Addressing “Political Anger” And “Non-Political Fear” In News Engagement: The Impact of Human-AI Interactions on Emotional and Behavioral Responses

The rise of AI technologies has transformed news consumption, yet little is known about the mechanism of how human-AI interactions influence audiences’ emotional reactions and civic behaviors. This study employed a 2 (China vs. U.S. AI chatbot) × 2 (cognitive vs. emotional language style) × 2 (political vs. nonpolitical news) factorial experimental design to address these questions. The results indicate that interacting with AI chatbots can mitigate discrete negative emotions related to the issues and enhance civic actions. AI was particularly effective in reducing anger, specifically in political news, and fear, specifically in nonpolitical news. Cognitive language style elicited higher trust than emotional language style, while country of origin had no significant effect.