Deepfake damage: How AI is scamming China’s e-commerce system

What happened AI-generated damage photos of damaged products are driving a sharp rise in fraudulent refund claims across China’s e-commerce platforms. What began as sporadic manipulation in categories like toys and apparel has escalated into a coordinated gray-market practice now spilling into beauty — one of the most vulnerable sectors due to high order volume and thin merchant margins. Flower Knows, the Hangzhou-born color brand, was among the first to publicly report cases of refund-without-return requests supported by AI-fabricated images. Both consumers and merchants are increasingly weaponizing AI imagery to influence refund decisions, eroding trust breakdown across platforms. The Jing Take: AI turns from efficiency engine to fraud multiplier The fraud wave did not originate in beauty. Toy sellers report AI-generated photos of dolls artificially aged, dirtied, or cracked; apparel merchants face doctored images of exaggerated tears and stains; and homeware shops receive pictures of cups and bottles with synthetic “fractures” indistinguishable from real structural defects. But beauty sellers are now experiencing the most accelerated spillover. The same AI tools that beauty companies once celebrated for boosting efficiency — content creation, virtual try-ons, and diagnostics — are now being used against labels and platforms. In Flower Knows’ case, a consumer submitted AI-generated images of a “shattered” blush compact to demand a refund. When the brand insisted on a return, the consumer physically smashed the intact product before shipping it back. Similar scams are emerging across the category. Examples include faux “broken lash bands,” doctored “leaking foundation pumps,” and artificially scratched palettes. Tutorials promising “30 successful refunds per account” are openly sold online, lowering the barrier to entry for fraudsters and accelerating copycat behaviour. The pattern is economically targeted. The majority of cases hit mid- to low-priced beauty SKUs — typically under 100 RMB ($14) — where margins are thin and the cost of disputing claims often exceeds the order itself. Small merchants are disproportionately exposed. A growing trust deficit for merchants and consumers alike For merchants, especially SMEs, the cost of disputing even a 50 RMB ($7) claim often exceeds the margin on the order. Investigating, appealing, and documenting evidence demands time and technical know-how. Many sellers have shifted to standardized packing videos, traceable packaging IDs, and manual review protocols, pushing operational costs higher. Consumers are not spared. Because fraud is rising across categories, legitimate buyers now face tougher evidentiary requirements: multi-angle videos, timestamps, and in beauty’s case, even dermatology documentation for alleged reactions. The result is a more adversarial after-sales environment in which genuine grievances become harder to resolve. Meanwhile, a troubling countertrend is emerging: some merchants guide customers to upload AI-generated photos specifically to trigger platform safeguards, allowing the system to misclassify negative reviews as malicious. This dual-sided manipulation further corrodes confidence in ratings and post-purchase services. Platforms and regulators confront an AI oversight gap AI promised unprecedented efficiency across China’s e-commerce supply chain. Instead, its misuse is hollowing out the trust that enables high-frequency digital purchasing — from toys to T-shirts to beauty. The sector’s challenge is no longer technological adoption but governance. Restoring confidence requires collective discipline: platforms upgrading detection, regulators tightening legal thresholds, merchants investing in responsible verification, and consumers adhering to good faith. Only then can AI return to its intended role — a productivity engine rather than a fraud multiplier. The Jing Take reports on a piece of the leading news and presents our editorial team’s analysis of the key implications for the luxury industry. In the recurring column, we analyze everything from product drops and mergers to heated debate sprouting on Chinese social media.

The refund hustle: AI-generated lies rock China’s beauty market

Upgrade to Pro

Sectors

Insights

Company

The refund hustle: AI-generated lies rock China’s beauty market