Over the last few months, a new type of scam has quietly started spreading across second-hand marketplaces like Vinted: Buyers using generative AI to fake damage on items to get refunds.

Yes… really.

Grey Nike ‘Just Do It’ T-shirt laid on a wooden floor with two holes near the neckline, showing visible damage to the fabric.

Here’s the new playbook scammers are using to scam sellers on Vinted:

  1. Order an item
  2. Receive it in perfect condition
  3. Take a picture
  4. Use AI (e.g. ChatGPT) to add “damage” (burn holes, stains, tears…) to the photos
  5. Submit a refund claim on Vinted
  6. Keep the item for free

Buyers get the product and their money back. Sellers lose their items and never get paid!

Some attempts are hilariously bad (like the one in the screenshot below), but others are scarily subtle. With the right tools, AI can generate realistic damage that easily passes as legitimate evidence.

Screenshot collage from a Facebook post in the ‘Idiots of Vinted Official’ group showing an AI-generated damage scam on Vinted. The seller explains a buyer used AI to fake burn holes on a Nike hoodie to claim a refund. The collage includes the fake damaged photo, the original undamaged hoodie, Vinted support messages, and angry chat messages confronting the scammer.

AI is Breaking Online Marketplaces like Vinted

  • Writing compelling listings faster
  • Enhancing photos
  • Automating admin

But when buyers weaponise AI to commit fraud, the entire trust system of online reselling platforms collapses.

And marketplaces face a brutal choice.

The Future of Reselling Platforms

With the rise of AI scams reselling platforms have two options:

  1. Either they fail to stop AI-driven fraud, in which case sellers lose trust. They stop listing. The marketplace liquidity dries up and buyers look elsewhere for goods to buy.

Everyone loses. Or…

  1. Marketplaces tighten verification!

This means, platforms start requiring:

  • Video evidence for damages (harder to fake, for now!)
  • Pre-shipment documentation such as pictures of the packaged good to verify the conditions pre-shipping,
  • Tighter profile verification to keep fraudsters off the platform

Implementing such checks and measures will undoubtedly help maintain trust on the marketplace, but so will it:

  • Increase operational costs for the platform
  • Add frictions for both sellers and buyers
  • Likely increase selling or buying fees on the platform (the cost always gets passed to the customer…)

Here again, user experience gets worse. And online platforms start looking less appealing compared to… other ways of shopping.

Shopping Second-Hand is Much Better In-Person

In fact, building trust has always been significantly easier in real life. When you visit a local second-hand shop, you:

  • See items yourself
  • Try them on, touch the fabric, and inspect them

None of that can be faked. On top of that, visiting charity and second-hand shops is a social activity where you can:

  • Talk to shop owners (if you want)
  • Thrift with friends - instead of scrolling on the sofa
  • Support real local businesses

That’s the beauty of in-person second-hand shopping - and it’s exactly what we’re accelerating at Ganddee by connecting shoppers to local shops and helping them see what items are in stock.

The more AI disrupts trust online, the more people will rediscover the joy (and reliability) of shopping second-hand offline.

Download Ganddee (free) to start exploring your local second-hand shopping scene!

If you’re a second-hand or charity shop willing to get listed fill this form with your shop details - it takes under 30 seconds and it’s free!