AI-ready biological data: $1.8B global commitment

The total combines cash with in-kind contributions. The Department of Energy is committing more than $500 million over five years to lab measurement, modeling and computation. NIH is contributing datasets and repositories built with more than $500 million in earlier federal funding, and Biohub will standardize them for AI training. Google DeepMind, Isomorphic Labs and Meta are adding $300 million between them. The Virtual Biology Initiative launched on April 29, 2026 as a five-year program to produce open datasets for predictive models of life. Biohub put in $500 million at the time: $400 million for new cell-measurement tools and $100 million for outside research. NIH takes part through its Bio Genesis Mission, which is its piece of the federal Genesis Mission under Executive Order 14363. For founders, the takeaway is that the main constraint in bio AI is data, not model architecture. A large open commons makes it cheaper to train models of how cells respond to interventions. It also weakens startups whose main advantage is proprietary perturbation data. As a result, competitive advantage is likely to shift toward in-house wet-lab feedback loops, disease-specific data and better use of the shared resource.