Mark Zuckerberg is betting big on artificial intelligence again—but this time, it’s not about social media or the metaverse. Along with his wife Priscilla Chan, he is backing the Chan Zuckerberg Biohub’s new Virtual Biology Initiative, a project that aims to use AI to better understand how human cells work.
The organisation has announced a $500 million effort to build AI systems capable of simulating how human cells behave. In simple terms, the idea is to create digital models of cells using vast amounts of biological data, allowing scientists to run experiments virtually instead of relying only on lab work. The long-term ambition through this project is essentially to use AI to help cure or prevent diseases.
The programme, called the Virtual Biology Initiative, will run over five years and focuses on developing what scientists describe as “predictive models” of human cells. These are essentially AI-powered versions of cells that can mimic how they function inside the human body, giving researchers a new way to study health and disease.
The broader idea behind this initiative is to shift biology from a slow, experiment-heavy field to one where researchers can first run simulations, test ideas, and predict outcomes using AI before stepping into the lab. If successful, this approach could significantly speed up how scientific discoveries are made.
Biohub says it is committing major resources to the effort, including large-scale computing power, advanced imaging technologies, and new tools to study biology at a much deeper level. Importantly, the data generated through the project will be made open and freely available to scientists worldwide.
Biohub plans to invest around $400 million into building advanced tools and infrastructure, ranging from high-resolution imaging to techniques that can observe millions, or even billions, of cells within living tissues. The remaining $100 million will support external researchers and institutions, helping to build a broader global ecosystem for data generation and collaboration.
Understanding and mapping human cells
So what exactly is the company trying to achieve? The goal is to build AI models that can understand how cells behave in both healthy and diseased states. If successful, these models could help scientists identify how diseases begin, predict how they develop, and even suggest ways to reverse them.
In practical terms, researchers could run thousands of “what if” experiments digitally, something that is currently slow, expensive, and often limited in real-world lab settings.
However, the initiative has its own set of challenges. You see, human biology is extremely complex, and current AI systems simply do not have enough high-quality data to model it accurately. Biohub itself acknowledges that achieving this vision will require far more data than exists today, along with new technologies to observe biology across multiple levels, from molecules to entire tissues.
Regardless, Zuckerberg’s broader goal through the Biohub is to “cure and prevent all disease” using AI-driven biology. While that ambition remains a long way from reality, it offers a glimpse of how artificial intelligence could help us tackle some of the most complex problems in science and medicine.
