Federal Agencies Promote Federated Learning For Pediatric Cancer

In a guest article, Michael Savas and Richard Ricciardi respond to the Trump administration's recent executive order 'Unlocking Cures for Pediatric Cancer with Artificial Intelligence,' urging federal agencies to adopt federated learning to train AI without centralizing patient records. They note early implementations at Johns Hopkins and Mayo Clinic, and warn of privacy, interoperability, and resource barriers. The authors call for an HHS-anchored public-private partnership to scale federated networks and improve pediatric cancer diagnosis and treatment.