Social good & healthCognition & learningTechnical failuresPrivacySystemic riskUnited KingdomResearch
AI chatbots matched human emotional support in everyday situations
Five studies involving 1,233 participants compared responses from ChatGPT 4o, Claude 3.5 Sonnet, Gemini 1.5 Pro, and human participants across everyday, non-clinical emotional situations. The AI responses were rated as more supportive for anger and fear, performed about as well as people for sadness, and still helped when recipients correctly suspected they came from a machine. The strongest factor was not generic validation but specific, actionable guidance.
AI can provide useful everyday emotional support, but performance in controlled, non-clinical scenarios does not establish safety for therapy, crisis response, sensitive-data handling, or sustained emotional dependence.
Why it matters
The result points to a practical design lesson: support improved when responses offered concrete next steps such as breathing exercises, reframing, or a usable plan. That insight can improve both chatbot design and human support, while avoiding the mistaken conclusion that a convincing response is equivalent to a durable relationship or professional care.
The research studied everyday scenarios rather than clinical treatment or emergencies. It did not establish whether chatbots recognize escalating risk reliably, protect intimate disclosures, avoid dependency, or remain helpful over time. Those unresolved questions become more important as people increasingly use conversational systems for emotional needs.
Primary trail
Go to the source
Read the evidence behind this analysis. External links open in a new tab.
