Young adults who use generative artificial intelligence to ask health-related questions are more likely to screen positive for clinically significant anxiety and depression. This finding comes from a new study led by University of Florida researcher Yusen Zhai.
The research analyzed data from nearly 100,000 young adults. It found that the odds of screening positive for anxiety and depression were 52 percent and 46 percent greater for people who reported using tools like ChatGPT or Claude for health questions.
However, the findings only show an association between the two behaviors. The data does not establish that using generative AI directly causes anxiety or depression. It also does not prove that people with mental health struggles are necessarily more likely to use AI for health questions.
Zhai, an assistant professor in the UF College of Education, said he was intrigued by the findings. While concerns about AI often focus on academic productivity, his research examines how technology can identify and address mental health risks.
“We are learning that young people are turning to AI, especially generative AI systems, when they feel worried or vulnerable or are trying to understand their health symptoms,” Zhai noted.
Among participants who used AI for health questions, about 52 percent screened positive for anxiety, compared to 43 percent among non-users. For depression, the figures were 47 percent and 38 percent. These associations remained even after researchers accounted for factors like age, race, biological sex, and socioeconomic status.
Zhai warned that repeatedly turning to AI for reassurance can create a harmful cycle. A user might receive an answer, develop additional uncertainties, and ask more questions. He explained that this repeated cycle could reinforce rumination, health anxiety, or other distress.
Unlike a traditional internet search that provides links, generative AI responds directly and conversationally. This interaction can feel highly persuasive, even when the generated information is incomplete or inaccurate.
Because the study relies on data from 2023 to 2024, Zhai views the findings as an early indication rather than a complete picture. He suggests that universities and mental health professionals should incorporate AI literacy into their guidance, helping young adults evaluate AI-generated answers safely.
