During this year's "Double Eleven", many people discovered when selecting products that more and more merchants are using AI to create product promotion images. The page effects are exquisite and eye-catching, but there is a huge gap after receiving the actual products, and their expectations are often frustrated. AI, which could have improved the quality and efficiency of e-commerce, has now become a new worry for consumers on their online shopping journey. When AI breaks away from reality, a test of consumer trust also ensues.
This week, the China Youth Daily Social Survey Center and the Questionnaire Network (wenjuan.com) conducted a special survey around the "AI Seller Show" and collected a total of 1,334 valid questionnaires. In the survey, more than half of the respondents have definitely stepped on the "pits" of AI seller shows, and more than 30% of the respondents suspected that they have stepped on them and are not sure whether the pictures are generated by AI. 95.8% of the respondents lamented that the AI seller show made them feel "uncertain" when shopping online, adding to the "uncertainty" of placing orders. 67.4% of the respondents called on the platform to upgrade the AI image recognition and review mechanism to intercept misleading AI images in advance.
AI seller shows two major "pitfalls": severe color difference and distortion of pattern details
Wu Zhiyuan, a post-85s generation in Zhejiang, is currently renovating. He took advantage of Double Eleven to buy a lot of household items, but he has also returned a lot of them. He feels that AI is "to blame". The most serious mismatch was a bed: the merchant's AI renderings showed solid materials and high quality, but when you received the actual product, it was completely different. After several "pitfalls", he became more concerned about online shopping: "I spent a lot of time comparing stores, but the results were all AI-generated pictures. It was a waste of energy. I don't dare to do it easily anymore."
Li Qiao, a post-90s generation from Yunnan, recently returned two pieces of clothing. “The styles and fabrics in the AI pictures are very different from the actual items.” Not only that, she also encountered several contradictory AI pictures of the same product. She was interested in a coat. In the detailed picture, some of the collars had buttons, some did not, and some had extra lace. When she asked customer service which picture was accurate, the other party kept being vague: "Dear, the model picture shall prevail" and "maybe the light will cause the picture to appear differently", but he never gave a clear answer, leaving her speechless.
Sun Jiayi, a mother of two who works in Beijing, also suffered from the AI seller’s show. She bought a toy for her children. The material in the product picture was thick and the colors were full, but when she got it, she found that the color difference was huge and the plastic was very thin. She bought a kitchen drain basket. The steel wire was thicker in the promotional picture, but the overall workmanship looked very solid. However, the actual steel wire was very thin and had no texture, so it was not worth the price at all. “I placed the order after reading all the comments in the comments section. When I opened it, I thought the wrong product had been shipped.” Later, she scrolled through the comments section and found that real buyers complained that the seller’s comments and positive reviews were all generated by AI, and she suddenly understood that she had been deceived.
The survey shows that during online shopping, more than half of the respondents (51.4%) have stepped on the "pitfalls" of AI seller shows, and nearly 10% (9.0%) of them complained that they have "stepped on them many times." It is worth noting that more than 30% of the respondents (35.2%) suspected that they had stepped on it, but were "not sure whether the picture was generated by AI", and only more than 10% of the respondents (13.4%) had not stepped on it. In addition, the younger the respondent, the higher the proportion of people who have clearly experienced the "pitfalls" of AI seller shows.
So, what pitfalls are common in current AI seller shows? 58.6% of the respondents complained about serious color difference, inconsistent hue, saturation, etc.; 50.2% of the respondents pointed to the distortion of the layout and details; 45.6% of the respondents believed that there was a supporting AI fake buyer show and positive reviews; 44.2% of the respondents felt that the AI-generated pictures were falsely claiming to be real shots; 36.9% of the respondents found that the material texture was fictitious; 31.6% of the respondents felt that the fictional scenes exaggerated the effect of the product.
"Online shopping has changed from a fun thing to a troublesome thing."
