ARPA-H puts $98.5M behind rare-disease AI

ARPA-H is putting up to $98.5 million over 4.5 years behind AI for rare-disease diagnosis and drug development through its RAPID program. The agency’s award announcement names teams at UNC, Sage Bionetworks, FDNA, and Probably Genetic to build privacy-preserving resources spanning clinical, genomic, patient-reported, photo, video, voice, wearable, and synthetic data; it says more than 350 million people live with one of 10,000 rare diseases and only about 5% have an approved therapy. AI-health builders should watch this as data-and-validation infrastructure, not a model launch, alongside the archive’s evidence on ambient clinical AI.