In 2024, online shopping scams were the most commonly reported fraud type among Americans under 80, according to the FTC. Most of those losses didn't come from obvious fakes. They came from ghost sites: stores that look completely legitimate, show real product photos, accept real credit cards, and then disappear with your money.
The design quality of these sites has gotten much better in recent years. They copy real logos, generate convincing "about us" pages, and sometimes even show fake customer reviews. What they don't have is any inventory, any customer service, or any intention of shipping anything to anyone.
AI can help you check a site in about two minutes before you hand over your card number.
The Red Flags to Look For
Wells Fargo's security team breaks down the most common warning signs in plain language. The short version:
Prices that are dramatically lower than any legitimate retailer offers
A domain name that's slightly misspelled or uses an unusual extension (.shop, .store instead of .com)
No physical address, phone number, or working contact information on the site
No genuine social media presence, or only recently created accounts
Payment only through wire transfer, Zelle, Venmo, or cryptocurrency rather than credit cards
Checkout pages that don't use https in the browser bar
The problem is that savvy ghost sites have learned to fake most of these signals. That's where running the URL through AI before you buy adds a layer of protection that visual inspection alone can't provide.
Before you shop anywhere unfamiliar, our piece on scammer tactics AI helped expose walks through the psychological manipulation playbook these sites use so you recognize the pressure when it's happening.
The Prompt
Copy this into ChatGPT or Claude:
I want to evaluate whether an online shopping website is legitimate or a "ghost site" scam. The website is: [URL]. I found it through: [Google search / Facebook ad / Instagram ad / email / other].
Analyze the common risk factors in plain language: domain age and credibility, pricing that seems unusually low, urgency or pressure tactics, quality of written language, presence or absence of real contact information, return and refund policies, payment methods offered, and overall transparency.
Classify the site as: likely legitimate, questionable, proceed with caution, or high risk (avoid). Include practical next steps I should take before buying, such as safer payment options or ways to verify the seller elsewhere.
AI can catch a lot. It can't guarantee a verdict on a site it hasn't seen, but it knows the common patterns and can identify when something doesn't add up.
Two practical tools to use alongside this: NordVPN blocks known scam sites at the network level before you even land on them. And 1Password ensures that if you accidentally enter credentials on a suspicious site, those credentials don't unlock anything else you own.
This week: before you click any social media ad that takes you to an unfamiliar store, paste the URL into AI and run the prompt. Thirty seconds of checking is worth the time.
WHERE TO GO NEXT
Someone Just Tried to Scam Me. I Think. — A real-world account of what a scam attempt looks like in practice and how to respond when you're not sure.
How to Avoid an Online Shopping Scam — FTC — The FTC's current guidance on avoiding fake stores and fraudulent sellers, updated November 2025.
Report Fraud — FTC — If you've already been targeted or suspect a site is fraudulent, this is where to file a report that actually reaches the people who investigate these cases.
Advertising Disclosure: We evaluate all recommendations of products and services independently. Clicking on links provided on this page may result in AI for Daily Living earning compensation, which supports independent publishers like us.
