😺 OpenAI found 18 rare diseases
The Neuron · Grant Harvey · 2026-06-19
A study published in NEJM AI used OpenAI's o3 Deep Research model to reanalyze 376 previously unsolved pediatric rare-disease cases at Boston Children's Hospital and Harvard, enabling clinicians to confirm 18 new diagnoses after expert review — while OpenAI separately expanded health capabilities in ChatGPT for its 230 million weekly health-question users.
Appears in
Extraction
Topics: ai-in-healthcarerare-disease-diagnosisclinical-ai-researchopenai-modelsconsumer-health-ai
Claims
- OpenAI's o3 Deep Research model helped researchers surface leads that led to 18 clinician-confirmed rare-disease diagnoses from 376 previously unsolved pediatric genetic-disease cases.
- More than 230 million people ask ChatGPT health and wellness questions every week, making it one of the largest consumer health information channels.
- GPT-5.5 Instant, available to free ChatGPT users, improved on health evaluations around urgent-care recognition, uncertainty acknowledgment, and context gathering.
- Roughly half of rare-disease patients remain undiagnosed after extensive specialist review and genomic sequencing, illustrating the scale of the unmet diagnostic need AI is targeting.
- The rare-disease study used AI to generate evidence-linked diagnostic leads for human clinician review rather than allowing autonomous diagnosis, preserving the human-in-the-loop model.
Key quotes
That is the useful near-term shape: AI as the tireless second reader, not the doctor of record.
The risk is that consumers will treat a polished answer like a final answer, especially when care is expensive or slow.
OpenAI is turning health into one of ChatGPT's biggest mainstream jobs. The test now is whether it can make people better prepared for care without convincing them they can skip care entirely, especially when the answer sounds calm, confident, and complete on a stressful night alone.