The US government, Google, and Meta Platforms are partnering with nonprofit Biohub to advance biological research through AI by investing a combined $1.8 billion in creating open-access datasets to train predictive AI models.
- Total funding of $1.8 billion for AI-driven biology datasets
- Public-private cooperation with phased data release
- Goal to build predictive models to speed up drug discovery
What happened
On October 7, 2026, Biohub announced a new $1.8 billion funding effort to develop open datasets for AI-driven biological research. This initiative unites the US Department of Energy, National Institutes of Health, tech giants Meta and Google, and startups like Isomorphic Labs to standardize and expand data collection. The funding includes over $500 million from DOE over five years, more than $500 million from NIH's earlier allocations, and $300 million from Meta, Google DeepMind, and Isomorphic Labs combined.
This collaborative approach supports Biohub’s Virtual Biology Initiative, which intends to generate comprehensive datasets mapping cellular responses across numerous conditions. The initiative seeks to compress typical decades-long biological research into a five-year timeline and produce predictive AI models capable of transforming drug discovery and biomedical science.
Why it matters
Currently, biological data sets are relatively limited, creating a bottleneck for AI model training and predictive accuracy. The Virtual Biology Initiative aims to dramatically expand this data scale from hundreds of millions to billions of cells, enabling AI models to learn the 'language of biology' comprehensively. This knowledge base could revolutionize how new drugs are developed and research is conducted, reducing timeframes and costs substantially.
The collaborative model also represents a novel framework for open science, balancing commercial funders’ early data access with public availability after embargo periods. This sets a precedent for integrating private investment with government research funding to advance scientific innovation in a transparent, scalable manner.
What to watch next
The first large dataset from this initiative is expected within approximately one year, with more datasets and predictive models anticipated over the next five years. Observers should track progress against the ambitious timeline and monitor subsequent partnerships, especially with pharmaceutical companies and philanthropies that Biohub aims to engage as the initiative scales.
Additionally, similar AI biology projects by other organizations like Anthropic and the OpenAI Foundation highlight increasing competition and collaboration trends in this space. The emergence of powerful predictive models based on these datasets will be a critical marker of success and could reshape drug discovery, biomedical research, and AI application in life sciences.