Boston Children’s Hospital has used a secure internal ChatGPT environment to diagnose more than 40 rare diseases that had previously gone unresolved. By embedding AI across clinical and operational workflows, the institution has captured 60,000 hours of time savings, translating into over $7 million in redeployed labor.
The initiative began with isolated tools—documentation assistants and translation bots—before the leadership pivoted to an enterprise AI layer. This shared foundation, governed by safety and monitoring protocols, allows teams to prototype and deploy new capabilities in days instead of months. More than a third of employees now use AI daily, spanning research, billing, and patient care.
Operational gains focus on supply‑chain automation, where AI routes invoices and responds to queries, and on surgical scheduling, where it estimates patient acuity from clinical notes. In both cases, AI improves utilization, enabling earlier and more efficient operating‑room allocation. Clinical teams leverage the same technology for decision support, drafting documents, and synthesizing complex literature.
A standout application is the co‑pilot geneticist, which integrates genomic data, phenotypic profiles, and global literature to pinpoint causative mutations. Since its deployment, the system has delivered diagnoses for 40+ previously unsolved cases and identified new gene targets that may open therapeutic avenues. Families now receive answers that were once impossible, and researchers gain data that can inform future treatments.
Boston Children’s plans to deepen integration, expanding AI usage across more departments and refining the genetic co‑pilot with additional data sources. The hospital’s experience demonstrates that treating AI as infrastructure, rather than a set of add‑ons, can accelerate both operational efficiency and clinical discovery.
Source: Boston Children's uses AI to unlock new diagnoses
Domain: openai.com
Comments load interactively on the live page.