"Nowadays, every shopping app is full of AI pictures, and I can't even pick out one after a long time." Li Qiao lamented, "Online shopping has changed from a fun thing to a troublesome thing. Before buying clothes, I have to carefully look at the model's body, light, shadow, and details to judge whether it is an AI picture. Shopping has become a 'detective scene.'" Now, Li Qiao will skip any product page if he thinks it is an AI picture.
Wu Zhiyuan has also encountered a store with "full-link AI packaging": from product display pictures, copywriting, to posting orders in the comment area, almost all are produced by AI. "There is a store with a sales volume of more than 1,000 for a product and more than 100 reviews. However, after browsing for a long time, I felt that all the reviews had a strong 'AI smell'. I didn't dare to place an order at all, for fear of stepping into a 'trap.' It would be too troublesome to return the product back and forth."
In the survey, 95.8% of the respondents lamented that the AI seller's show made them feel "uncertain" when shopping online and increased the "uncertainty" of placing orders. Over 40% (44.2%) of the respondents said that this feeling has significantly increased. Specifically, those born in the 2000s (98.8%) had the highest proportion, followed by those born in the 1980s (96.4%) and those born in the 1990s (95.6%). This feeling was more obvious among women (97.7%) than among men (92.7%).
Because it is unclear whether the pictures are generated by AI, Sun Jiayi's online shopping return rate is also high. She couldn't figure out that the products would eventually be delivered to consumers. If the promotional pictures did not match the actual products, it would only end up eroding everyone's trust in the store. "Either return the product, or reluctantly keep it because it's too troublesome, but it's just a one-time deal, and it also involves a lot of after-sales work. Is it cost-effective for the merchant to do this?"
Nearly 60% of the respondents expect the introduction of detailed rules for the use of e-commerce AI and clarify the red lines for AI promotion and use
Zhou Ting, a college student born in the 2000s, bought a skirt that had "nothing to do" with the AI seller's show. She asked customer service why the dress she received was "exactly the same" as the one on display, but the other party always insisted that it was a real shot of the product. "In the end, I had no choice but to return it and block the store."
In her view, relying solely on consumers to protect their rights and avoid unscrupulous merchants cannot fundamentally solve the problem. It is still necessary to rely on the platform to strictly regulate the use of AI, upgrade the AI image recognition and detection system, accurately identify AI-generated images and automatically label them. For AI pictures that are deliberately exaggerated and misleading to consumers, platforms should restrict uploads and delete them promptly. "We hope that the platform will set up a dedicated and responsive reporting portal, and that the rectification will not be superficial or painless."
Wu Zhiyuan discovered that some merchants are using more and more AI methods. In addition to seller shows and buyer shows, they even use AI to generate stunning offline store renderings and display them on the details page. "There is nothing you can't think of, but nothing you can't do." He hopes that there will be dedicated e-commerce AI usage specifications to clarify the boundaries of AI use. "The widespread application of AI is the general trend, but businesses must maintain their bottom line and cannot let AI take care of everything. Even taking a few more real-life pictures is a reflection of honest management. Only by knowing how to respect consumers can AI truly empower e-commerce."
In response to the chaotic show of AI sellers, 67.4% of the respondents called on the platform to upgrade the AI image recognition and review mechanism to intercept misleading AI-generated images; 56.8% of the respondents expected to issue detailed rules for the use of e-commerce AI and clarify the red lines for the use of AI propaganda; 52.9% of the respondents Respondents suggested opening a special complaint portal for AI false propaganda to simplify the process of proof and rights protection; 44.9% of respondents emphasized regulating the commercial use of AI by merchants; 44.2% of respondents believed that merchants who deliberately conceal the attributes of AI and use AI to exaggerate their propaganda should increase penalties.
Among the respondents, 60.0% were women and 40.0% were men. Post-00s accounted for 17.3%, post-90s accounted for 42.8%, post-80s accounted for 31.3%, post-70s accounted for 7.0%, post-60s accounted for 1.3%, and others accounted for 0.3%.
(At the request of the interviewees, Sun Jiayi and Li Qiao are pseudonyms)
China Youth Daily·China Youth Daily reporter Du Yuanchun
Source: China Youth Daily
November 12, 2026 Version 08
