AI-generated videos exploit elderly and cause public panic in China
AI-generated videos of romantic figures were deployed on short-video platforms to manipulate elderly Chinese users into purchasing products, while a separate AI-generated video of a building collapse sparked public panic before authorities confirmed the incident never occurred and penalized the creator.
What happened
Video marketing accounts on short-video platforms deployed AI-generated videos of romantic figures called 'AI tyrants' (AIιΈζ») targeting elderly users with sweet talk and declarations of love, designed to manipulate them into purchasing products from affiliated shops and live-streaming storefronts.
An AI-generated video falsely depicting a building collapse with multiple workers falling circulated on social media, accumulating 19,000 likes and over 5,000 comments and shares before police investigations determined the incident never occurred. The individual responsible, He moumou (δ½ζζ), used mobile AI tools to generate the fake video and added geographic location tags to boost visibility.
Reported impact
Elderly citizens in Jiangxi province and elsewhere were misled into believing AI-generated characters were real people, leading them to purchase low-cost trinkets (priced at 39.9 to 99 yuan) and leave comments engaging with the fake personas.
The building collapse video caused widespread public alarm and panic, with viewers sharing unverified information further spreading the false report before police confirmed no such accident had occurred in Hejiang County.
Index analysis
Marketing accounts were designed with identical scripts targeting the emotional and companionship needs of elderly users, deploying AI video generation technology to create romance scenarios. Online tutorials taught operators how to build followers in the 'silver-hair demographic' and monetize them through product sales.
The AI-generated building collapse video bore an AI generation indicator in its lower left corner, but law enforcement noted that ordinary users lack the ability to reliably identify AI-synthesized content, enabling the false information to spread rapidly.
Governance controls
The Luzhou Hejiang police network security division conducted investigations and confirmed no building collapse accident had been reported in the jurisdiction, verified through dispatch records. He moumou was administratively penalized and required to delete the video from all platforms.
Police issued public warnings that AI technology should serve society rather than enable rumor-spreading and fraud, and called for citizens to improve their ability to identify false information.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.