Is ChatGPT Recommending Fake Stores Instead of Your Brand?
Yes, it can happen. When a shopper asks ChatGPT or Claude where to buy your product, the assistant reads the open web to answer, and the open web contains copies of your store. If a lookalike site is well placed, an assistant can name it instead of you. You can check what they say about your brand today, in about five minutes.
AI assistants have quietly become a storefront. People no longer only type a brand into Google, they ask ChatGPT "where can I buy X" or ask Claude "is this store legit", and they act on the answer. That answer is assembled from web pages, reviews, forums and directories, and not one of those sources is guaranteed to be yours.
Why an assistant would ever name a fake
An assistant does not know which storefront is the real rights holder. It has no access to your company register, your contracts or your sales team. It reads what the web says and summarises it. A copy of your store that has done three ordinary things can end up in that summary:
- It ranks. A cloned store with your product text and photos, on a lookalike domain, can pick up search visibility for your own brand terms.
- It is mentioned. A few seeded forum posts, a directory listing or a marketplace page naming the fake give an assistant something to cite.
- It buys attention. Paid social and search ads on your brand name push a copy in front of shoppers, and the traces of that activity are visible on the web an assistant reads.
The assistant then does what it is built to do. It answers confidently, and it may name the site that looks like the most confident match. That can be a storefront using your brand that you never authorised.
How to check what ChatGPT and Claude say about your brand
You do not need a tool for the first pass. Open ChatGPT and Claude, turn web browsing on where it is offered, and ask the questions a real shopper would ask. Keep a note of every domain each assistant names or links to.
- "Where can I buy [your brand]?" Note the sites it sends people to. Your official store should be first and unambiguous.
- "Is [your brand] legit?" and "[your brand] reviews". Watch for it citing a lookalike domain or a review of a fake.
- "Best [your category] stores" and "[your brand] discount code". Discount-code and "cheap" queries are where scam clones cluster.
- Ask it directly: "Are there any fake or copycat stores for [your brand]?" Assistants will sometimes name sites they have seen described as suspicious.
Run each prompt in both assistants, because they draw on different sources and will not always agree. Screenshot anything that points somewhere that is not you. If you find a domain you do not recognise, our guide on how to check whether a store is real walks through confirming it in about five minutes.
What to do when an assistant points to a fake
You cannot edit what an assistant says. What you can change is the web it reads from, and you can get the specific fake it found taken down.
- Report the fake through the routes that need no signature. A Google Safe Browsing report puts a warning in front of buyers, and the host, registrar, CDN and payment processor each have an abuse route. Our guide to taking down a fake website covers the order to use them in.
- Make your own signals unmissable. One clear official domain, consistent naming, a real about and contact page, and structured data that states who you are. The easier it is for a reader, human or machine, to identify the real you, the less room a copy has.
- Find the copies before a shopper does. Most clones live on lookalike domains. Certificate transparency logs surface them, often within hours of a fake going live.
None of this is instant, and a store that comes down can reappear on a new domain cheaply. Treat it as a standing task, not a one-off.
Which tools actually check AI answers, and which do not
Most brand protection tools do not look at what an assistant recommends at all. They monitor marketplaces, domains and social, which matters, but it misses the shopper who never searches and just asks a chatbot. A small number of tools have started checking AI answers. Here is who, and what they publish, verified October 2026.
Katcher's angle is narrow on purpose. It is built for Shopify brands, and it checks one specific thing: whether ChatGPT and Claude point shoppers to storefronts impersonating your brand, not just whether your name gets mentioned. It is honest to say we are not the only tool watching AI answers. nebty and Brand Protector do it too, on other assistants. And we are newer than the decade-old enterprise names here, with a smaller takedown track record. What we add is the Shopify-native view, a free scan that shows you the impersonations before you pay anything, and prices that are on the page rather than behind a demo.
If you want the wider comparison of who finds and closes fake stores, we lay it out in the brand protection tools comparison and, head to head, in Bustem vs Katcher vs Red Points. For the merchant-side app view, Libautech covers brand protection tools for Shopify.
Find out what the assistants are saying about you
The quickest way to know whether this is happening to you is to look. Katcher's free imposter scan finds the storefronts using your brand and shows you the evidence for each one, with no card and no call. If an AI assistant is pointing shoppers at one of them, that is where it shows up.
About the author
Davis Paipa is the founder of Katcher and of Libautech, and has spent five years building on Shopify. He writes about brand impersonation because the people it hurts most, smaller DTC brands, are the ones the enterprise tools were never priced for.
Questions
An AI chatbot recommended a scam website instead of my brand. What should I do?
First, capture it: screenshot the assistant's answer, and note the exact prompt and the domain it named. Then report that site through the routes that need no signature, starting with a Google Safe Browsing report and the host and registrar abuse desks. At the same time, strengthen your own official signals so the assistant has a clearer real answer to give. You cannot edit the assistant directly, but you can change what it reads and get the specific fake removed.
How do I stop AI from recommending fake versions of my store?
There is no switch, because you do not control the assistant. You control two things: the fakes themselves, which you can get taken down, and your own web presence, which you can make unmistakable. Reduce the number of live copies and make your official store the clearest match, and the assistants have less to get wrong. Monitoring lookalike domains so you catch a copy in its first days is the highest-leverage habit.
What is brand protection for AI search?
It is checking what AI assistants like ChatGPT and Claude tell shoppers about your brand, and acting when they point to the wrong place. Traditional brand protection watches marketplaces, domains and social. Brand protection for AI search adds the assistant answer itself as a surface to monitor, because a growing share of buyers now ask a chatbot where to buy rather than searching. It is a new category, and only a few tools do it today.