AI chatbots pose risks to mental health and cognitive diversity

Research and psychological experts are examining the complex impact of AI chatbots on mental health, social development, and cognitive diversity. A framework proposed by researcher Warren Mansell suggests that AI can either expand a person’s understanding or reinforce unhealthy thought patterns by providing repetitive reassurance rather than deeper reflection.

In Switzerland, the crisis helpline ‘Dargebotene Hand’ has implemented an AI chatbot to provide immediate responses when human counselors are unavailable. While this offers immediate interaction, psychologists warn of the risks regarding the quality of AI responses and the lack of transparency in how large language models function.

A study from the University of North Carolina highlights social implications among youth, finding that many middle school students prefer discussing emotional issues with AI over human friends. Notably, students who frequently use chatbots for emotional support reported higher levels of loneliness.

Furthermore, there are growing concerns regarding cognitive homogenization. As more individuals rely on the same AI tools for brainstorming and decision-making, there is a risk that collective human creativity and diverse ways of thinking may diminish, leading to a convergence of ideas based on the dominant probabilities of AI models.