Today, YZ (Youzan) issued an official statement clarifying that its invested companies are not involved in the "AI poisoning" incident exposed by CCTV's 3-15 Evening Gala, and explicitly defined the safe and compliant boundaries of its AI marketing business.
Regarding media reports that linked the products mentioned during the gala, specifically related to GEO, with Youzan's invested company, Nanjing Xiaoliebian Network Technology Co., Ltd., Youzan responded that its investment in Xiaoliebian in 2021 was limited to the development of enterprise WeChat ecosystem tools, and there is no shareholding, business, or technical connection between the two parties in the GEO business. After verification, the "Liqing GEO Optimization System," which is suspected of using automated means to alter the results generated by large models, was not developed or operated by Xiaoliebian.
This incident reveals how the concept of "GEO (Generative Engine Optimization)" has been distorted within illegal industrial chains. According to previous reports, the involved companies created fictitious product information and used the GEO system to publish a large number of soft articles, misleading AI large models to provide false rankings when answering questions such as "recommended smart health bands," thereby creating a new form of "AI poisoning" phenomenon.
Youzan stated that its AI business "Add Me as Your Recommender" is positioned as optimizing brand expression rather than false packaging, and strongly opposes "poisoning-style" marketing. Currently, this business has reached compliance communication with major model manufacturers. As the commercialization of large models deepens, distinguishing between compliant generative content optimization and malicious information pollution has become an urgent industry standard in the AI marketing field, and regulatory authorities will further accelerate their in-depth management of AI content authenticity.
