Research
Perlis RH, Gunning FM, Uslu AA, et al. Generative AI Use and Depressive Symptoms Among US Adults. JAMA Netw Open. 2026;9(1):e2554820. doi:10.1001/jamanetworkopen.2025.54820
Faculty Authors
How is AI use connected with depression?
Since Chat GPT 3.5 emerged in late 2022, generative artificial intelligence has become a source of fascination and fear for the public. As more people began to use this technology, reports surfaced about its harms to mental health, from the potential of chatbots to increase suicide risk to their ability to exacerbate delusions. Researchers have found associations with AI use and depression and isolation, as well as potential benefits for chatbots trained to discuss mental health.
New research by lead author Dr. Roy Perlis at Harvard Medical School and co-authors including Matthew Baum, Martin Kalb Professor of Global Communications at Harvard Kennedy School, explored the associations between how—and how much—American adults used AI, and the extent to which they showed symptoms of depression.
Greater use of AI associated with greater depression, anxiety, and irritability
In a recent article in JAMA Open Network, “Generative AI Use and Depressive Symptoms Among U.S. Adults,” researchers analyzed a survey of over 20,000 people, exploring demographic features such as race and income alongside answers to screening questionnaires for anxiety and depression.
They found that daily or frequent use of AI is significantly associated with greater levels of depressive symptoms. The odds of reporting moderate depression were 30% higher among people who used AI each day. The researchers found similar patterns for symptoms of anxiety and irritability. The odds of reporting at least moderate depression were 50% greater for people aged 45-65 who used AI daily.
The researchers also found that frequent AI use is that more common among certain groups, including:
- Men
- Younger adults
- People in cities
- People with higher levels of education
- People with higher incomes
Symptoms of depression were highest among people using AI for their own personal use, and for people between 25-44 years old.
What’s next for understanding the relationship between AI and mood?
The researchers point to two randomized clinical trials that found that abstaining from social media could have a positive (but very modest) effect for some people. The authors consider exploring similar approaches around the use of AI. Since their findings are associations, not causations, the researchers also suggest exploring any potential causal relationship between AI use and mood.