Biohub, Google DeepMind and Meta back $1.8 billion AI biology push

Big Tech has spent years teaching AI to write emails, generate images and summarize meetings. Now some of its biggest names want it to understand the human cell.

Google DeepMind, Meta Platforms and Isomorphic Labs will collectively put an additional $300 million into the Chan Zuckerberg Biohub’s Virtual Biology Initiative. The expansion, announced on October 7, 2026, brings the program’s total commitment to $1.8 billion.

Who is paying for what

The US Department of Energy is contributing more than $500 million over the next five years. That federal money targets advanced measurement, modeling and computational work.

The National Institutes of Health is taking a different role. Rather than writing a new check, it will coordinate existing datasets that have already received over $500 million in federal funding.

The Chan Zuckerberg Biohub launched the program in April 2026 with an initial $500 million commitment. Six months later, the price tag has more than tripled.

The total is being billed as the largest commitment for AI-ready biological data to date.

The data problem nobody could ignore

According to the research summary of the announcement, existing datasets contain only hundreds of millions of cellular data points. Effective AI models for biology are projected to need billions, if not trillions.

The initiative aims to close that gap by generating open datasets that detail how cells respond to various perturbations. A perturbation is any change you introduce to a cell, such as a drug, a genetic edit or an environmental stress, and then observe what it does in response.

The timeline is measured in years, not quarters. The first dataset is expected in about a year, and accurate virtual cell models are anticipated within a five-year timeframe.

Open science, with a commercial asterisk

The program follows an “open science” approach. Data produced by the initiative is meant to become publicly available.

There is a catch for the corporate partners, though. Commercial backers may receive embargo periods before the data goes public. The source material does not specify their length.

Why drug developers are paying attention

The stated goal is to speed up drug development dramatically. The initiative aims to cut timelines from decades to approximately five years.

What this means for the players involved

Google DeepMind, Meta and Isomorphic Labs are collectively contributing $300 million. With the Department of Energy committing more than $500 million and the NIH coordinating prior federally funded work, this is not a purely private venture.

For investors in healthtech and biopharma, the key milestones are clear. The first is whether the initial dataset actually lands in about a year as planned. The second is whether those datasets reach the scale researchers say AI needs, moving from hundreds of millions of data points toward billions. The third is whether virtual cell models become accurate enough to change real drug development decisions within the five-year window.