DefinePK

Abstract

This study interrogates how algorithmic systems mediate fairness when Chinese and Western digital contents circulate across cultural borders. Using 5.9 million TikTok, YouTube, Facebook and Bilibili posts (2021-2023), we quantify three tensions: visibility (English #chinatravel clips attain 4.7× raw and 100× adjusted impressions vs. Chinese #中国旅行); values (8.2 % of Chinese political commentary on YouTube is demonetized vs. 1.9 % English equivalents; 11.4 % of Western LGBTQ+ videos are suppressed inside mainland China vs. 2.3 % globally); and voice (cross-border appeals succeed 27 % less often, driven by linguistic and contextual blind spots). Borrowing from Chinese smart-court audit trails, we propose culture-aware procedural logs, layered cultural translations for model explanations, and rotating creator councils with veto power pilots already cut false takedowns by 19 %. Fairness, we argue, is not a universal equation but a negotiated equilibrium between algorithmic logics and culturally embedded notions of justice.