> ## Content Index > Fetch the complete content index at: https://nexi.fund/llms.txt > Use this file to discover other available public pages before exploring further. # Six Aging Clocks Agreed on Insilico's AI Drug. The Hard Endpoint Is Still Missing. - URL: https://nexi.fund/insilico-ai-longevity-2026/ - Published: 2026-09-27T18:00:18.000Z - Updated: 2026-09-27T18:00:18.000Z - Description: Insilico Medicine moved an AI-designed drug from target to preclinical candidate in 18 months. In September it reported biological-age reduction across six independent proteomic clocks, then opened a longevity toolkit and a vaccine program. What is missing is the endpoint investors actually price. - Author: Nexi.fund Labs - Tags: Biotech & Health, #mode-5, #hook-character, #track-E Alex Zhavoronkov has spent years arguing that aging is a measurable process, which would make it a druggable one. In September 2026 his company put that claim in front of a journal, and in front of the market. Insilico Medicine runs generative AI across the whole drug pipeline: target discovery, molecule design, trial design. It listed on the Hong Kong exchange in December 2025, raising about $291 million, and trades as 3696.HK. Its lead asset, rentosertib, is an AI-discovered and AI-designed candidate for idiopathic pulmonary fibrosis (IPF), a scarring lung disease with few good options. ``` CHRONICLE: Insilico Medicine's longevity program ───────────────────────────────────────────────────────────── 2014 ─────────── June 2025 ─────────── Sep 7 2026 🏭 🧪 ◉ NOW Founded; Rentosertib Six proteomic clocks AI drug Phase IIa met report biological-age discovery safety endpoint reduction ``` Company disclosures, 2014–2026 ``` ───────────────────────────────────────────────────────────── Sep 15 2026 ───── Sep 17 2026 ───── WHAT COMES NEXT 🧬 🔬 🔥 Longevity Open toolkit Clinical outcomes, Vaccines (Cell cover) not clock readings ``` Company disclosures, September 2026 ## The six clocks that all moved On September 7, Nature Biotechnology published an analysis built on 12-week longitudinal Olink proteomic data from 42 IPF patients. Six aging clocks ran on the same samples. They were developed independently by groups at Harvard, Oxford, Peking University and Insilico itself, and include ProtAge, OrganAge and PAC. Every clock pointed the same direction. Predicted biological age fell in treated patients. 6 aging clocks agreed #### Independent proteomic clocks Six models from four institutions moved the same way on the same blood samples. *Nature Biotechnology, 2026* Forced vital capacity (FVC), the standard lung-function measure, reversed dose-dependently against placebo. The program moved from target identification to preclinical candidate in roughly 18 months, quick for a field where discovery usually runs in years. 42 IPF patients analysed #### Proteomic cohort, Phase IIa Twelve weeks of Olink proteomics per patient. Small, early, and the only human dataset of its kind so far. *Nature Biotechnology, 2026* Zhavoronkov, the founder and chief executive, presented the results at a Nature conference at Sorbonne University in Paris on September 8. | Endpoint | What it measures | Demonstrated | | ------------------------------ | ------------------------ | ------------------------- | | **Proteomic aging clocks** | Predicted biological age | ✔ Six clocks declined | | **FVC** | Lung function | ✔ Dose-dependent reversal | | **Mortality, hospitalisation** | Hard clinical outcome | ✗ Not yet shown | Nature Biotechnology, September 2026 🎯 Six independent clocks agreeing is rare in a field that argues about its own instruments. FVC moved with them, so the signal is not purely computational. None of it is a clinical outcome. A clock measures a prediction, not a life. ## Why six clocks beat one A single aging clock is a model with a training set and a bias. Different labs calibrate against different cohorts, weight different proteins, and disagree about what biological age means in the first place. When six of them, built by rival teams, move together on one drug, the shared signal is harder to dismiss as a quirk of the instrument. That is the strongest reading of the September paper. The weaker reading is structural. All six clocks ran on the same 42 samples, from the same trial, in a study whose