“Xiaohai, how come your parents still aren’t here? Everyone else’s family has come to help,” a teacher asks a teenage boy preparing for a student-led product sales competition.
“I’m fine, teacher,” the boy says, sweat streaming down his face. “I can do it alone — I believe the aunties and moms out there will help me.” He then makes an at-camera plea, sweat now turned to tears, asking viewers to help him by buying a luxury face cream and saying, “Are children from poor families really not suited for this stage?”
This scene comes as videos featuring the AI-generated teen male character, “Xiaohai,” surge in popularity, with a related hashtag amassing more than 620 million views on Douyin, the Chinese version of TikTok, as of Tuesday. The videos, marketed to older women, appear across dozens of accounts and push a range of products, from soy sauce to cosmetics.
But in response to the popularity of Xiaohai videos — in which the teen often relays fictional sob stories, likening them to elderly-targeting scams — other user-made clips portraying him as a greedy, ill-mannered troublemaker, under arrest or admitting to fraud, have also begun to emerge, domestic media reported this week.
In the original Xiaohai videos, the character addresses viewers as “aunties and moms,” asks after their health, and occasionally performs a song before marketing various products.
Many older viewers reportedly believe the AI-generated character is real. In one video in which Xiaohai donates blood, the comment section is flooded with statements that appear to reflect people’s belief that he is real, such as “You’re such a good boy,” and “Why are you so pale? Please take good care of yourself.”
Using the character to sell products has proven lucrative, according to domestic media. A report by state media outlet Shandian News found that content creators producing videos featuring Xiaohai earn up to over a half of total sales, and some creators say they earn at least 1,000 to 2,000 yuan ($140 to $280) a day from videos featuring the AI character.
In contrast, the recent wave of Xiaohai videos shifts the narrative, aiming to paint him in a light that proves he’s not worthy of viewers’ affection and to dissuade them from purchasing his products. In some videos, the boy brags about how much money he’s squeezed out of his devoted followers. In others, he smokes, drinks, insults older women, or flaunts his wealth.
In response, some commenters said they were angry they had bought Xiaohai-backed products. Others interpreted the videos as evidence that Xiaohai had accidentally gotten into trouble or been coerced into lying. “Know that your aunties and moms won’t think badly of you,” reads one comment. “This will pass … I’ll think of something; tell me your address and phone number.”
Xiaohai creates a “sense of intimacy” that can make his appeals persuasive to seniors, Yin Xiaoting, associate dean of the School of Humanities and Public Administration at Shanghai Open University, told Sixth Tone. “That sense of familiarity and intimacy can override the need to verify whether the character is real or whether the story is true,” she said.
Helping older adults navigate AI-generated misinformation means encouraging them to question seemingly trustworthy personas and to verify claims independently, even when the story feels emotionally convincing, Yin added. She also stressed the need for both digital literacy programs for seniors and increased regulation of AI-generated content.
In recent years, regulators and state media have increasingly scrutinized the use of fabricated personas and empathy-driven narratives to sell products online. In 2024, a local court in southwestern China’s Sichuan province convicted eight people for selling products online using false narratives. At last year’s Two Sessions meeting, industry experts called for legislation to curb the misuse of AI-generated deepfakes.
Editor: Marianne Gunnarsson.
(Header image: Screenshots of Xiaohai, an AI-generated character. From Weibo)