Mental Health Innovation Is Scaling Faster Than Accountability

Digital mental health tools, including chatbots, symptom checkers, artificial intelligence (AI) companions, digital therapeutics, and passive monitoring systems, are rapidly moving from peripheral wellness products into care-influencing roles. In mental health, this shift creates a distinctive accountability problem: tools may encourage disclosure, shape symptom interpretation, influence help-seeking decisions, or provide therapeutic-like interaction without assuming the duties traditionally attached to clinical care. This editorial argues that access and scalability, although important, are insufficient markers of trustworthiness. Governance should be based not only on product labels or medical-device claims, but also on the functions these systems perform and the foreseeable ways in which users rely on them. We propose a tiered model of digital mental health vigilance, with obligations that increase as tools move from low-risk wellness support toward symptom interpretation, triage-like guidance, therapeutic intervention, autonomous monitoring, or clinician-supervised treatment. Such vigilance should include clear scope limitations, privacy safeguards, user reporting channels, escalation pathways, adverse-event and near-miss reporting, performance-drift monitoring, periodic audit and explicit harm-remediation responsibilities. The analogy with pharmacovigilance is useful but incomplete because digital mental health tools evolve rapidly, lack defined exposure, and often operate without a professional intermediary. Proportionate, function-based oversight can help protect users without undermining access, particularly in low-resource settings where treatment gaps are wide and regulatory capacity may be limited. Mental health innovation may improve care, but systems that ask for clinical trust must also be prepared to account for their effects.