corresponding author runs the company that designed the drug. Independent replication on a separate cohort has not happened. The paper reads its numbers from an Olink proteomic panel taken in a single blood draw. That is a broad panel attached to a small trial. The economics compound the ambiguity. A validated aging endpoint would turn a decade of drug development into something measurable in months. A regulator willing to qualify that endpoint would change how every longevity company raises money. Neither exists. What exists is a listed company, a plausible mechanism, and a result its own founder frames as the start of a longer argument. The market, for now, is pricing the argument and not the endpoint. ## The bet beyond one molecule Two days before the toolkit paper, Insilico opened a second front. On September 15 it launched Longevity Vaccines, a research program that pairs generative AI with programmable RNA medicine. 🧬 **Longevity Vaccines (September 15, 2026)** Single-dose, self-limiting in vivo cell therapies designed to clear senescent, fibrotic and autoreactive cells. The first target is immune rejuvenation, built on Pharma.AI and aging foundation models. The scale of the ambition is easy to state and hard to price. The drugs are meant to be preventive, programmable and given once. ## The clock problem Proteomic aging clocks are trained on correlation, not causation. A drug can move a clock without moving the disease. In June 2025 the rentosertib Phase IIa trial (NCT05938920) was published in Nature Medicine and met its primary safety endpoint. Safety, not efficacy. As we wrote in September, pelacarsen cut lipoprotein(a) sharply and cardiovascular events stayed flat. The cleanest recent reminder that a surrogate is not a result. > We have developed a conceptually new type of personal longevity AI that will not only let you live longer, but also help you and others discover new longevity drugs that work for everyone.— Alex Zhavoronkov, founder and CEO, Insilico Medicine #### Does a falling aging clock mean a longer life? Not on its own. Clocks are surrogate biomarkers: they estimate biological state from protein patterns. Turning a lower reading into fewer deaths or hospitalisations needs a larger trial with hard endpoints. ## What investors are pricing Insilico is not a private bet discovered late. It raised about $291 million in its December 2025 listing, and in May 2026 it signed a multi-million-dollar AI co-development deal with Human Longevity to build longevity foundation models. $291M HK IPO, Dec 2025 #### Raised at listing, 3696.HK Roughly HK$2.28 billion. The company now reports scientific milestones to public markets, not just to journals. *SCMP, 2026* The demand side is real. UBS sizes the global longevity market at about $5.3 trillion, heading toward $8 trillion by 2030\. The market did not celebrate the foundation-model deal, though: Insilico shares slipped 4.21 percent to HK$48.28 around that period. 📊 **Signals to watch** A Phase IIb or III rentosertib trial powered for FVC or survival Whether other labs reproduce the six-clock result on independent cohorts Any regulator moving proteomic clocks toward a qualified endpoint Whether Longevity Vaccines reaches an investigational new drug filing [ Six aging clocks agree in Insilico's Phase IIa The primary study behind the September milestone: proteomic aging clocks integrated into a Phase IIa trial of rentosertib. Nature Biotechnology ](https://www.nature.com/articles/s41587-026-03286-y?ref=nexi.fund) The peer-reviewed anchor for every number in this piece. [ Insilico opens an AI longevity toolkit The Cell-cover release of LongevityBench, Longevity-LLMs and LongevityClaw, built with Liquid AI, the Buck Institute and Harvard. Insilico Medicine ](https://insilico.com/news/cell170926en-ai-longevity-discovery-toolkit?ref=nexi.fund) Company primary source for the open-toolkit and vaccine programs. [ Insilico and Human Longevity build predictive AI models The May 2026 collaboration, the market size numbers, and the share-price reaction. South China Morning Post ](https://www.scmp.com/business/china-business/article/3354923/insilico-teams-us-firm-build-ai-models-predicting-disease-decades-ahead?ref=nexi.fund) Independent business coverage; source of the investment framing.