Common Sense Media and Stanford’s Brainstorm Lab released a risk assessment showing that while teens increasingly use AI-powered mental health apps, the market remains largely unregulated and some products can be harmful. The report compared school-oriented “institutional” apps with direct-to-consumer apps and found the school-based models that keep humans in the loop performed better on safety and helpfulness metrics. The assessment evaluated whether apps could recognize warning signs across conditions such as anxiety, ADHD, depression, and psychosis, and whether they accurately assessed severity and directed users to crisis resources or professional care. Common Sense Media also reported that about three in 10 teens have used AI mental health apps, with many also using general purpose chatbots such as ChatGPT for emotional support. For districts dealing with psychologist and counselor shortages, the practical takeaway is not broad adoption but guarded use—positioning school-based tools as supplemental supports rather than replacements for clinical staff. Common Sense Media’s Youth AI Safety Institute head Robbie Torney said these tools can help, but they cannot be the only part of support. The development adds to the policy and procurement challenge for K–12 and higher education support systems: balancing demand for rapid assistance with requirements for safety, human oversight, and crisis escalation reliability.