Our 2025 AI & Mental Health Emergencies Survey of 1,000 Americans reveals unexpected acceptance of AI crisis detection — with crucial limitations.
Key Finding: 49% would allow AI to monitor their mental health patterns for early crisis detection, but 73% insist humans make all final care decisions.
The research challenges assumptions about AI resistance in healthcare. While consumers accept AI surveillance for potentially life-saving detection, they firmly reject automated decision-making in mental health emergencies.
Notable insights:
- Privacy concerns are less of a barrier than expected
- Significant demographic differences across gender, age, and income
- AI most effective as rapid alert system, not decision-maker
- Human oversight remains non-negotiable
The findings provide critical guidance for healthcare systems implementing AI crisis response protocols, supporting an “enhancement, not replacement” model for behavioral health technology.
Survey conducted August 2025 among a national sample of 1,000 U.S. consumers.
