OpenAI announced on September 1, 2026 that ChatGPT for Healthcare can now connect directly to Epic's EHR system, with UCSF Health as an early pilot site testing read-only, chart-grounded clinical queries.
A 2026 real-world study of the popular SkinVision AI mole-scanning app found accuracy affected by ordinary users' photo quality, reviving debate over whether AI skin-cancer apps expand access to screening or create false reassurance that delays care.
Hearing aid makers led by Starkey are adding AI-driven fall detection, activity tracking and heart rate monitoring to devices already worn all day by older adults, though researchers say real-world fall-detection accuracy still needs independent validation.
A Florida hospital faces a lawsuit alleging an AI diagnostic tool's low-risk flag delayed a patient's cancer diagnosis, raising new liability questions for hospitals and vendors deploying clinical AI tools.
A Florida pastor is suing OpenAI, alleging ChatGPT downplayed his symptoms for months before he suffered a near-fatal pulmonary embolism, in a case testing how courts treat AI chatbots giving health-adjacent advice.
Chai Discovery closed a $400 million Series C at a $3.8 billion valuation, tripling its price tag in seven months, as OpenAI, Eli Lilly, and Index Ventures back its AI-driven antibody design platform.
Germany's nationwide PRAIM study found AI-assisted mammogram reading raised the breast-cancer detection rate by 17.6%, adding to a wave of 2026 studies from Sweden and South Korea supporting AI as a screening aid.
Abridge's ambient AI documentation tool, built with Epic and Mayo Clinic nurses, is now live at more than 250 U.S. health systems, extending AI scribing from physicians into nursing, ER, and post-acute care.
The EQUAL trial found Apple Watch-based AI heart monitoring detected atrial fibrillation in 9.6% of high-risk older adults versus 2.3% with standard care, bolstering the case for wearable-based stroke prevention.
Illinois's new WOPR Act bans AI chatbots from delivering psychotherapy or clinical decisions, with fines up to $10,000 per violation, making it the latest of four states to restrict AI mental-health tools in 2026.