Videos of a boy appealing to viewers about his hardship that spread across Chinese short-form platforms turned out to feature a virtual character created by artificial intelligence. The videos drew a combined 600 million views, and the accounts behind them exploited viewers' sympathy to sell health supplements and other products.
Videos featuring a boy called Xiaohai have spread one after another across major Chinese short-form platforms in recent weeks, according to China Central Television and other local media on the 24th. The faces and voices differed from video to video, but the storyline was the same — that he lived in poverty with a single mother and could not eat properly. The accounts used the videos to draw attention, then sold goods for profit.
The boy approached middle-aged and older female viewers by calling them "auntie" and "mom." One recurring line had him hiding in a classroom and making a meal out of a steamed bun. A considerable number of viewers believed the story and bought products, local media reported.
In May, a 79-year-old woman who could not read boarded a long-distance train without telling her family, saying she wanted to meet Xiaohai, and was found by police. She said only that she was going to visit a distant relative and could not name her destination. She returned home after police convinced her that the boy did not exist.
The market for AI-generated video is expanding rapidly in China. China's AI short-drama market reached 22 billion yuan in the first half of this year, according to research firm DataEye. Kuaishou said views of AI videos as of July were up 321% from six months earlier.
Rules are already in place. Since September 2025, China has required AI-generated content to be labeled as such. The Cyberspace Administration of China said in February, ahead of the Lunar New Year holiday, that it had penalized 13,421 accounts for spreading false or misleading information without AI labels and removed more than 543,000 posts.
In late April, the State Administration for Market Regulation launched a six-month crackdown on internet advertising, including the misuse of AI, and on May 3 the CAC said it had penalized 98,000 self-run media accounts for providing information without disclosing its source.
As the controversy grew, local platforms moved to penalize the accounts involved. Some operators changed their account names, deleted videos or stopped selling products, according to local media. Even with labeling requirements and enforcement campaigns in effect, however, videos fronted by an AI-generated character racked up 600 million cumulative views.