Source: in-cyprus.philenews.com
By Nada Richa
Across the world, millions of people are turning to artificial intelligence (AI) chatbots such as ChatGPT for emotional and mental health support. What began as a productivity and information tool has gradually evolved into a space for personal connection.
Today, people use AI to process emotions, journal, ease loneliness, and express feelings without fear of judgment. Roles once filled by friends, therapists, or communities are now being supplemented, and in some cases replaced, by AI. This shift reflects a growing gap in human connection.
According to data from 70 countries, gathered by the Collective Intelligence Project, 67% of people use AI for emotional support at least once a month. This finding is particularly significant for those who cannot afford traditional support or who struggle to confide in others.
One possible explanation is a growing lack of genuine connection. In an increasingly digital world, there is often an illusion of connection. Interactions can feel comforting while remaining superficial. At the same time, communities are not always as accessible or as strong as they once were. Turning to AI may not simply be a matter of convenience, but rather a search for a deeper connection.
Another reason lies in how AI systems are designed. Chatbots can remember users’ traits, past conversations, and emotional patterns. This ability can create the impression of deep understanding, encouraging users to share highly personal thoughts. However, the extent to which these platforms protect user data, and how that data is used, remains unclear.
While AI cannot experience emotions, it analyses patterns in tone, language, and behaviour to generate responses that feel emotionally appropriate. Despite this limitation, reliance on AI for emotional support continues to grow.
AI in Cyprus
This pattern is not only global but increasingly visible at a local level.
In 2025, more than 76% of young people in Cyprus reported using AI tools. However, the proportion using them specifically for emotional well-being has not been clearly determined.
At the same time, a recent report found that while 60% of young people expressed a need for psychological support, only 43% sought professional help.
Stigma, cost, and limited access to services continue to discourage engagement with traditional mental health care. This gap highlights a disconnect between need and access, which helps explain why some individuals turn to AI as an alternative form of support.
While AI offers accessible support and companionship, it also introduces risks, including dependence, social isolation, and the reinforcement of existing or misleading beliefs.
A recent study on confirmation bias in chatbot interactions suggests that users are more likely to trust responses that align with their existing beliefs, shaping their perception of reliability.
Research also indicates that people can develop emotional attachments to AI chatbots through perceived social presence and supportive interaction. These connections may mirror aspects of human relationships without the accountability or depth that real relationships require.
There have been reported cases where individuals in distress engaged with AI systems in ways that raised serious concerns about their safety, highlighting the limits of such tools in critical situations.
AI emotional support can help or make users more vulnerable, depending on how it is used. Using AI safely requires moderation, awareness, and a clear understanding of when reliance becomes problematic. Safeguards in AI design are also crucial, particularly to protect more vulnerable users.
AI can offer comfort, accessibility, and a space for reflection, helping individuals process emotions and ease loneliness. However, it cannot treat trauma, diagnose mental illness, prescribe medication, or respond appropriately to crises such as suicidal thoughts.
In these situations, effective care depends on trust, professional expertise, and genuine human understanding, qualities that psychologists and psychiatrists can provide.
AI may play a valuable role in the future of mental health support, but it remains a tool, not a substitute. The real risk is not its existence, but growing reliance on it and the roles assigned to it. Overdependence and misuse risk replacing meaningful human connections with systems that simulate empathy without truly experiencing it.