OpenAI Foundation Bets 25 Billion on Data to Fight Disease
Two Paths to the Same Wager
The OpenAI Foundation approaches disease from two directions at once. One path is direct: pour money into specific illnesses, such as Alzheimer’s disease, and hope that concentrated funding accelerates answers. The other path is indirect: build the raw material — scientific data — that every researcher can draw on, whether they work on cancer, tuberculosis or something else entirely. These are not competing strategies. They are two ways of describing the same bet, which is that artificial intelligence becomes useful against disease only when there is enough observation of the world to learn from.
That bet now has a balance sheet behind it. The foundation, created in October 2025 after a turbulent restructuring of OpenAI’s governance, holds a 26 percent stake in the company. [1] OpenAI is seeking a valuation of around $1.4 trillion. On that basis, the charity is on track to become one of the wealthiest non-profit organizations on the planet — and could end up the richest. It has committed to giving away at least $25 billion.
Most research funders operate like gardeners tending a few beds. This one is trying to change the soil. Jacob Trefethen, who leads life sciences at the foundation in San Francisco, California, frames the ambition plainly: the mission is to ensure AI benefits all of humanity, and the life sciences are where that benefit may be largest. [1] He came to the role from Coefficient Giving, a billionaire-funded foundation — a detail that matters, because it suggests the foundation hired for grant-making judgment, not just for technical enthusiasm.
The first cheques are already moving. The foundation has committed $125 million to fund freely available data for health and life-sciences research. [1] Within that round, $40 million goes to the University of North Carolina at Chapel Hill for cancer vaccine development. Half a million dollars supports a pilot to repurpose data from bankrupt biotechnology companies — an unusual target, and a telling one. Another $15 million funds work to better predict how drugs cross the blood-brain barrier and other membranes. These awards follow an earlier tranche of $100 million for Alzheimer’s research.
The Foundation’s Three Focus Areas
Trefethen describes three focus areas in the life sciences. The first is AI for Alzheimer’s disease. The second is public data for health. The third is what the foundation calls high-burden diseases — defined as any disease causing more than half a million deaths worldwide each year, or comparable levels of ill health. It sets a threshold that any candidate disease must clear before money follows.
Tuberculosis is the test case named. It is the largest infectious killer, and the foundation’s stated question is what is newly possible there for people who do not currently have good-enough options. It does not promise a cure. It asks what has become possible that was not possible before — a framing that leaves room for failure and still justifies the spending.
The unifying philosophy is stated as a combination: intelligence plus more observations of the world equals breakthroughs against disease. Observations, in this formulation, mean scientific data. On Alzheimer’s specifically, Trefethen is candid about the state of knowledge: collectively, we do not know enough to deliver treatments that are excellent for patients. More data of various forms is probably required to get there. It admits that the causal chain from data to therapy is not guaranteed. What the foundation can do is support the underlying scientific infrastructure so that all researchers can build on it and all patients can benefit — a public-good argument, and also a hedge, since infrastructure pays off regardless of which specific hypothesis wins.
Rescuing that material is unglamorous work with no obvious champion. Funding it suggests the foundation is willing to buy information that no market currently prices.
