Fake Photos, Real Losses: The AI-Driven Refund Scam Sweeping China’s Online Market

In China’s bustling e-commerce landscape, sellers face a new challenge: shoppers exploiting artificial intelligence to create fake photos showing damaged or defective products as a basis for refund claims. This emerging scam involves AI tools generating hyper-realistic images, often of moldy fruit, cracked ceramics, or torn clothing, which appear convincing but are outright fabrications.​

Many sellers reported cases where buyers submitted AI falsified images to claim refunds without returning the product, effectively obtaining goods for free. AI detection software reveals high probabilities of image manipulation, but the rapid rise of such scams overwhelms sellers who lack resources to verify every claim.​

Popular e-commerce platforms like Taobao and Tmall now incorporate buyer credit ratings and reject automatic “refund only” claims to curb abuse. Additionally, new regulations mandate clear labeling of AI-generated content and introduce penalties for deceptive practices.​

This fraudulent trend undermines sellers’ trust and increases operation costs, fueling price hikes and threatening smaller vendors disproportionately. Experts urge a combined approach involving legal enforcement, technological safeguards, and ethical consumer behaviour to mitigate the escalating issue.​

Key Highlights

Surge in AI-generated fake photos used by shoppers in China to fraudulently claim refunds.​

Common fake damages include moldy fruit, cracked mugs, torn clothes, and rusted items.​

Sellers face significant financial loss and reputational damage due to these scams.​

E-commerce giants Taobao and Tmall have tightened refund policies and buyer rating systems.​

New regulations require AI-generated content to be clearly indicated to consumers.​

Legal penalties exist for fraud offenses, including fines and imprisonment in severe cases.​

Experts call for combined technological, legal, and ethical remedies to restore trust.​

Sources: Hindustan Times, South China Morning Post.​