, unproven reliability of aging clock measurements, and the drug is still years away from regulatory approval even for its primary lung disease indication.
Insilico Medicine's AI-designed drug candidate, rentosertib, originally developed to treat a rare chronic lung disease, now shows signs of slowing biological aging markers. The company published results in Nature Biotechnology showing the drug reduced aging indicators across six different "aging clocks"—AI systems trained to measure biological age. This comes from the same clinical trial that initially tested rentosertib for lung disease, so the anti-aging findings emerged as an unexpected secondary observation rather than from a dedicated aging study.
The company used neural networks to build two separate AI systems for drug discovery. The first analyzed massive datasets of patient health records, blood tests measuring microscopic proteins, and academic literature to understand disease mechanisms. The second system studied protein structures and how molecules bind to targets, then generated entirely new molecular candidates that could hit specific drug targets. This approach accelerated what normally takes years of traditional chemistry and testing.
The results come with serious caveats. The clinical trial involved a small sample size, and the aging clocks themselves remain scientifically controversial—experts debate how reliably they actually measure aging. More critically, Insilico only tested the drug in patients with lung disease, not in healthy people. The drug is still years away from regulatory approval even for its original lung disease indication, let alone for anti-aging use in otherwise healthy patients. Alex Zhavoronkov, the company's founder, acknowledged these limitations while still presenting the findings as promising evidence that AI can accelerate drug discovery.