'Efficient Money-Making' Among Strangers... AI-Driven Anti-Chinese Content Spreads in Japan

AI-Generated Anti-Chinese Videos Spread Online in Japan
"Targeting the Elderly for Clicks and Profit"

Online in Japan, generative artificial intelligence (AI) videos that glorify the country and disparage Chinese people are spreading rapidly. There are concerns that hate-driven content, mass-produced through AI in pursuit of high view counts, is fueling anti-Chinese sentiment, especially among the elderly.

Videos related to patriotism spreading on YouTube in Japan. YouTube

On May 8, Asahi Shimbun reported, "Internet video platforms are inundated with videos that praise Japan and criticize foreigners, and in many cases, it is not even indicated whether the content is factual or fictional."

Most of these videos are structured as stories composed of AI-generated images and synthetic voices. They feature provocative titles such as "The world is amazed by Japan's Shinkansen" or "Chinese people who damaged cherry blossoms are experiencing strange symptoms," focusing on glorifying Japan or depicting Chinese people negatively. Some of these videos have garnered hundreds of thousands of views, generating advertising revenue for the creators.

A man in his twenties who previously worked as a subcontractor editing these videos stated, "I had no particular ideology or grudge against foreigners," and added, "I started it with a lighthearted attitude." He began this side job two years ago after seeing a job posting on a recruitment site for "YouTube video editing on overseas reactions criticizing China."

The work involved generating character and scene images within the videos using AI, based on provided scripts. He recalled that, at the time, "it felt like something simple I could do at home."

YouTube logo. Pixabay

The editing process even included a specific 'manual.' According to the newspaper, the instructions stated that, even if the story was fictional, place names and facility names should be set as real locations, and since the main viewers were elderly, the audio should be made as easy to understand as possible.

For example, inputting a phrase like "Chinese person loudly protesting" into the AI would produce a convincing image within minutes. He said that initially he felt uneasy, but thought, "It would be obvious to viewers that this is a work of fiction."

However, about half a year after starting, he began to question who was watching these anti-Chinese videos and how they were being consumed. He searched for the script titles on the video platform himself and found that the videos he had created were being uploaded one after another.

Asahi Shimbun pointed out, "Behind these internet videos that provoke viewers' hatred and sometimes distort perceptions, there exists an 'efficient money-making structure' where people who do not even know each other's faces are connected and mass-produce videos using AI."

This content was produced with the assistance of AI translation services.