Unlock this publication
Purchase this publication to read all pages online.
Access sales are currently unavailable. Please contact CGPR.
Chapter 2 in Democracy in the era of Artificial Intelligence: Opportunities and Risks
Abstract
Cross-cultural communication has been a fundamental aspect of democratic participation and intercultural understanding for a long time. The landscape is changing drastically as humans engage with one another in new ways across cultures thanks to the emergence of Artificial Intelligence (AI) tools, such as neural machine translation, large language models, and recommendation algorithms. AI systems, however, that are mostly trained on English language and global north data can embody structural language biases that marginalize minority and low-resource language communities. In this study, a qualitative systematic literature review of twenty peer-reviewed article sources found from 2015 to 2025 is used to investigate the enabling and threatening aspects of cross-cultural communication with the aid of AI. The same technologies that help to bring about multilingual engagement also threaten cultural homogenization and epistemic erasure. The promotion of multilingual engagement is hypothesized to be the same technologies that are helping to democratize multilingual engagement which also help to risk cultural homogenization and epistemic erasure. The results identified 4 thematic dimensions: AI as a communicative enabler, linguistic bias, cultural homogenization and ethical governance. An inclusive development is proposed to follow a culturally responsive AI five-layer model.