GAO finds no FDA-authorized mental health chatbot; diabetes psychologists ease distress but not HbA1c; resistance training eases depression
Listen to this post
read by AI voices · all episodesThe Government Accountability Office (GAO) said Nov. 20 that the Food and Drug Administration has authorized no AI chatbot for mental health, though nearly one-fifth of Americans aged 12 to 21 used one in 2025.
A Swedish trial found diabetes-trained psychologists eased distress without lowering HbA1c, a meta-analysis linked resistance training to fewer depressive symptoms and a Norwegian trial of Goal Management Training for adolescents with ADHD was negative.
- AI Nov 20, 2026The Government Accountability Office (GAO) on Nov. 20 published a Science & Tech Spotlight on AI chatbots and mental health, reporting that nearly one-fifth of Americans aged 12 to 21 used an AI chatbot for mental health advice in 2025. The Food and Drug Administration (FDA) has not authorized any AI chatbot to support mental health, GAO said, and some chatbots tend to keep conversations going and agree with the user in a way that can reinforce delusions. Two companies have advertised an FDA designation for their purpose-built chatbots, which does not guarantee future authorization, according to GAO.Source: GAO
- Research Nov 20, 2026In a 52-week randomized trial at six Swedish clinics, published in Diabetes Care, 151 adults with type 1 diabetes and HbA1c above 7.8% received conventional care or care plus individual counseling from psychologists with structured diabetes training, attending a median of nine sessions. HbA1c fell 0.58% with the psychologist and 0.36% with conventional care, a difference that was not significant (mean difference -0.2%; 95% CI -0.6 to 0.2; P = .31). Diabetes distress decreased (P = .040) and treatment satisfaction improved (P = .004), while diabetes-related quality of life did not differ.Sources: Diabetes Care, PubMed
- Research Nov 16, 2026A JAMA Network Open meta-analysis of 80 randomized trials with 6,123 participants found resistance exercise training associated with a small antidepressant effect across all comparisons (effect size 0.32; 95% CI 0.25 to 0.39). Against active comparisons including relaxation, mind-body exercise, education and aerobic exercise, the effect was 0.21, larger in participants with a mental illness (0.40; 95% CI 0.29 to 0.50) than without one (0.11; 95% CI 0.04 to 0.18). The authors said the findings support recommending resistance training in the treatment of depression.
- Research Nov 16, 2026Using Population Assessment of Tobacco and Health Study data from 2013 to 2021, researchers studied U.S. young adults aged 18 to 24 with elevated internalizing problems and social media use several times a day, across 3,788 person-interval observations. About 14% cut back to once daily or less, and reduced use was associated with a greater likelihood of nonelevated internalizing symptoms at the next wave (0.57 versus 0.49; prevalence ratio 1.15; 95% CI 1.04 to 1.27). The authors of the cohort study called reduced use a potentially modifiable target for prevention and early intervention.Sources: JAMA Pediatrics, PubMed
- Research Nov 18, 2026In a preregistered multicenter randomized trial led from the University of Oslo, 116 adolescents with attention-deficit/hyperactivity disorder (ADHD) received seven weekly two-hour Goal Management Training group sessions plus treatment as usual, or treatment as usual alone, according to Research on Child and Adolescent Psychopathology. At 12 weeks, the parent-reported BRIEF-2 Global Executive Composite, the primary end point, did not differ significantly between groups (adjusted b = -2.12; 95% CI -7.64 to 3.41; P = .449), nor did parent-reported ADHD symptoms or other secondary outcomes.
The weekly innovation roundup for psychology and social work, published every Sunday for the week before; every item links to its sources. Latest psychology and social work news · All specialties