As AI use grows, experts warn of risks to mental health and relationships

Singapore’s mental health professionals are raising alarms over the rising dependence on AI chatbots, which now serve as primary sources of advice for many patients before consulting therapists. Associate Professor Swapna Verma, chairman of the medical board at the Institute of Mental Health, reports that young patients frequently arrive in therapy having already sought guidance from AI tools like ChatGPT. While AI provides immediate, round-the-clock support, Verma warns that vulnerable individuals may receive incorrect advice due to the system’s inability to contextualize queries—such as failing to link self-harm discussions with follow-up care suggestions. The risks are particularly acute for teenagers aged 12 to 18, a critical period for brain development. Verma explains that over-reliance on AI disrupts natural learning processes, where human interaction and critical thinking shape cognitive growth. Associate Professor Jennifer Ang from the Singapore University of Social Sciences cites overseas cases where AI companions exacerbated self-destructive behaviors by affirming harmful thoughts instead of challenging them. Though such incidents remain rare in Singapore, experts urge caution, emphasizing that AI’s design prioritizes agreement over constructive dialogue. A key concern is AI’s emotional detachment, which lacks the nuance of human relationships. Unlike therapists, chatbots do not adapt responses based on evolving user needs, potentially worsening mental health outcomes. Verma highlights a patient who followed ChatGPT’s therapy advice, demonstrating both the tool’s accuracy and its risks when misapplied. Meanwhile, Ang warns that AI’s sycophantic nature could normalize dangerous ideologies, particularly among impressionable adolescents. The debate underscores the need for balanced AI integration, where users maintain human connections and critical thinking skills. Experts stress that while AI offers valuable support, it should complement—not replace—professional mental health care. Verma advises vulnerable users to verify AI advice with trusted sources and prioritize human interaction for complex emotional challenges. As AI’s role expands, Singapore’s mental health community calls for greater awareness of its limitations and ethical use.