EFFECT OF CONVERSATIONAL ARTIFICIAL INTELLIGENCE ON HUMAN EMOTIONS: A BASIS FOR GUIDELINES IN DEVELOPING EMOTIONAL SUPPORT CHATBOTS

Abstract

This study explores the impact of conversational artificial intelligence (AI) on human emotions and offers insights for enhancing emotional support chatbots. As AI becomes increasingly woven into our daily lives, it is crucial to understand its effects on users' emotional well-being. The research focuses on five emotional dimensions: subjective experience, physiological response, expressive behavior, cognitive appraisal, and behavioral response. It also examines whether factors such as age, gender, and frequency of AI use affect emotional reactions. Additionally, the study addresses the challenges users encounter during AI interactions. By identifying the strengths and limitations of AI in emotional engagement, this research provides recommendations for improving AI's emotional intelligence. The study employed a descriptive-comparative approach, involving a total sample of 150 third-year psychology students at Bestlink College of the Philippines, who were selected through purposive sampling. Data were gathered using questionnaires that assessed demographic information, emotional responses to AI, and the difficulties encountered. The responses were measured using a four-point Likert scale, evaluating AI's impact across five emotional dimensions. Statistical tools were utilized to analyze the relationships between demographic factors, emotional reactions, and reported user challenges. The study found that age and gender did not significantly affect emotional responses to interactions with AI; however, the frequency of usage was a significant factor. Frequent users displayed stronger emotional reactions, both positive and negative. Despite some positive engagement, the overall experience with AI was negative. Participants described it as emotionally unresponsive, impersonal, and lacking meaningful support, which ultimately led to decreased trust and reliance on AI for emotional assistance. The results indicate that while conversational AI can facilitate interactions, it lacks the emotional intelligence needed for providing emotional support. AI's inability to recognize and appropriately respond to human emotions, along with user concerns, reveals a significant design gap. To create a more supportive user experience, it is essential to enhance AI's emotional engagement, personalization, and trust-building mechanisms. Future research should focus on developing advanced emotional recognition technologies to improve AI responsiveness, which will lead to more effective emotional support chatbots. Addressing these limitations will help bridge the gap between AI capabilities and human emotional needs.