All news
    AI and the Jewish world
    3 min read

    Israeli AI model aims to flag Alzheimer's risk years early

    Alzai Health, an Israeli company, reports that an AI model built on routine medical records was validated across 1.6 million patient records in Israel and the US. The results are the company's own.

    Israeli AI model aims to flag Alzheimer's risk years early

    An Israeli health-technology company says it has built an artificial intelligence model that can estimate a person's risk of developing Alzheimer's disease years before symptoms begin. Israel National News reported the announcement, and other outlets, including The J and Matzav, carried it on November 15 and 16, 2026. As with most announcements of this kind, the findings are the company's own, and none of the coverage we found points to an independent, peer-reviewed publication.

    What was reported

    • The company is Alzai Health. Coverage says its model grew out of research at Clalit Health Services' Beilinson Hospital, in collaboration with Ariel University.
    • The model works from routine data: blood tests, existing diagnoses and demographics. Reports stress that it does not need expensive brain scans or invasive procedures.
    • The company says the model has now been validated across more than 1.6 million patient records, drawn from both Israel and the United States, covering roughly 200 million individual medical data points. The sources we read do not say how many records came from each country.
    • In July 2026 the company announced a separate validation on US data alone, from more than 1.1 million patients, using a licensed dataset. Medical Device Network reported that result as comparable to earlier validations on Israeli healthcare data. The 1.6 million figure is therefore a combined total, not a US-only one.
    • The developers say the tool is not designed to diagnose Alzheimer's or to replace a doctor's judgement. Patients it flags would still need a proper clinical assessment and, where appropriate, further testing.
    • The company describes the validation as an important step toward possible commercialisation in the US market.

    Why it matters

    Alzheimer's is a condition that touches many families, including many in our communities who care for an aging parent or grandparent. Earlier identification of risk, if it holds up, could in principle help doctors and families plan sooner and arrange screening and support. That is the promise being described. It is a promise, not a result: a risk estimate is not a diagnosis, and a model that performs well on records can still behave differently in a particular clinic or population.

    The story also shows where Israel's AI sector is putting effort. Health data held by large health funds, combined with university research, is one of the country's practical strengths. Announcements like this one are worth reading closely: look at who ran the validation, whether the results are published for others to check, and what the tool is said to do and not do.

    What it means for families, schools and frum professionals

    • Families: a screening tool of this kind, if it is ever offered, would be something to discuss with your own physician. Nothing here is advice to test, or not to test, anyone.
    • Privacy: models built on medical records raise questions about consent and data handling. When a company says it used patient records, it is fair to ask how they were protected and whether they were de-identified. The coverage we read does not go into those details.
    • Health and chessed professionals: for those who work with elderly patients and their caregivers, this is an early development to watch, not a tool to rely on yet.
    • Mechanchim: it is a concrete example for older students of AI doing something beneficial, alongside the habit of asking how a claim was checked.

    Medical and halachic questions both belong with qualified people. A doctor decides what a test means for a patient, and a rav decides questions of Torah. An AI model does neither.

    What it changes for someone using Navōn

    Nothing directly. Alzai's model is a separate product from Navōn, and Navōn is not a medical tool. Navōn does not diagnose or predict illness, and we are not suggesting it can. The lesson that applies is the general one: judge any AI claim by who checked it and what it says it cannot do.

    See how Navōn approaches kosher AI for families and schools.

    Common questions

    What is Alzai Health's Alzheimer's AI model?

    An Israeli company's model that estimates Alzheimer's risk from routine records such as blood tests, existing diagnoses and demographics. It grew out of research at Clalit's Beilinson Hospital and Ariel University.

    Can AI predict Alzheimer's years before symptoms?

    One Israeli company says its model can flag risk years early. The results reported so far are the company's own, and we found no independent peer-reviewed publication yet.

    Does the Alzheimer's AI model diagnose the disease?

    No. The developers say it is not designed to diagnose or replace clinical judgement. Flagged patients would still need medical assessment.

    How many patients was the Alzai model tested on?

    The company reports more than 1.6 million patient records from Israel and the US combined. A July announcement covered more than 1.1 million US patients alone.

    Sources

    This story was researched and written by Navōn.

    Inside Navōn, News writes an article on any topic from sources you choose, checked against your family's or school's standards.

    More stories

    Published by Navōn. How these guides are written and checked.

    Read next