As more people turn to AI for emotional support, a licensed clinical social worker explains what it can and cannot – and when it's time to seek human help
Late at night, when anxiety gets louder and reaching out to another person feels like too much, an AI chatbot can seem easy to talk to.
It does not judge. It does not interrupt. It answers immediately.
For people who are lonely, overwhelmed or afraid of being misunderstood, that can feel comforting. But mental health experts caution that while artificial intelligence may offer convenience or a place to organize thoughts, it is not a substitute for trained, ethical and accountable care.
The World Health Organization recently identified the growing use of generative AI tools for emotional support, particularly among young people as an emerging public health concern.
"In a world where people are increasingly isolated and overwhelmed, it makes sense that some would turn to AI as a companion, as a confidant," said Natalia Giordano, assistant teaching professor in FIU's School of Social Work. "There are major concerns, however, about what information is being shared by AI. It's not trained, regulated or monitored in the ways that mental health professionals are."
Why people open up to AI
As Mental Health Awareness Month comes to a close, Giordano says the growing use of AI for emotional support raises an important question: What are people really looking for when they turn to a chatbot?
"People don't want to be judged, and AI is seen as a safe, neutral place to share their deepest thoughts and feelings," Giordano said. "It's right at our fingertips and won't judge or make assumptions about us."
For some, AI may be a low-risk place to organize thoughts or seek general information. Giordano said it can be useful in limited ways, but people should approach AI-generated responses like any online search: with critical thinking and caution.
When support becomes risky
The line between support and risk becomes clearer when someone begins to rely on AI as their primary source of emotional care.
A 2025 Brown University study found that AI chatbots used for mental health support can violate core ethics standards, including by mishandling crisis situations, reinforcing negative beliefs and creating a false sense of empathy. Recent research has also found that AI chatbots can be overly agreeable when people seek personal advice, sometimes reinforcing a user's beliefs or behavior rather than challenging them. In mental health conversations, that tendency can be risky, especially when someone needs careful assessment, grounding or immediate support.
Giordano said those risks become especially concerning when someone is already vulnerable.
"My top concern is that AI will share inaccurate information and encourage an adolescent or young adult to harm themselves or others," Giordano said. "I also worry about the potential disconnection or isolation from human beings."
That does not mean every use of AI is dangerous. For someone preparing for a difficult conversation, sorting through their thoughts and emotions, or identifying resources, AI may serve as a tool. But Giordano cautions against allowing it to replace self-reflection.
"It can be a useful tool for fine-tuning," she said, "but it can be harmful to rely on it to create a sense of self."
Natalia Giordano is a licensed clinical social worker and assistant clinical professor in FIU Robert Stempel College of Public Health and Social Work. Her work focuses on child welfare, youth development, and the transition to independent living for adolescents and young adults. She teaches practice courses and child welfare electives for undergraduate and graduate students and brings experience providing counseling, case management, and life coaching, as well as developing programs in employment, education, and life skills. Her work is grounded in healing-centered, trauma-informed, and empowering relationships with students, interns, and professionals serving young people.
