In the e-commerce industry, the use of AI to create fake product images to defraud refunds is becoming increasingly rampant, which not only harms merchants' interests but also disrupts the platform's ecosystem. According to AIbase, Taobao and Tmall announced at a recent merchant service conference that it will officially launch an AI image recognition model for after-sales fraud, as one of the key measures to optimize the business environment in 2026.

image.png

This new tool means that merchants now have strong technical support when dealing with after-sales disputes. Merchants can now use the Wangwang chat window to perform one-click detection on images that are suspected to have been modified by PS or added with fake defects (such as holes, stains). Once the system identifies the image as an AI-generated fake, the identification result will be directly accepted by the platform for subsequent dispute resolution, refund review, and appeal processes, effectively curbing the "black and gray industry" of malicious after-sales services.

In addition to technological governance, Taobao and Tmall also revealed a series of impressive data. According to AIbase, its account integrity system has cumulatively intercepted abnormal refund behaviors, helping merchants recover losses exceeding 4 billion yuan; meanwhile, through optimizing the logistics compensation mechanism, the platform is expected to save merchants over 1 billion yuan in costs throughout the year. In the future, the platform's traffic will be further tilted towards high-quality service stores, creating a more fair and healthy e-commerce business environment through continuous optimization of algorithms and rules.