A new risk assessment from Common Sense Media and Stanford’s Brainstorm Lab warns that AI-powered student mental health apps are largely unregulated and can be harmful, particularly when used directly by consumers without safeguards. Researchers found that school-appropriate “institutional” apps with humans in the loop performed better on safety and helpfulness measures than consumer versions. The assessment tested apps’ ability to recognize warning signs across conditions such as anxiety, ADHD, depression, and psychosis, and to provide appropriate crisis resources and referrals. It also evaluated whether apps might delay professional care or provide harmful guidance. With three in 10 teens using AI mental health apps and many more using general-purpose chatbots for emotional support, the report urged schools to treat these tools as one component of support—not a replacement—amid ongoing staffing shortages for counselors and psychologists.