Half of distressed Israeli users turn to chatbots for emotional support

Researchers at Tel Hai College of the Galilee and Bar-Ilan University published findings showing that 47% of Israeli adults with severe psychological distress used general-purpose AI for emotional support. The study, released in the journal Cyberpsychology, Behavior, and Social Networking, analyzed data from 584 adults aged 18 to 70. This group included individuals who had used a generative AI tool at least once in the year before the survey was conducted in November 2025.

The data indicates that psychological distress is the primary driver for seeking this type of assistance. Among all respondents, 29.3% reported turning to AI for coping advice or mental health related support. This figure rises sharply to 47% among the 66 participants whose distress levels were rated as severe. An additional 119 participants reported medium levels of distress. The lead author, Dr. Dorit Hadar Shoval, noted that these measures track emotional stress rather than clinical mental illness.

Demographic factors also influenced usage patterns. Women and younger adults were more likely to report using AI for emotional support than men and older participants. While higher attachment anxiety initially appeared linked to such use, the connection disappeared once overall distress levels were accounted for in the analysis. The study could not determine if distress led to AI use or if AI use contributed to distress.

Hadar Shoval explained that users are drawn to these systems because they are available and non-judgmental. She stated that conversational responses can encourage emotional attachment. However, the researchers warned that consumer tools like ChatGPT and Gemini are not designed for clinical care. These platforms lack the specific crisis safeguards and professional responsibilities found in traditional therapy.

The survey did not measure whether these interactions helped users cope or delayed professional treatment. It also did not clarify if participants used chatbots as a substitute for clinical help. The reported usage ranged from single exchanges to sustained engagement. Hadar Shoval said shortages in accessible public mental health services help explain why people turn to these alternatives.

The researchers called for closer scrutiny from clinicians and policymakers. They highlighted the need to investigate whether AI supplements or replaces professional assistance. Hadar Shoval cautioned against trusting chatbots to recognize acute mental health crises. She stressed that professional help remains necessary for serious problems.

The study was led by Dr. Dorit Hadar Shoval of the School of Psychology at Tel Hai College of the Galilee in Kiryat Shmona. Elad Refoua of Bar-Ilan University in Ramat Gan also led the research team. The findings describe this specific sample of AI users and are not considered representative of the entire Israeli population.