Boston Children’s uses AI to unlock new diagnoses

Boston Children's Hospital has integrated OpenAI technology across its organization as a core part of its clinical, research, and administrative infrastructure. Facing financial constraints, administrative burdens, and cognitive limits in diagnosing rare conditions, the hospital shifted from fragmented AI pilot projects to an enterprise AI layer. This secure internal ChatGPT environment supports more than one-third of employees in their daily work, spanning supply chain management, surgical scheduling, research cohort building, and clinical documentation. According to the hospital, this technological integration has captured approximately 60,000 hours in time savings across more than 50 automations, representing over $7 million in redeployed labor. In addition to operational improvements, the hospital developed an AI co-pilot geneticist designed to synthesize genetic data, phenotypic information, and global medical literature. John Brownstein, Chief Innovation Officer at Boston Children's, notes that this system addresses human cognitive limits in synthesizing complex medical data. As a result of this clinical discovery tool, the hospital has successfully diagnosed more than 40 rare conditions that previously went unresolved, while also identifying new gene targets and potential therapeutic pathways. Boston Children's plans to continue expanding AI usage, embedding the technology more deeply into clinical decision-making and refining models in collaboration with OpenAI.