Misinformation Resilience and Democratic Participation in the Era of Generative AI: A Comparative Social Science Study

Abstract

The study explores how misinformation is perceived and accepted (misinformation resilience) and how citizens participate in democracy in the current context of generative artificial intelligence, especially the impact of AI-generated information on citizens' trust, political engagement, and ability to find misleading information. The rise of the power of generative AI to generate lifelike text, image, audio and video content poses new challenges to democratic societies in terms of electoral integrity, public opinion manipulation and information disorder. With a comparative social science lens, the study examines variations in how much exposure to misinformation someone has, their level of digital literacy, trust in institutions, platform dependence, and civic engagement in selected democratic environments. The findings suggest that societies with robust media literacy, high trust in institutions, and clear platform governance are more resilient to the threat of misinformation spreading via artificial intelligence. The results further indicate that digital literacy interventions for younger, more connected citizens also increase their exposure to generative AI misinformation, but can also improve their ability to verify information. The current study reveals that misinformation resilience is not solely a technological problem but a social, educational, and governance-related problem as well, through comparative patterns. This study emphasises the need for awareness of the public, systems of fact-checking, civic education, platform accountability and policy coordination to safeguard the integrity of democratic participation from manipulation through AI. In conclusion, the paper suggests that a multi-stakeholder approach is needed to build democratic resilience in the era of generative AI, encompassing citizens, governments, education providers, media outlets, and technology platforms.