What happened: A University of Ottawa study of nearly 40,000 Ontario students found about one in five in grades 4-12 used an AI chatbot for emotional support, not just homework, per findings in JAMA Pediatrics.
Key points:
These students were nearly twice as likely to report clinically significant emotional problems like anxiety or depression than their peers.
They also reported higher levels of loneliness and lower feelings of 'mattering,' or feeling valued by other people.
Why this is important: The study shows a correlation, not causation, between AI use and poorer mental health. Researchers suggest this behavior is a signal of underlying distress, prompting conversations about a child's feelings rather than just restricting technology.
What happened: Seattle startup NewDays expanded its seed funding to $16 million and launched its AI-driven dementia care platform in Nevada. The service combines telehealth visits with an AI companion to provide cognitive strategies.
Key points:
In a limited study of 24 patients, the company reported its platform preserved cognitive function for 18 months.
The service is covered by traditional Medicare and is now undergoing a randomized controlled trial with Kaiser Permanente.
Why this is important: NewDays aims to make clinically proven but historically inaccessible cognitive strategies scalable using its AI companion, Sunny. This blended model of human oversight and AI could expand specialized care for people with cognitive decline.
What the author argues: Undisclosed AI use by clinicians or clients—a “shadow artificial third”—can threaten the therapeutic relationship by creating a sense of betrayal. Wei argues that even when AI-generated messages seem empathic, they risk homogenizing language, diminishing authenticity and undermining the human connection central to therapy.
What the author argues: California’s SB 903 is useful but fails to address the main source of AI-related psychiatric harm. Gillen argues that the law’s narrow focus on products marketed as “therapy” misses the general-purpose chatbots that people in crisis actually use, leaving a significant regulatory gap.
News at the Intersection of Mental Health and Technology
DAILY
Archive edition assembled in September 2026 from saved reporting; coverage cutoff 6 a.m. Eastern on 2026-09-03. Featured-image illustration generated with AI.
Leave a Reply