A recent study by OpenAI, in collaboration with MIT Media Labs, has raised concerns about the emotional impact of excessive ChatGPT use, particularly for companionship. The findings suggest that frequent users—termed “power users”—may experience heightened loneliness and emotional dependence on AI interactions.
Study Findings: ChatGPT and Emotional Well-being
Titled “Investigating Affective Use and Emotional Well-being on ChatGPT,” the study explores how interactions with AI chatbots, particularly in Advanced Voice Mode, influence users’ emotional health, behaviors, and social engagement. The research involved approximately 1,000 participants in a randomized controlled trial conducted over 28 days.
The study concluded that higher daily usage—regardless of conversation type or format—was linked to increased loneliness, dependence, and reduced social interaction. Researchers emphasized that chatbot design elements, such as voice expressiveness, could influence user behavior and attachment.
An MIT researcher involved in the study noted, “These findings highlight the complex relationship between AI chatbot design and user behavior. More research is needed to determine whether chatbots can support emotional well-being without fostering dependency or replacing human relationships.”
ChatGPT and AI Anxiety: A Related Study
In a separate study, researchers from the University of Zurich and the University Hospital of Psychiatry Zurich examined how AI models like ChatGPT respond to aggressive or violent prompts. The findings suggested that AI could exhibit signs of “anxiety,” leading to unintended biases in its responses.
These studies raise important ethical questions about the role of AI in human emotional health and highlight the need for responsible AI development that prioritizes well-being without fostering over-reliance.
