JAMA +AI Update July 4, 2026

As mental health systems struggle to meet demand, AI is starting to fill the gap. But can AI really replace complex human interactions, and will it improve mental health outcomes?

In this episode of JAMA+ AI Conversations, JAMA+ AI Editor in Chief Roy Perlis, MD, MSc, and JAMA+ AI Associate Editor Yulin Hswen, ScD, MPH, discuss the expanding debate on AI ethics as more people explore AI use to augment clinical care. They cite recent studies published in JAMA Network Open and JAMA Psychiatry on the use of AI for mental health advice, psychiatric documentation, and thinking like a therapist. Their wide-ranging discussion highlights an emerging evidence gap.

Listen now on Spotify | Apple Podcasts | YouTube | JAMA.com.

Editor’s Picks in this week’s JAMA+ AI Update:   

• A machine-learning model improved the ability to predict aneurysm rupture risk in patients with intracranial aneurysms. The study suggests that future data-driven approaches could go even further in predicting rupture risk, identifying the drivers of pathogenesis, and exploring ways to reduce the risk. (JAMA Network Open)
• In a diagnostic study of 278 children with cochlear implants to treat bilateral sensorineural hearing loss, deep transfer learning algorithms better predicted spoken language development compared to traditional machine learning. The improved prediction ability may enable customized interventions to improve language outcomes in children following cochlear implant. (JAMA Otolaryngology–Head & Neck Surgery)
• Pediatric clinical AI tools need to strike the right balance between speed and safety: premature deployment risks harm but delaying validated tools can lead to preventable illness and death. A conceptual child-in-the-loop framework for AI puts children and families at the center of conversations about consent, communication, and shared decision-making to enhance pediatric health care. (JAMA Pediatrics)