ARPA-H awarded up to $98.5 million for AI-enabled rare disease diagnosis, discovery, and treatment efforts, funding teams building capabilities around artificial intelligence and large multimodal datasets. The awards include projects led by the University of North Carolina, Sage Bionetworks, FDNA, and Probably Genetic, with an emphasis on using AI systems and multimodal data to accelerate how rare diseases are identified and targeted. The program’s structure suggests a move toward data-intensive platforms that can generalize across rare disease cohorts, rather than siloed single-indication approaches. For biotech, ARPA-H’s investment is a signal that government-backed AI infrastructure remains a priority, especially when paired with diagnostic and translational aims.