Buying second-hand on Facebook Marketplace, eBay, Craigslist or Vinted has become second nature, but it has also become a playground for scammers. A tempting listing almost always rests on its photos: they are what build trust, and they are the first thing fraudsters manipulate. Learning to read an image before you pay is today the single most profitable anti-scam skill you can have.
Why photos are the weak point of any listing
On a second-hand platform you don't have the product in hand, nor the seller in front of you. Everything rests on a few visuals and a block of text. The human brain places disproportionate trust in images: a single "realistic" photo is enough to disarm caution. Fraudsters know this and pour their effort into the visual credibility of the listing rather than the product, which often doesn't exist at all.
There are three broad families of visual deception:
- Stolen photos: the seller lifts images from a legitimate listing (another platform, a retailer's site, a real Vinted account) and reposts them under a fake identity.
- AI-generated visuals: an object, an interior, or even an "owner" that is entirely synthetic, created to give substance to a ghost listing.
- Deceptive edits: hiding a defect, adding a missing accessory, altering a serial number, or replacing a background to mask the true origin.
In every case the goal is the same: trigger a payment or a deposit before you've had a chance to verify anything. The whole scam is a race against your scrutiny — which is exactly why slowing down is your strongest defense.
The economics of why this works
Fraud scales the way any business does. A single stolen photo set can be cloned across dozens of throwaway accounts in an afternoon; an AI generator can spin up a fresh, plausible-looking inventory on demand. The marginal cost of one more fake listing is essentially zero, while each victim represents real money. That asymmetry is what fuels the volume you now see on every platform. Understanding it reframes the problem: you are not facing a lone opportunist but an industrialized, repeatable process — and the photos are the assembly line's most reusable part.
The most common scams on marketplaces
The ghost listing with stolen photos
This is the classic. An iPhone, a console, an e-bike priced well below market. The photos are crisp, professional, sometimes too professional. The seller pushes you off-platform ("let's continue over text / WhatsApp"), invents urgency ("I'm leaving the country tomorrow"), and asks for payment by link, money order or bank transfer. Once the money is sent, the account vanishes.
The fake profile propped up by AI
On Vinted and similar apps, some accounts use AI-generated wardrobe or model photos to simulate a well-stocked closet and a "real" seller. AI can now produce dozens of coherent visuals in minutes, which makes these fake profiles far more convincing than before.
Fake payment screens and "delivery" phishing
A formidable variant: the buyer (or seller) receives a fake email or text imitating the platform, the courier or the bank, with a fraudulent payment link. The visual mirrors the official branding. If you receive this kind of message, first learn how to verify a photo or attachment received by message or email before you click.
The product exists... but not the one in the photos
Some scams are subtler: the item is real but damaged, counterfeit or different. The photos have been retouched (scratches erased, a screen "repaired" in post) or borrowed from the brand-new model sold in stores. This is the hardest family to catch, because there is a genuine object behind the listing — the deception lives in the gap between what you see and what you'll receive.
The overpayment and "wrong amount" scam
A reversal of the usual pattern: here you are the seller. A buyer offers to pay more than your asking price, sends a forged payment confirmation (a doctored screenshot of a transfer or a fake platform email), then asks you to refund the "overpayment" or ship before the funds clear. The screenshot looks perfect because it was generated to look perfect. The rule is simple: trust the money in your account, never an image of money.
Table: red flags in a listing's photos
| Red flag | What it may hide | Reflex to adopt |
|---|---|---|
| Very "pro" photos, studio background, perfect light | Visuals stolen from a retailer | Reverse image search |
| No real in-situ photo (at the seller's place) | Ghost listing | Ask for a specific live photo |
| Same object with inconsistent angles (shadows, reflections) | Composite or AI | Forensic image analysis |
| Blurred details: hands, text, logo, serial number | AI-generated image | Check detail coherence |
| Price far below market | Classic bait | Maximum caution, don't rush |
| Missing or inconsistent metadata | Recovered / edited image | Check EXIF |
| Pressure to pay off-platform | Bypassing protections | Refuse and report |
How to verify photos before buying
1. Reverse image search
First reflex, free and fast. You upload the listing photo into a reverse image search engine to see where it appears elsewhere on the web. If the same shot turns up on a retailer's site, an older unrelated listing, or an unrelated account, that's a clear red flag. To master this technique, read our complete guide to reverse image search and its limits.
A warning: reverse search only detects already-indexed images. A stolen photo that's been lightly cropped or recompressed can slip through. That's why it must be combined with other checks.
2. Reading the metadata (EXIF)
A photo taken with a smartphone often carries metadata: device model, date, sometimes geolocation. A "today's" listing with a photo dated two years ago, or a complete absence of metadata (a sign of a screenshot or an edited file), should put you on guard. EXIF data is easy to manipulate, so its absence isn't proof, but an inconsistency is telling.
3. Detecting AI-generated images
This has become essential. Image generators produce stunning objects and scenes, but they leave traces: textures that are too smooth, impossible reflections, unreadable text on labels, malformed hands or fingers, backgrounds that "melt." To know where to look, read our dedicated article: how to detect an AI-generated image.
