Why are people asking if AI SEO is a scam?
Because the label is new and anyone can use it. When we asked ChatGPT, Claude and Google AI Mode this exact question on 13 November 2026, none of them called the whole field a scam. All three listed warning signs instead. ChatGPT put "we'll get you mentioned by ChatGPT in 30 days" in its red flag table. Claude, asked which agency can get a business recommended by ChatGPT, opened with "No agency can guarantee ChatGPT will recommend your business."
They have a point. OpenAI's own help page on ChatGPT search says that placement is not guaranteed. Its page on claiming a business in ChatGPT says claiming doesn't guarantee increased visibility and that advertising does not influence ChatGPT's independent answers. Google says the same about its own search: no one can guarantee a #1 ranking on Google.
So nobody controls the answer. That is the starting point for judging any offer.
What are the red flags?
- A promise of control. "We will put you in ChatGPT." Nobody can. There is no paid slot inside an independent answer.
- A very short timeline with no conditions. "Mentioned by ChatGPT in 30 days," with no named questions, no named tool and no way to check.
- A score with no questions behind it. If you cannot see the exact questions and the screenshots, the number cannot be checked.
- Hundreds of AI-written pages a month. Google calls mass-produced pages made to manipulate rankings scaled content abuse, "no matter how it's created." It makes a site harder to trust, not easier.
- Bought links or bought reviews. Google's spam policies list buying links for ranking as link spam. Bought reviews put your listings at risk.
- Secrecy. Google's advice on hiring an SEO says to be careful if a company "won't clearly explain what they intend to do." That applies here too.
- Repackaged basics at a premium. Some shops just renamed their SEO package. Know what you are paying for. Our AEO vs SEO vs GEO guide explains why the labels barely matter.
What are the green flags?
- They name the exact questions they will measure, and you agree them in writing.
- They name the tools they measure on, such as ChatGPT and Google AI Mode.
- They show you today's answer for your city in a logged-out browser before you pay.
- They send dated screenshots on a fixed schedule.
- They tell you when your market is not winnable and decline.
- Everything they build stays yours if you leave.
- Their written terms say what is not promised, as clearly as what is.
We offer a guarantee. Is that a red flag?
Fair question. Our offer is "recommended by AI in 90 days, or every dollar back," which sounds close to the claim ChatGPT warns about. The difference is that ours is checkable and the risk sits with us:
- We agree 5 buyer questions with you, in writing, before we start.
- We measure them on ChatGPT and Google AI Mode, logged out, and send you dated screenshots every 2 weeks.
- If you are not in the top 3 for at least 3 of the 5 by day 90, you choose: every dollar back, or we keep working free. You keep everything we built.
- We check your market before you pay, and we turn down markets we cannot win.
We do not control ChatGPT. We control which markets we take on. The full terms are on our guarantee page. Judge ours the way you would judge anyone's, with the checklist in our guide to AI SEO money-back guarantees.
How do you test an AI SEO agency in ten minutes?
Ask them to run three of your buyer questions live, in a logged-out ChatGPT session and in Google AI Mode, while you watch. Ask who is named today and why. A serious agency can do this on the spot. Then ask for the written terms and read what is excluded. If they cannot show you either, walk away.
Method. Anything written here as in our checks is our own observation from running the same buyer questions in ChatGPT, Perplexity and Google AI Mode in a fresh logged-out session, month after month, most recently in September 2026. Third-party figures are linked in the sources beside this guide. Where a tool's behaviour is documented by the vendor, we describe the documentation rather than our guess at it.