- A new study investigates the link between the use of artificial intelligence (AI) and symptoms of depression.
- They concluded that those who used AI most frequently were more likely to experience symptoms of depression than those who never used AI.
- Interestingly, the results varied widely between age groups.
A group of researchers published a paper.
Researchers analyzed survey data from nearly 21,000 participants across 50 states and concluded that higher rates of AI use increased the likelihood of experiencing symptoms of depression, anxiety, and irritability.
Although AI has only been around for a few years, it is already a constant part of many people’s lives. Whether you ask ChatGPT how to make chicken curry or use ChatGPT to summarize data tables at work, you can’t ignore it.
As with any new technology, understanding its impact on users’ health is an important area of research.
There’s already some anecdotal evidence that chatbots can make people more paranoid and potentially encourage suicidal thoughts. Despite these worrying early signs, there is little academic research focused on AI use and mental health. The latest research goes some way toward filling that gap.
Researchers accessed survey data collected in April and May 2025. This included 20,847 participants across 50 states, all of whom were 18 years of age or older. This data included a balanced mix of information from different age groups, ethnicities, and genders in each state.
As part of the survey, all participants were asked, “How often do you use the following technology or product? – Artificial Intelligence (AI).”
Possible responses were “never,” “once or twice,” “about once a month,” “about once a week,” “more than once a week,” “every day,” and “more than once a day.” They were also asked if it was for personal, work, or school use.
In addition to this information, the researchers had access to personal data such as household income and educational background.
The analysis found that 10.3% of the sample used AI daily, and 5.3% of these used AI multiple times each day.
Among daily users, about half use it for work, 11.4% for school, and 87.1% for personal reasons.
Unsurprisingly, AI usage varies by demographic. Those who used AI the most were more likely to be younger, male, more educated, have higher household incomes, and live in urban rather than rural areas.
Today’s medical news contacted corresponding author Roy H. Perlis, MD, MSc, professor of psychiatry at Harvard Medical School and clinical researcher at the Mass Research Institute in Boston, Massachusetts.
“The odds of developing at least moderate depression (the threshold at which clinicians typically refer people for evaluation and treatment) were 30% higher in people who used AI at least daily,” he explained. A similar pattern emerged when looking at anxiety and irritability.
When we looked a little deeper into the data to see if type of use mattered, we found that the association with mental health problems was only significant for personal use, not work or school use.
The researchers found no differential association between AI use and gender, but they did find an association with age group. Specifically, in the 25-44 and 45-64 age groups, AI use was associated with more depressive symptoms.
As with any research, especially in a new field, each new study generates more questions than answers.
First of all, it’s a question of cause and effect. “We don’t know if it’s the cause, the effect, or neither,” Perlis said. MNT. “Only randomized trials will tell us for sure, but following people over time may be helpful.”
Regarding this question, he also said, “It is certainly possible that people with more severe depressive symptoms may use AI more frequently.”
Another unanswered question is its relationship to age. A similar association between depressive symptoms and AI use was not shown for people younger than 25 years or older than 64 years. “This group may be more likely to use AI in certain ways and in certain settings, but we don’t know yet,” Perlis said. MNT.
Certain specific forms of AI usage may be more problematic than others. Although the survey used in this study was not designed to collect such information, Perlis believes this could be an important area for future research.
“As a psychiatrist, I’m certainly concerned that people are using AI in place of social interaction. In fact, this was what motivated me to look into this connection in the first place.”
MNT We also contacted John Pulse, LCSW, MCAP, a psychotherapist and addiction specialist at Full Life Comprehensive Care, who was not involved in the study.
“I wasn’t surprised at all by this result,” he told us. “Using multilingual AI often prevents you from having meaningful conversations with others and getting support.”
This can lead to increased feelings of isolation and loneliness, he explained. Pals also said, “When AI does everything for us, including thinking and problem-solving, it’s natural to start to feel like we’ve lost purpose, which can lead to depression.”
While the study’s findings may be seen as alarming, Perlis is hopeful that personal use of AI could still have benefits if implemented in the right way.
“There’s no question that chatbots could be a promising intervention for people who don’t have access to talk therapy,” he explained. “But only if those chatbots are carefully designed and monitored, and ideally used in conjunction with human therapists.”
Dr. Mil Brown, chief medical officer at Spring Health, who was not involved in the study, agreed that AI has the potential to:
“We’re already seeing evidence that high-quality, clinically designed mental health AI tools can reduce symptoms, facilitate engagement with others, and help people access support who otherwise wouldn’t have it. AI also has the potential to help clinicians do their best work while improving outcomes.”
Importantly, Pals also said that the current research does not exclude the possibility that some individuals may benefit from AI.
“Certainly, AI can be a force for good,” Pals says. MNT. “Reducing someone’s work load and improving their mental health. But if that is done at the expense of real human relationships, it will have significant negative effects.”
The large size and scope of the survey allows us to place some confidence in the results. But as Perlis said, MNT“As an internet survey, our sample is not completely representative and bias may exist.”
MNT We also contacted Radial’s psychiatrist, Dr. Owen Muir (DFAACAP), who agreed with this view. “If you think about who is more likely to talk to a chatbot multiple times a day, you are selecting people who are lonely and more likely to be depressed.”
“We hope that other groups will try to replicate these results using other samples to increase confidence in our results. We are also interested in longitudinal studies to see how mood changes over time with the use of AI,” Perlis said.
“And, of course, the most direct test would be a randomized trial, but this type of study is difficult to conduct.”
Many experts are optimistic about the potential of AI to improve mental health if used properly. “I have spent my career building AI algorithms for use in mental health care, and I think AI can be a very powerful force for good,” Muir said.
