Audience Responses to Cultural and Linguistic Gaps in English–Arabic Auto-Subtitles on YouTube

Abstract

This study investigates audience reception of English–Arabic auto-generated subtitles on YouTube, focusing on linguistic accuracy, cultural representation, trust, and viewing experience. A large-scale survey of 4,500 participants across diverse age groups, genders, and linguistic backgrounds was conducted to examine demographic and behavioral factors shaping subtitle use. Descriptive results revealed that YouTube is firmly embedded in daily routines, with 72.4% of respondents integrating it into everyday media habits. Subtitles were widely relied upon, with nearly 60% routinely enabling them, and motivations included clarifying unclear audio (68.2%), navigating accents (61.9%), and supporting language learning (54.3%). However, mistranslations and cultural misrepresentations were highly salient, with over 70% noticing linguistic errors and 66.5% reporting missed cultural references, both of which significantly reduced enjoyment and comprehension. Trust in auto-generated subtitles was conditional: fewer than half expressed blanket trust, while 73% preferred human-translated subtitles, and trust varied strongly by content type, being lower for educational and formal material. Inferential analyses showed that proficiency in English reduced sensitivity to linguistic errors, while proficiency in Arabic heightened awareness of cultural mistranslations. Frequent viewers and those with strong media habits reported greater reliance on subtitles, while motivations such as accessibility and language learning predicted consistent use. The findings highlight both the indispensability and limitations of auto-subtitles, emphasizing their role as accessibility tools while exposing persistent deficiencies in linguistic accuracy and cultural mediation. These insights have implications for AI translation design, user trust, and media accessibility in multilingual digital environments.