Why are people turning to AI chatbots for therapy?
The rise of artificial intelligence in mental health care reflects both unmet need and growing concerns about safety.
A gap in services and a quick, 24/7 option
In regions where licensed therapists are scarce or unaffordable, people are using AI chatbots to get emotional support, practical coping tips, and even what they perceive as therapy outside regular office hours. For some users, the immediacy of an always-available conversational tool fills a gap when human care is unreachable — late at night or between appointments. Policymakers, clinicians, and advocates report that access barriers are a major reason for this shift.
What research and regulators are finding
- Multiple studies and reports have found that chatbots and large language models can give incorrect, incomplete, or potentially harmful medical advice.
- A randomized study on LLMs as medical assistants highlighted reliability concerns for the general public, and other investigations have documented instances where chatbots produced dangerous guidance or reinforced self-harm behaviors.
Immediate implications
- Safety risks: When models give wrong or harmful advice, vulnerable people can be put at risk.
- Need for oversight: Calls for clearer regulation and stronger guardrails are growing, because current technologies can behave unpredictably in clinical or crisis situations.
What remains uncertain
- How well regulated chatbot-driven care will become and what standards will be imposed for accuracy, transparency, and crisis escalation.
- How to balance innovation and access with protections that prevent harm, particularly for minors and people in crisis.
Why it matters
AI tools can expand access to mental-health support but do not replace skilled clinicians. The challenge for health systems and regulators is to keep the benefits — increased reach and convenience — while limiting harms through verification, clinical integration, and enforceable safety standards.