January 21, 2026

expert reaction to study on the association between Generative AI use and depressive symptoms among US adults

A study published in JAMA Network Open looks at the association between Generative AI use and depressive symptoms among US adults. 

Prof Chris Ferguson, Professor of Psychology, Stetson University, said:

“This study provides very little evidence that AI is correlated with mental health symptoms.  Most effect sizes appear to be near zero, suggesting there is next to no correlation between AI use and mental health.  Also of concern, all responses are self-report. 

“Unfortunately, people with depression are known to overestimate their technology use causing false correlations.  As such, there is very little reason to believe these tiny correlations are real as opposed to false positives and, even if they are real, they are so trivial in size as not to be worth worrying about. 

“Unfortunately, during periods of technology moral panic, weak evidence is often hyped irresponsibly to cause fear in the public.  That fear can be profitable for multiple stakeholders and the general public need to be wary of the tendency to hype weak evidence to support that fear.”

Prof David A. Ellis, Chair of Behavioural Science, University of Bath, said:

“What this paper shows is that among a sample of adults, those who report using AI also report slightly higher depression scores after controlling for some other factors. However, there are a number of considerations to keep in mind before drawing any strong conclusions. First, while the title implies that ‘use’ has been measured, the research relies on self reported use of AI. We know that this is a far from perfect measure of technology use when it comes to smartphones and social media. It will be equally problematic here. If this measurement issue is corrected, it may reduce what is already a very small effect to zero.

“Assuming the effect is not statistical noise, what is suggested is that AI use may harm mental health. However, there are three equally plausible explanations for the direction of the relationship reported in this work. AI use leads to depression, depression leads to AI use or there is in fact a third variable that has not been measured (e.g., social isolation). Indeed, the direction could be entirely reversed from what the title of the paper implies.”

Dr David Harley, chartered member of the British Psychological Society (BPS) and co-chair of the BPS’s Cyberpsychology Section, said:

“This appears to be a well-conducted piece of quantitative research which starts with a decent sample, uses standardised measures and shows a significant correlation between AI use and depression (and anxiety). This is a fairly straightforward replication of previous studies that have shown the same relationship existing between depression/anxiety and other forms of digital engagement (including social media use, smart phones and video games).

“Previous studies have also suggested a ‘displacement’ hypothesis, whereby digital activities are seen as displacing healthier face-to-face activities, such as socialising or playing sports, deemed to be essential for one’s mental health. No such causation has been suggested in this article in relation to AI. This makes sense given the nature of the study, but I think is hinted at if you know the research literature. 

“There are some caveats about the results of this study. It is not clear whether these results would generalise to other countries. Is American use of AI the same as the UK and other countries (how common is it in the workplace for instance)? Also, we know that the context of use is very important if we really want to understand AI’s relationship to mental health.

“The distinctions used here, i.e. personal, work or school, are not very informative, especially when it comes to AI where the context of use is not just concerned with physical location but also the psychological context, e.g. was it being used for emotional advice, to ascertain unknown information or to improve writing style?

“Finally, there is one interesting result that is not fully explored in the article and that is the lack of correlation between social media and AI use. This suggests that it could be affecting a whole different set of users. If this is indeed another form of displacement (which future research will explore) then it is probably not displacing social activities but something even more essential for mental health, such as reliance on our own consciousness.”

‘Generative AI Use and Depressive Symptoms Among US Adults’ by Roy H. Perlis et al. was published in JAMA Network Open at 16:00 UK time on Wednesday 21 January 2026. 

DOI: 10.1001/jamanetworkopen.2025.54820

Declared interests

Prof Chris Ferguson: “I have no conflicts of interest to declare.”

Dr David Harley: Dr Harley has no interests to declare in relation to this study.

Prof David A. Ellis: “My research has been or is currently funded by UKRI (ESRC/EPSRC) and NIHR. I’m a Member of the DSIT research commission “Feasibility Study of Methods and Data to Understand the Impact of Smartphones and Social Media on Children and Young People”. I have received funding or consultancy payments from We Adapt, Duco and Plexal, as well as payments from the BBC and Girton College, University of Cambridge.”

For all other experts, no reply to our request for DOIs was received.