Decoding Global AI Risk Perception Evolution: Intercultural drivers and Public Discourse Patterns in Algorithmic Societies

Abstract While artificial intelligence (AI) has become a transformative force across societies in the 21st century, its rapid advancement has sparked significant public discourse about potential risks and societal implications. This study investigates global patterns in public risk perceptions of artificial intelligence, examining three critical dimensions: spatial distribution characteristics, intercultural determinants, and temporal evolutionary trends. Through an integrated methodological framework combining natural language processing and convex hull analysis of 4.3 million social media posts from 2010-2023, we identify three dominant risk clusters: algorithmic bias in decision systems, data privacy violations , and labor market disruptions. The findings reveal significant geographic disparities, with privacy concerns increasing 23.7% annually in developed nations since 2018. Five key determinants emerge: national income levels, infrastructure completeness, health security index, regulatory quality, and electoral democracy index. This research offers valuable insights for evidence-based policymaking and contributes to the growing discourse on responsible AI development, ultimately supporting the sustainable integration of AI technologies into society.