Insilico Medicine, a company that uses artificial intelligence to find new drugs, has released free software meant to act as a personal longevity assistant. It was built in collaboration with David Sinclair, an aging researcher at Harvard Medical School1. The company announced the release on November 6, 2026, and launched a website for it called longevitygrok.si2.
The tool is called Longevity Grok, also described as a “Personal Longevity Superintelligence” (PLSI). It is designed to bring together lab results, data from wearable devices and clinical records, follow how a person’s health changes over time, and compare its own predictions with later test results1. No results showing how well it does any of this have been published.
What was released
The software was made for a platform called Grok Bot. Insilico describes it as “open source,” meaning its code is public, and “multi-agent,” meaning it runs many specialized AI helpers that split up the work1. According to the company, it coordinates up to 30 of these helpers1.
Insilico frames the problem it targets as scattered health data. Blood tests, continuous glucose monitors, wearable streams, epigenetic tests (lab tests that read chemical marks on DNA linked to aging) and medical records often sit in separate online portals. The company says this leaves patients unable to follow their own health over time and leaves doctors working from incomplete histories1. It also argues that AI chatbots now used in health care rarely beat traditional methods because they lack lasting, checkable memory and do not follow up on whether their advice helped1.
To combine records from different sources, the software converts them into standard medical coding systems, named in the release as LOINC, RxNorm and UCUM, which are shared vocabularies for lab tests, medicines and units of measurement1. It also runs routines in the background that watch for new longevity research and sort incoming data, so the user does not have to keep prompting it1.
How it is meant to work
Insilico describes three roles for the same system1.
A personal assistant
For an individual, the software builds one combined health timeline on the user’s own device. When it suggests a lifestyle, food or other longevity change, it attaches what the company calls a “prediction card.” The card lists which test result should change, a likely range, a time frame and a grade for how strong the supporting evidence is1. Each plan is first challenged by other AI helpers in what the company calls an adversarial review1.
When follow-up test results arrive, the software scores whether they landed inside the predicted range and moved in the expected direction. Over time it tracks how well its forecasts match reality1. The company describes this as turning each choice into an “N-of-1 experiment,” a test run on a single person1.
A helper for doctors
For physicians, Insilico says the tool supplies an organized history in place of piles of PDF files and portal downloads. It says prediction cards and evidence grades can support decisions and that scoring after follow-up allows discussion of how a patient responded and whether they stuck to a plan1. The company says the software has safeguards that keep formal diagnosis and prescriptions with licensed physicians1.
A source of research data
Users can choose to send anonymized health timelines to a shared research system called Longevity Mothership. The company says this goes through an anonymous relay with an “irrevocable kill switch” that deletes the records on demand1. Users can also give unused daily computing allowance to a background feature called Idle Research. Insilico compares it to volunteer computing projects such as Folding@home: AI helpers scan published studies, form hypotheses and design analyses using public databases1.
Who can use it, and what happens to the data
Insilico says the tool is available to all Grok Bot users, who can install it by pointing their bot to the project’s page on GitHub, a public site for sharing code1. Both the personal software and the shared research system are free and open source, so anyone can review or modify them1.
The company says privacy is the default. Raw medical files and identifiable records stay in the user’s own setup unless the user opts into shared research1. It says separate data-control helpers strip information as it leaves and enters the system, to prevent people from being re-identified1. These are the company’s descriptions of the design; the sources do not report any outside security review.
What the evidence shows, and what it does not
This is a software launch, not a study. The announcement describes what the tool is built to do. It does not report how accurate its predictions have been in real users, whether people who follow its suggestions see their test results improve, or whether it helps doctors make better decisions1.
The built-in scoring of predictions is the feature most likely to produce evidence later, since it records whether each forecast came true. But scores gathered inside the tool, from people who choose to use it, are not the same as a controlled trial comparing users with non-users. A change in a lab value after a suggested habit change can also have other causes.
Sinclair described the hopes behind the project in the release:
“Longevity Grok turns Grok Bot into an open platform where individuals, physicians, and scientists can use AI to better understand their health, predict the effects of interventions, and find ways to live longer, healthier lives.”
He added that “people can contribute to scientific discovery while improving their own health, helping us understand why we age and how we might slow or even reverse it”1. Those are aims; the release does not present data showing the tool improves health or slows aging.
Who is behind it
Insilico Medicine calls itself a clinical-stage generative AI drug discovery company1. It is listed on the Hong Kong stock exchange3. Sinclair is identified in the release as Professor of Genetics and Co-Director of the Paul F. Glenn Center for Biology of Aging Research at Harvard Medical School1. The sources do not say whether money changed hands between Insilico and Sinclair or what his formal role in the project is beyond “collaboration.”
The release fits a broader push by Insilico into AI for aging. In May 2026 it announced a multi-million-dollar deal with Human Life Foundation Models, a company set up by Human Longevity, Inc., to build large AI models for longevity science using Human Longevity’s anonymized health datasets3. Those models are expected to be sold commercially3. In October, Insilico highlighted ARDD 2026, an aging research conference held at Harvard in its 13th edition and its first in North America, which featured AI-driven drug discovery and FDA perspectives4.
What to watch next
- Whether Insilico or outside researchers publish data on how often the tool’s prediction cards come true.
- Whether independent experts review the open code, including its privacy and anonymization features.
- How many people opt in to share data with the Longevity Mothership research system, and whether any findings come out of it.
- Whether doctors adopt it, and whether any controlled study tests its effect on health.
For now, the tool is a freely available piece of software with a stated design. How well it works in practice remains unmeasured.
