A global coalition committed $1.8 billion to build open data infrastructure for training AI models of biology. The group includes Biohub, the U.S. Department of Energy, NIH, Google DeepMind, Isomorphic Labs, Meta, and other partners. The funding is intended to generate open, AI-ready biological data aimed at predictive modeling of cells and disease. The stated goal is to reduce friction for model developers by standardizing access to datasets that can be directly used for training and evaluation. In practice, this could accelerate foundation-style approaches in biology by improving dataset availability, interoperability, and benchmarking—areas that have constrained model performance and reproducibility.