Google DeepMind and the Wellcome Sanger Institute launched an AI genomics consortium aimed at building large genomic datasets for training predictive biology models. Google.org and Google DeepMind committed $5 million per year and plan to collaborate for five years. The consortium’s dataset strategy focuses on materials designed specifically for machine learning model training, with the goal of accelerating discovery in life sciences. Partners also intend to recruit additional collaborators and extend prior joint programs like AI genomics fellowships. The initiative signals how major AI players are shifting from model development to dataset infrastructure—an increasingly decisive bottleneck for advancing biology-focused predictive algorithms.