Whose Voice Counts? Synthetic-Voice Labour and Attributional Displacement on Douyin

A smooth synthetic voice can make a long act of augmentative and alternative communication (AAC) sound immediate. This article examines how that audibility is produced and how audiences decide whether machine-mediated words, emotions and competence belong to the nonspeaking creator. Drawing on a six-month netnographic study conducted between September 2025 and February 2026, the analysis combines 14 adult creators, 487 public videos, 412 hours of platform observation, 423 purposively retained voice- and authorship-related comments/replies, and 21 communication-accessible interviews totalling 1,087 minutes. Rather than claiming a universal sequence, the findings identify a recursive configuration: creators compose through heterogeneous AAC and voice arrangements, format expression for platform genres, encounter audience attribution, and sometimes undertake verification labour. The article develops attributional displacement—the movement of presumed authorship, competence, or affect away from the creator toward a device, communication partner, AI system, or commercial intermediary when external involvement is interpreted as replacing rather than supporting the creator’s role. This reframes platform accessibility as a problem of circulation, attribution, tempo and governance, not interface operation alone.