Individuals who interact with social chatbots tend to be younger and experience higher levels of psychological distress than those who do not use such technology. The findings suggest that while these artificial intelligence tools are often designed and marketed to provide companionship, their use is frequently associated with feelings of loneliness and emotional struggle rather than improved well-being. The study was published in the Journal of Social and Personal Relationships.
Automated programs known as social chatbots use natural language processing to mimic human conversation. Some developers market these tools as potential friends or sources of emotional support.
Despite the growing popularity of these applications, scientific understanding regarding who actually uses them remains limited. Previous inquiries often focused on single applications or small groups, leaving a gap in knowledge about broader user characteristics.
It is unclear whether these tools attract individuals who are already socially secure or those seeking to alleviate deficits in their social lives. The authors aimed to identify the demographic and psychological profiles of social chatbot users across different cultural contexts.
“Humans have an innate need to build and maintain meaningful relationships, and in today’s digital world, many of these connections increasingly unfold through technology,” said study author Iina Savolainen, a senior research fellow at Tampere University and vice director of the Emerging Technologies Lab.
“With the rise of AI, more people are using social chatbots to explore new forms of communication or to seek companionship, emotional support, or simply, everyday interaction. Yet despite this growing trend, surprisingly few empirical studies have examined who uses these tools and how such usage relates to well-being. We saw a research gap and wanted to better understand these dynamics through a large, cross-national study.”
To address these questions, the research team analyzed data collected in the autumn of 2023 as part of the longitudinal “Self & Technology” study. The analysis included responses from adults in Finland (1,095 participants), France (1,014), Germany (900), Ireland (588), Italy (1,099), and Poland (967). These nations were selected to represent diverse geographic and cultural regions within Europe, ranging from the Nordic technology hub of Finland to Southern and Eastern European contexts.
Participants completed online surveys assessing their usage of “chatbot friends” such as Replika or My AI. They provided information regarding their frequency of use, ranging from never to multiple times a day. The researchers created a binary category to distinguish between those who used the technology and those who did not.
The team measured psychological distress using the Mental Health Inventory, which asks about feelings of nervousness, depression, and calmness over the past month. Loneliness was assessed using a short version of the University of California Loneliness Scale. The researchers also evaluated self-esteem using a single-item measure.
In addition to mental health metrics, the survey gathered data on the frequency of face-to-face social contact with friends, relatives, or colleagues. Participants reported their general attitudes toward new technologies. They also provided sociodemographic details including age, gender, income, and relationship status.
The researchers found that social chatbot usage varied by country, ranging from a low of 8.67% in Ireland to a high of 17.93% in Italy. Across all six nations, a clear demographic trend emerged regarding age. Individuals who used social chatbots were consistently younger than those who did not.
A primary finding concerning mental health appeared in every country sampled. The analysis showed a positive association between social chatbot usage and psychological distress. This indicates that people engaging with these AI companions were more likely to report symptoms of anxiety, depression, or low mental well-being compared to non-users.
“The cross-cultural consistency was striking; social chatbot use was related to poorer mental well-being in all six countries,” Savolainen told PsyPost. “That broad consistency across different European contexts was quite surprising.”
The relationship between chatbot use and loneliness was significant in four of the six countries: France, Germany, Italy, and Poland. In these nations, users reported higher levels of loneliness than non-users. This pattern was not statistically significant in the data from Finland or Ireland.
Regarding self-esteem, the results were mostly neutral, with one exception found in the French sample. In France, social chatbot usage was associated with higher reported self-esteem.
“An unexpected finding was the positive association between chatbot use and higher self-esteem in France,” Savolainen said. “It is possible that, for some users, chatbots offer a space that feels supportive or affirming and tools designed to be attentive and nonjudgmental can sometimes create a sense of being ‘heard.'”
“Perhaps French users engage with these tools in a way that encourages such mental boosts. Still, it’s important to acknowledge that chatbots have limitations in truly empathizing or providing sustained emotional support, so these effects should be interpreted with caution.”
The researchers also found that the frequency of face-to-face social contact was not related to chatbot use in any of the countries. This lack of association with in-person contact implies that using AI companions does not necessarily replace human interaction but exists alongside it.
General attitudes toward technology played a role in some regions. A positive view of new technologies was linked to chatbot use in Finland, France, Italy, and Poland. Other demographic factors showed inconsistent patterns. For instance, users in Germany, Ireland, and Poland were less likely to be female.
The consistent link between distress and chatbot usage raises questions about the efficacy of these tools as sources of support. While often intended to help, the data do not support the idea that they successfully alleviate mental health burdens. It is possible that individuals currently struggling are simply more prone to seeking out these tools.
The study highlights that social chatbots might function as “weak ties” in a person’s social network. They can provide conversation and distraction but may lack the depth required to buffer against serious emotional challenges. The user knows the interaction is simulated, which may limit its ability to reduce deep-seated loneliness.
“Across the six countries we studied (Finland, France, Germany, Italy, Ireland, and Poland), we consistently found that younger individuals and those experiencing psychological distress were more likely to use social chatbots,” Savolainen explained. “Loneliness was also a significant factor in several countries.”
“Taken together, these results suggest that social chatbot use may emerge as a response to emotional or social challenges rather than as a tool that inherently improves well-being. This does not mean chatbots can’t be helpful, but it reminds us that the dynamics of use are complex and may reflect underlying needs that technology alone cannot fully address.”
The researchers used robust statistical methods to ensure the reliability of these associations. They controlled for various background variables to isolate the specific relationship between well-being and chatbot use.
However, as with all research, there are some limitations. The study provides a snapshot of associations but cannot prove causality due to its cross-sectional design. It remains unclear whether psychological distress drives people to use chatbots or if the usage itself contributes to distress.
Future investigations could employ longitudinal designs to track changes in well-being over time. This would help determine if chatbots successfully alleviate loneliness or potentially exacerbate it. The authors also suggest that future work should explore the specific motivations behind why individuals turn to these tools.
“We are living a unique time of human–computer interaction, as different technologies are becoming more sophisticated and culturally embedded,” Savolainen said. “Moving forward, we aim to examine social chatbot and user relationships through longitudinal research, which will allow us to trace changes over time and better understand causal dynamics.”
“Social psychological aspects, such as how people form bonds with chatbots, whether they regard them as companions or even “friends,” and how these perceptions influence well-being, is another future direction for us. As chatbots diversify and become more personalized, understanding the evolving nature of these relationships will be crucial.”
“Our study contributes to ongoing conversations about digital well-being and the role of technology in addressing emotional needs,” Savolainen added. “Issues like loneliness, distress, and self-esteem are deeply human challenges that technology often tries to help with -and sometimes succeeds. But our findings highlight the importance of looking critically at these tools, understanding who uses them, and considering how they fit into the bigger picture of human–technology relationships.”
The study, “Individual and well-being factors associated with social chatbot usage: A six-country study,” was authored by Rita Latikka, Jenna Bergdahl, Iina Savolainen, Magdalena Celuch, and Atte Oksanen.
