Insilico Medicine announced on July 7, 2026 that it had dosed the first patients in a Phase III clinical trial of rentosertib, an oral small-molecule drug the company says is the first AI-discovered and AI-designed therapy to reach late-stage testing for idiopathic pulmonary fibrosis, or IPF, a progressive and ultimately fatal scarring disease of the lungs. The trial will enroll 320 patients across 47 sites in China, with participants taking a once-daily dose for 52 weeks while researchers track changes in lung function.
A Disease With Few Good Options
IPF causes irreversible scarring of lung tissue, gradually robbing patients of the ability to breathe. Only two drugs, pirfenidone and nintedanib, have been approved for it over the past decade, and both primarily slow decline rather than reverse it, often at the cost of significant gastrointestinal and liver side effects that lead many patients to stop treatment. Rentosertib targets TNIK, an enzyme implicated in the fibrotic signaling pathway, in what Insilico describes as a first-in-class mechanism for the disease.
From Algorithm to Molecule in Record Time
Rentosertib’s origin story is the point Insilico most wants regulators and investors to notice: both the biological target and the drug candidate itself were identified using the company’s Pharma.AI platform, a suite of generative-chemistry and target-discovery models. Insilico has said the compound moved from target identification to a clinical-ready molecule in roughly 18 months, a fraction of the multi-year timeline drug discovery has traditionally required, though the company has not published a full peer-reviewed accounting of that timeline for outside verification.
What the Earlier Data Showed
The Phase III trial follows a Phase IIa study, called GENESIS-IPF, in which the 60-milligram once-daily dose of rentosertib was associated with a mean improvement in forced vital capacity — a standard measure of lung capacity — of 98.4 milliliters at 12 weeks, alongside a manageable safety and tolerability profile. Those numbers, while encouraging, come from a small early-phase study; the Phase III trial is designed specifically to determine whether that signal holds up in a much larger, more diverse patient population over a full year rather than three months.
Serious Institutional Backing
The trial is being led by Professor Zuojun Xu of Peking Union Medical College Hospital, part of the Chinese Academy of Medical Sciences, with academician Nanshan Zhong and Professor Chang Chen serving as co-principal investigators — a roster that gives the program significant credibility within Chinese pulmonology, a field Zhong in particular helped bring to global attention during prior respiratory disease outbreaks.
Believers and Skeptics
Supporters of AI-driven drug discovery point to rentosertib as proof the field is finally producing molecules that survive clinical scrutiny rather than just generating candidates faster on paper; industry trackers have counted well over 100 AI-enabled therapeutic assets now in human trials across dozens of companies. Skeptics counter that speed of discovery says nothing about whether a drug actually works, and that of those AI-originated assets in trials, only a small fraction have even completed Phase II, let alone Phase III — the stage where most experimental drugs, AI-derived or not, ultimately fail. For rentosertib, the real test is not how it was invented but whether a 320-patient, yearlong study shows it meaningfully slows a disease that has defeated most therapies aimed at it.
What to Watch
Results from the 52-week trial are not expected until well into 2027, meaning rentosertib’s fate as a marketed drug remains years away even in the best case. But its progression to Phase III is being closely watched across the pharmaceutical industry as a bellwether: if an AI-designed molecule can clear the highest bar in drug development, it would give considerable momentum to a wave of similarly AI-originated compounds from other companies now sitting in earlier stages of testing, waiting to see whether the technology’s promise translates into medicines patients can actually use.
Photo: qamarabbas20 / PIXABAY via Pixabay
