Incident guide

AI is telling shoppers my store is a scam.

"Is [your brand] legit?" is one of the most-asked pre-purchase questions on every AI engine. When the answer comes back "be careful, this may be a scam", you lose the sale silently — the shopper never emails you to check. Legitimacy answers are assembled from reviews, complaint boards, and scam-checker sites, which means they can be corrected at the source. Here's how.

Understand where 'scam' answers come from

Engines rarely invent a scam claim from nothing. They synthesize it from what they read: scam-checker sites that auto-score domains, complaint threads, review platforms with a handful of bad reviews, or simply the absence of any trust evidence at all. A new store with no reviews, no about page, and a domain registered last month pattern-matches to 'risky' in these sources. Ask the engine "is [brand] legitimate?" several times and note which sources it cites — that's your fix list.

The correction playbook

  1. 1.Answer the question yourself, prominently

    Publish a page that directly answers "Is [brand] legit?" — who you are, how long you've operated, your address, your policies, real contact channels. Put the direct answer in the first 100 words. Engines prefer quoting a direct answer over inferring one.

  2. 2.Fix the cited scam-checker and review pages

    Scam-checker sites usually have a claim or dispute process; use it. Respond publicly to negative reviews on the platforms the engine cites. A claimed, responded-to profile reads very differently to a crawler than an abandoned one.

  3. 3.Ship machine-readable trust signals

    Organization JSON-LD with your legal name, address, founding date, and sameAs links to your official social profiles gives engines an unambiguous identity record to check claims against.

  4. 4.Build third-party evidence

    Reviews on independent platforms, a Google Business Profile, press mentions, supplier or industry listings — legitimacy answers weigh independent sources far more than anything on your own domain.

  5. 5.Document every wrong answer

    A false 'scam' claim about a functioning business is potentially defamatory. Keep timestamped, tamper-evident captures of each one. Even if you never litigate, the record disciplines your correction campaign and proves the before/after.

What not to waste time on

Don't rely on the answer's report button — it's feedback, not an override. Don't buy fake reviews; review platforms and engines both detect patterns, and getting flagged makes the legitimacy problem worse. And don't expect parametric engines (Gemini, Claude, ChatGPT without browsing) to update quickly no matter what you fix — their answers change at the next retrain. Fix sources for the long term and monitor in the meantime.

Frequently asked questions

Why does AI say my store might be a scam?

Usually because the sources it reads — scam-checker sites, review platforms, complaint boards — score you as risky, or because there's no trust evidence to find at all: no reviews, no about page, a young domain. Ask the engine for its sources; that list is your fix list.

Is a false 'scam' answer defamation?

Possibly — it's a factual claim that damages a business. Six AI defamation suits were filed in the two years to mid-2026, and a German court has held Google responsible for AI Overview statements. But the dismissed Walters v. OpenAI case shows these are hard to win. Keep timestamped evidence of every wrong answer either way.

How fast can I fix a 'scam' answer?

Engines that browse (Perplexity, ChatGPT with browsing) usually reflect source fixes in days to two weeks. Google AI Overviews take weeks. Gemini and Claude answer from training data and can't be corrected directly — expect months, and focus on the sources their next retrain will read.

Can PositionBird remove the wrong answer for me?

No — and be skeptical of anyone who says they can. Nobody controls AI engine outputs. What PositionBird does: continuously checks what engines say about your legitimacy, alerts you when a wrong claim persists past statistical noise, keeps a hash-stamped evidence record, and gives you the per-engine correction playbook.

Find out what AI says about your store — before customers do

Run a free one-shot accuracy check right now, or let PositionBird monitor your business facts weekly across ChatGPT, Perplexity, Gemini, and Claude — with statistical debouncing so you're only alerted to real patterns, and a hash-stamped evidence record of every answer.

Run a free accuracy scanGet the Shopify app ↗

Wondering how common this is? Read the research: how often AI gets small businesses wrong →