Generative AI Outperforms Human Support in Emotional Well-Being

Recent research reveals that generative artificial intelligence (gen AI) can provide emotional support that some users find more comforting than responses from human beings. A 2025 survey of U.S. individuals aged 12 to 21 indicated that 13% sought advice from gen AI when feeling sad, angry, or nervous, with the figure rising to 22% among those aged 18 to 21. This trend highlights a significant shift in how younger generations are engaging with technology for emotional and mental health support, suggesting that businesses in the mental health and technology sectors may need to adapt their strategies to meet this evolving demand.

The study conducted by researchers from the universities of Manchester and Durham involved five experiments comparing emotional-support messages from humans and large language models, such as those behind ChatGPT. Participants were presented with scenarios involving anger, sadness, and fear, and they rated the responses without knowing their source. Notably, responses generated by gen AI received higher ratings for emotional support in scenarios involving anger and fear. This finding suggests that gen AI may not only be a viable alternative for emotional support but could also outperform human responses in certain contexts.

The research also aligns with a broader review of 23 studies indicating a trend where users rated gen AI messages as more empathetic than human-generated content. This phenomenon, termed the "AI advantage," may stem from the consistency and structure of AI-generated responses. Unlike humans, who may struggle to articulate their thoughts due to fatigue or uncertainty, gen AI can consistently provide actionable suggestions that users find helpful. This aspect of gen AI could signal a shift in how emotional support is conceptualized and delivered, particularly in industries focused on mental health and wellness.

However, the preference for human interaction remains strong. The same studies revealed that when participants believed a supportive message came from a human, they rated it as more empathic and supportive. This suggests that while gen AI can effectively deliver structured support, the emotional connection and perceived effort associated with human responses still hold significant value. The nuances of personal relationships and the context in which support is given play a critical role in how individuals perceive and benefit from emotional assistance.

As businesses consider integrating AI into their customer support and mental health services, they must recognize the limitations of gen AI. Current research primarily examines brief interactions and does not account for the long-term implications of relying on AI for emotional support. The dynamics of ongoing human relationships, characterized by personal knowledge and emotional investment, cannot be replicated by AI. Consequently, companies should focus on how to leverage AI's strengths while maintaining the essential human elements that foster deeper connections.

Looking ahead, organizations in the mental health and technology sectors should explore hybrid models that combine the efficiency of gen AI with the emotional depth of human interaction. Training staff to utilize AI tools effectively while ensuring that human support remains accessible could enhance overall service delivery. As gen AI continues to evolve, its role in emotional support will likely expand, but businesses must remain vigilant in balancing technological advancements with the irreplaceable value of human empathy and connection. This strategic approach could position companies to better serve a generation increasingly inclined to seek solace in technology.