Scammers in China Are Using AI-Generated Images to Get Refunds
During the holiday shopping season, many shoppers face the disappointment of receiving damaged goods. Traditionally, online shoppers submit photos to support refund claims, but the rise of generative AI is disrupting this system.
Fake Damage Claims Using AI
On the Chinese social app RedNote, multiple sellers and customer service agents have reported receiving suspicious refund requests backed by AI-generated images and videos. Examples include photos of torn bed sheets with illegible shipping labels and cracked coffee mugs that appear more like shredded paper.
Fraudsters frequently target categories such as fresh groceries, inexpensive beauty items, and fragile products like ceramics. These items are often refunded without requiring returns, making scams easier to execute.
Notable Case of AI Fraud
A crab seller on Douyin encountered a refund claim featuring videos of allegedly dead crabs. However, the videos contained inconsistencies such as unnatural crab limb positions and mismatched numbers of legs, raising suspicions. Police confirmed the videos were fabricated, leading to the temporary detention of the accused buyer. This case marked the first AI-driven refund scam in China to prompt law enforcement intervention.
Global Rise of AI-Driven Refund Scams
This issue is not unique to China. The fraud prevention company Forter reported a 15% increase in AI-manipulated images used in refund claims worldwide since early 2024. Criminal networks use these tactics at scale, flooding systems with fraudulent claims while using IP rotation to evade detection.
Countermeasures and Challenges
Some merchants use AI tools to analyze refund images for signs of manipulation, but these solutions are imperfect. Moreover, e-commerce platforms often continue to favor buyers, which complicates efforts to curb fraud without penalizing honest customers.
As AI-generated content becomes more prevalent, online marketplaces must devise stronger verification methods, revise refund policies, or implement more robust accountability standards to maintain trust and protect legitimate shoppers and sellers alike.
For more details, see the original article on WIRED.
