ChatGPT Health performance in a structured test of triage recommendations

Ashwin Ramaswamy / nature - Nature Medicine, Published online: 23 February 2026; doi:10.1038/s41591-026-04297-7A stress test of ChatGPT Health triage revealed missed high-risk emergencies and inconsistent activation of suicide-crisis safeguards, raising safety concerns for consumer-…

AI Summary: A structured, independent evaluation found that ChatGPT Health—an AI tool offering consumer triage and health guidance—missed or misclassified high-risk cases and gave inconsistent advice. Researchers and clinicians raised safety concerns about relying on the system for urgent medical decision-making, calling for tighter oversight and validation before broad public deployment.

#healthcare #publichealth #governmentpolicy #fda #digitalhealth #medicaldevices

Foundation AI model uses MRI data to predict multiple brain disorders

2 months / medicalxpress

Jame Abraham: Exploring the Impact of Artificial Intelligence in Cancer Care

2 months / oncodaily

Douglas Flora: The Future Is Agentic

2 months / oncodaily

World-first safety guide for public use of AI health chatbots

3 months / medicalxpress

Smartphone photos may be misleading doctors and putting patients at risk: New research

3 months / medicalxpress

Scientists made AI agents ruder — and they performed better at complex reasoning tasks