The Shanghai Consumer Protection Committee has released the "2026 618 Online Shopping Experience Consumer Survey Report". The survey results from 4,308 valid questionnaires are thought-provoking: under the new trend of AI-assisted consumer decision-making, 84.56% of consumers have tried AI shopping features, but only 16.06% believe that AI can accurately match products, and negative feedback on the user experience dominates.

The survey found that 38.79% of consumers feel that AI completely fails to meet the demand for low-price screening, instead prioritizing high-priced items; another 29.71% said the recommended results mix high and low prices, making manual filtering very time-consuming. Meanwhile, in terms of platform algorithm recommendations, only 24.21% of consumers believe the recommendations are highly matched with their needs, 58.33% frequently receive repeated recommendations of already viewed or added-to-cart items, and 26.69% explicitly stated that the repeated recommendations are a poor experience.

More concerning is the suspicion of algorithmic "price discrimination" - 38.51% of consumers noticed differences in the discount levels of the same product across different accounts, with over 40% stating that this caused shopping anxiety. Regarding discount rules, 21.4% of consumers criticized "the logic is complex and it's impossible to calculate the actual price", while nearly half of consumers felt the rules were complicated and required repeated verification.

The huge gap between high expectations for AI and poor experience

Despite the current poor AI shopping experience, consumer expectations for future AI one-stop shopping are exceptionally high - over 85% of respondents hold a positive attitude towards this, with 38.65% expressing strong anticipation for one-click Agent shopping and 46.87% willing to try it in appropriate scenarios. The report also shows that the 2026 618 shopping festival has moved away from large-scale price wars, with consumption patterns shifting from "people searching for goods" to "AI assisting in selection", and the growth rate of instant retail has reached ten times that of traditional e-commerce.

The Shanghai Consumer Protection Committee clearly pointed out that the new generation of AI shopping models represented by intelligent Agents have broad potential, but platforms must prioritize consumer experience, calibrate algorithm logic, reduce commercial bias, and strengthen demand orientation, truly achieving "algorithm for good". As early as April this year, another survey by the committee showed that over 90% of respondents sensed information bias or algorithmic manipulation in platform recommendations - two months later, the issue remains serious.