A Better Newspaper

## Overview Chinese-language videos suspected to be AI-generated deepfakes, featuring different narrators berating Singapore for its treatment of China and alleged alignment with the United States, circulated on Chinese platforms including Douyin and WeChat (SCMP, April 2026). Digital experts assessed the clips were most likely AI-generated, citing clues including identical scripts across videos featuring ostensibly different individuals in different settings. ## Content and Platform Distribution The videos accused Singapore of being 'ungrateful' toward China. They circulated on major Chinese social platforms, suggesting either organic amplification or coordinated distribution within China's information ecosystem. The cross-platform spread and identical scripting despite surface variation are consistent with AI-generated influence operation tradecraft. ## Technical Indicators According to digital experts cited by SCMP (April 2026): - Different narrators with different apparent settings used **identical scripts** - Visual and audio characteristics consistent with AI deepfake generation - Distribution pattern suggesting coordinated seeding across Chinese platforms ## Strategic Context Singapore occupies a delicate position in US-China tensions — it maintains strong trade and security relationships with both powers while serving as a regional financial hub. Influence operations targeting Singapore may aim to: - Pressure Singapore toward more explicitly pro-China policy positions - Undermine US-Singapore security cooperation - Exploit domestic Chinese-language media consumption patterns in Singapore's majority-Chinese population ## Implications - **New frontier in AI disinformation**: The use of AI to generate superficially diverse but substantively identical influence content at scale represents a qualitative evolution in state-adjacent information operations. - **Platform liability**: Douyin and WeChat's role as distribution vectors raises questions about content moderation obligations for Chinese platforms operating internationally. - **Attribution ambiguity**: 'Most likely' AI-generated and suspected to be pro-China stops short of state attribution, a recurring challenge in disinformation investigations. ## Related Trends - Iran AI Propaganda – Lego-Style Viral Video Campaign (existing page) - AI Swarm Influence Operations – Election Interference (existing page)