AI Weakens, But Does Not Strengthen, Political Attitudes

Conversational artificial intelligence (AI) may influence political polarization by shaping the strength with which political attitudes are held. Generative AI systems are now capable of conducting adaptive, human-like conversations, raising new questions about their capacity to strengthen or weaken political belief certainty. We report results from a preregistered, two-wave randomized experiment (N = 1,968) testing whether GPT-4 can strengthen or weaken individuals’ attitudes on salient, evolving political issues. Participants engaged in personalized, multi-turn conversations in which the AI was instructed either to strengthen or weaken participants’ stated views or to have a non-political conversation. Across issues including housing, immigration, and economic policy, AI-mediated conversations consistently reduced attitude strength, specifically, belief confidence, but showed no evidence of strengthening. These effects persisted for up to 21 days. Our findings indicate that generative AI can durably influence political attitudes, primarily by weakening rather than strengthening belief confidence. By examining both pro- and counter-attitudinal persuasion in dynamic political contexts, this study provides a stronger test of AI’s persuasive capacity and clarifies its implications for political polarization and democratic discourse.