Boston Children's Used ChatGPT to Bring 40 Families a Diagnosis!

Hey there, it's Shiichan! Today I've got a story that warms my heart a little. I honestly didn't expect AI to be working this hard behind the scenes at a hospital.

OpenAI News openai.com

What was announced?

In OpenAI's News, there's a story about how Boston Children's Hospital in the US is using ChatGPT across the whole organization. This is a big hospital with nearly a million outpatient visits a year, and it's bringing AI into everything from paperwork to diagnosing rare diseases.

The headline is that the hospital was able to bring new diagnoses to at least 40 families whose conditions had gone unexplained for a long time. That's a huge deal.

Why it matters

Diagnosing rare diseases, especially rare genetic ones, is incredibly hard. You have to read medical literature from all over the world and line it up against a patient's genetic data and symptoms, and there's more information than any expert can get through, no matter how much time they put in.

The team at Boston Children's frames it like this: the problem isn't effort, it's the limits of human cognition. There are many families who have been waiting years for an answer, and that's exactly where AI helps.

What changes

The people who gain the most are the families who were waiting for answers. Cases that were stuck at "cause unknown" finally get a name, and that alone opens up a path for treatment and care.

It's a big shift for doctors and geneticists too. One person could never carry thousands of cases, but when AI helps with the background research and hypothesis generation, the experts can focus on the judgment calls that really need them.

And it shows up on the operational side. Handing routine, back-office work to AI lightens the load on staff, so that time can go back into actual care.

Dive Deep

The star of the project is what the hospital calls a "co-pilot geneticist." It combines four things — genetic information, a patient's symptoms (phenotype), searches across the world's medical literature, and AI reasoning — to surface candidate explanations for diseases that couldn't be solved before. Think of it as a brilliant assistant doing all the legwork, with the final call left to a human expert.

The back-office side is impressive too: the hospital has more than 50 operational automations running. For example —

• Surgical scheduling and operating-room optimization
• Billing and invoice processing
• Supply chain (inventory and logistics) management
• Document drafting and administrative coding
• Research data analysis and building patient cohorts

Adding all of this up, the hospital estimates it freed up around 60,000 hours of work and saved more than $7 million in labor costs. The foundation is ChatGPT Enterprise (used for healthcare), running as an AI layer across the whole hospital.

(Since this is an OpenAI customer story, it's not something we can go try at our desks right away — but it's a real, concrete example of how AI helps out in the field.)

Wrap-up

• Boston Children's Hospital uses ChatGPT organization-wide and brought new rare-disease diagnoses to at least 40 families
• The key is a "co-pilot geneticist" — genetics, symptoms, literature search, and AI reasoning combined
• More than 50 automations run in the back office, saving an estimated 60,000 hours and over $7 million
• Source: OpenAI's News

If you want to see how AI actually helps in a medical setting — or if you have been wondering "what is AI even good for?" — this one lands gently but firmly!