As mental health demands soar and human resources remain limited, AI-powered tools are stepping into the gap. In 2025, we must ask: Can machines truly listen and support when people turn to them for help?

According to www.futuremarketinsights.com, the global market for AI-enabled behavioral therapy is estimated at USD 2.36 billion in 2025, with expectations to grow fivefold by 2035. Meanwhile, a survey found that 32% of individuals would consider using AI-based therapy, though the majority (68%) still prefer human therapists (artsmart.ai).

AI chatbots and virtual companions are now being deployed in schools, workplaces, and digital wellness platforms to provide 24/7 text-based emotional support. Some early clinical results are promising: one review noted that AI tools delivered comparable outcomes to traditional therapy in certain mild-to-moderate conditions.

However, significant risks remain. A 2025 study assessed 29 AI chatbots’ responses to suicide-risk scenarios and found none met adequate standards; roughly 48% of responses were judged inadequate. Similarly, researchers at Stanford University warn that AI systems may reinforce stigma, deliver unsafe responses, and lack the relational depth of human therapists. The takeaway: AI can assist, but it’s not ready to replace human care. It may offer timely access or augment services, but emotional intelligence, crisis intervention, and ethical oversight still hinge on human judgment.

AI therapy is no longer science fiction. It offers real possibilities, especially for those waiting for help or feeling reluctant to reach out. But, as experts emphasize, machines listen, yet may not understand in the way people do. The question for 2025 and beyond is whether we build AI to support human-centered care, rather than replace it.

고은별 강남포스트 학생기자  webmaster@ignnews.kr

<저작권자 © 강남포스트, 무단 전재 및 재배포 금지>