The human eye quickly hits its limits. A multi-layer forensic tool like TruthLens simultaneously analyzes sensor statistical signatures, compression artifacts, and the markers typical of generative models to estimate the probability that an image is synthetic. You can test a suspicious listing photo for free by dropping it on the image analysis page.
4. Forensic analysis of edits
When you suspect a composite (a defect erased, an object added), error-level and noise analysis reveals zones that were recompressed differently from the rest of the image. A patch that was pasted in, or a scratch painted out, carries a different compression history than its surroundings — invisible to the eye, but legible to the right tool. This is exactly the kind of manipulation professional methods can isolate, and which we detail on the e-commerce side in authenticating product visuals.
5. Ask for a live, dated proof
The cheapest and most decisive test costs nothing and defeats most scams at once: ask the seller for a fresh photo of the exact item next to a handwritten note with today's date, or from a specific angle you choose on the spot. Stolen photos can't comply; an AI generator can't improvise a coherent new shot of the same object on demand; and a borrowed retailer image has no "owner" to hold it. A seller who genuinely has the item produces this in two minutes. Hesitation, excuses, or a counter-offer of urgency in response to this simple request is, in itself, the answer.
The checklist before you pay
Before sending any money, run the listing through the wringer:
- Reverse search on 2 or 3 of the listing's photos.
- Request fresh proof: a photo of the item with a hand-dated note, or a specific angle impossible to improvise.
- AI check: is the scene plausible? Do the details hold up to close inspection?
- Profile coherence: account age, reviews, history.
- Payment method: stay inside the platform's protected system, never a direct transfer or external link.
- Refuse urgency: any pressure for immediate payment is a signal in itself.
Secure payment reflexes
The golden rule: never leave the official payment tunnel. Platforms offer secure payments with buyer protection (funds released only on satisfactory receipt). An honest seller has no reason to ask for a bank transfer, gift-card payment, cash money order, or a link sent by message. These methods are irreversible and untraceable: that's precisely why fraudsters demand them.
On the delivery side, beware of emails and texts imitating the courier or platform. Never enter your banking details via a link you received: open the official app yourself.
How the risk differs by platform
The mechanics of fraud adapt to each platform's design, and knowing the local flavor sharpens your guard.
On generalist classifieds (Facebook Marketplace, Craigslist, Leboncoin and the like), the dominant scam is the high-value ghost listing: phones, consoles, e-bikes, even apartment rentals advertised below market, with photos lifted wholesale and immediate pressure to pay off-platform. The buyer protection is weakest precisely because so many deals are arranged to happen "in person" or by direct transfer.
On fashion resale apps (Vinted, Depop, Poshmark), the volume game dominates. Fake closets propped up by AI-generated garment shots, counterfeit luxury goods photographed against the real brand's official images, and "I'll pay your shipping if you message me directly" lures are the staples. Because the items are lower-value, buyers let their guard down — which is exactly what the scammer counts on.
On auction and global marketplaces (eBay and similar), the subtler "real-but-not-this-one" fraud thrives: a listing for a pristine collectible illustrated with the manufacturer's press photos, where the actual item is damaged or counterfeit. Whatever the platform, the constant is the same: the photo sets the expectation, and the fraud lives in the distance between that expectation and reality.
What to do if you're a victim or you have a doubt
If you've already paid and doubt sets in, act fast: contact your bank to attempt a chargeback, report the account to the platform, file a complaint, and keep all evidence (screenshots of the exchanges, the listing, the photos). A reverse search and a forensic analysis of the photos can strengthen your case by showing the visuals were stolen or faked.
For high-value purchases (premium phone, watch, vehicle), a documented analysis report adds real weight with your bank or the authorities. A bank's fraud team or a magistrate is far more responsive to a timestamped report stating "these visuals show statistical markers of AI generation" or "this photo appears on a retailer's site dated eighteen months earlier" than to a verbal account of suspicion. TruthLens lets you generate a certified report from a listing's visuals, useful for substantiating fraud — and you can run a first check for free on the image analysis page the moment a doubt arises, ideally before you pay.
The order of operations matters: the strongest position is to verify before money changes hands, when walking away costs you nothing. Once a payment is made, you shift from prevention to recovery, where outcomes depend on the payment method, the platform's policies, and how quickly you act.
FAQ
Is reverse image search enough to detect a scam?
No. It's the first filter and reveals already-indexed stolen photos, but it's blind to cropped, lightly edited or AI-generated images. Always combine it with a check of detail coherence and, if in doubt, a forensic analysis.
How do I recognize an AI-generated listing photo?
Look for fine inconsistencies: unreadable text on labels or screens, malformed hands and fingers, reflections and shadows that don't match, textures that are too smooth, melting backgrounds. Our AI image detection guide details each signal, and a tool like TruthLens automates this analysis.
The seller refuses to send an extra photo, is that serious?
It's a strong red flag. A good-faith seller has no problem proving they own the item (a photo with a dated note, a specific angle, the serial number). A refusal, especially coupled with urgency or a request to pay off-platform, should make you walk away.
Is a bank transfer safe for a peer-to-peer purchase?
No. A transfer is irreversible and carries no buyer protection. Always prefer the integrated payment offered by the platform, which releases funds only after satisfactory receipt of the parcel. Systematically refuse external payment links and untraceable methods.