As anti-China sentiment intensified in Japan, related videos were even mass-produced through artificial intelligence (AI).
Japan’s Asahi Shimbun reported that online videos praising and criticizing Japan and foreigners are rapidly spreading, especially ‘hateful’ videos. Most of these videos contain AI-generated images and machine voices.
A man in his 20s, who subcontracted and produced AI anti-China videos, said, “I had no grudge against foreigners at all, and I started with a light heart.”
“I worked to create images in the video with AI according to the prepared script,” he said. “Even if the story was created, the name of the facility and the place name had to be set as existing. Since most of the viewers are elderly, there was also an instruction that the voice should be made easy to understand.”
He added, “When I received a lot of money, I earned 50,000 yen a month.”
He said, “I often felt doubts during the work, but at that time, I thought the viewer would know that it was a creation.”
Then, he found a video he made on the Internet, and later realized that Internet users did not doubt the content at all. A said, “I knew this could have distorted someone’s thoughts and created social division,” adding, “I feel like I did the wrong thing as a human being.”
The Asahi Shimbun also interviewed a man in his 60s who actually ran the channel.
The man, a former civil servant, entered the YouTube market after looking for a new job after retirement. After learning that anti-Chinese content is in vogue, B began producing similar videos.
“I hated Chinese people, too. What I wanted to do and the demands of viewers were the same,” he said, explaining why he started his career. “It’s strange that I’ve been allowed to do it for years and many others are doing it now,” he said.
Shinichi Yamaguchi, a professor at International University, pointed out, “The more negative emotions are stimulated, the more anti-Chinese content is produced.” He said, “If you stimulate antipathy toward a specific country or group, it is easy to attract people’s attention, and producers earn profits,” and he was concerned about viewers’ distortion of their perception of reality.
There are no explicit discriminatory expressions among the contents, but there are also those that indirectly promote hatred and prejudice. Professor Yamaguchi said, “It is also a question of whether platform companies can respond appropriately to content that narrowly avoided policy violations,” and argued that faster regulation or algorithm improvement is needed.
JENNIFER KIM
US ASIA JOURNAL
