How to Spot a Fake AI App in an App Store

Most fake AI apps are not broken. They work well enough to seem real, charge a weekly subscription, and send your text to someone else's server. To spot one in an app store, read the listing against the rules the store itself publishes, because a fake has to break several of them at once and a legitimate app usually breaks none.

That framing is more useful than a vibe check. The platforms write down what is not allowed, reviewers are supposed to catch it, and the ones that slip through tend to fail the same published rules.

Check the developer, not the app name

Apple's App Review Guidelines are direct about identity. Guideline 4.1 says: "Don't simply copy the latest popular app on the App Store, or make some minor changes to another app's name or UI and pass it off as your own," and 4.1(c) adds that "You cannot use another developer's icon, brand, or product name in your app's icon or name, without approval from the developer." Guideline 5.2.1 asks developers not to include "misleading, false, or copycat representations, names, or metadata in your app bundle or developer name."

Google Play's misrepresentation policy says the same thing from the other side: "Don't impersonate other apps, brands, or government entities," and apps "must not attempt to mimic functionality or warnings from the operating system or other apps." Its metadata rule is that listings "must precisely reflect the app's functionality."

So the first click is the developer name, not the app name. Three questions:

1. Does the developer name match the company whose product this claims to be? A well-known assistant published by a studio you have never heard of is the whole story.

2. Does the developer's other output look like a portfolio or a factory? A dozen near-identical utilities shipped in a month is the pattern Apple's guideline 4.3 calls out for apps "indistinguishable from what's already widely available".

3. Does the linked developer website exist, resolve, and describe this app? A parked page or a generic template is a bad sign.

Read the subscription terms before the reviews

The commercial model is where a fake pays for itself, and it is also where the published rules are most specific. Apple's guideline 3.1.2(c) requires that "Before asking a customer to subscribe, you should clearly describe what the user will get for the price," and 3.1.2(a) states that apps attempting "to trick users into purchasing a subscription under false pretenses or engage in bait-and-switch and scam practices" will be removed. For recurring payments, guideline 4.9 lists what must be disclosed: the length of the renewal term, that it continues until cancelled, what is provided each period, the actual charges, and how to cancel.

Read the in-app purchase list on the store page itself, which shows real prices rather than the ones in the screenshots. The patterns worth stopping at:

• A three-day trial that converts to a weekly charge, priced so the annual total is far above the thing it imitates

• A single tier called something vague like Pro Access with no description of limits

• Screenshots promising unlimited anything while the purchase list shows credit packs

• A price that differs from what the listing text claims

How to spot a fake AI app when the store listing looks polished

A good fake copies the icon, the screenshots and the description. It struggles to copy history and volume.

Apple's guidelines treat review manipulation as its own offence, warning that attempts to "inflate your chart rankings with paid, incentivized, filtered, or fake feedback" can lead to expulsion from the developer programme. You cannot audit that from outside, but you can read the one-star reviews first, which is where refund complaints and billing surprises collect.

The question that settles most cases

Ask what the app actually is. A large share of AI apps in any store are interfaces to a model someone else operates, which is legitimate when disclosed and a problem when the listing implies otherwise. If the listing claims a proprietary model, look for any technical page that says which model runs underneath. The methods for finding out are in which AI model a tool actually uses.

Then ask where your text goes. An app that sends your prompts to an intermediary adds a party you did not choose, and that party's retention terms are the ones that apply to your data. Checking whether a tool trains on your data covers how to read that quickly.

If you already installed one

Remove the app, then cancel the subscription separately, because deleting an app does not cancel a store subscription. Check the subscription list in your account settings rather than in the app. If you signed in with an email and password you reuse, change it elsewhere. If you pasted anything sensitive, treat it as disclosed to an unknown party.

The broader patterns, including the ones that arrive by message rather than through a store, are in how to spot an AI scam and telling when a chatbot is impersonating a real company. Browser extensions deserve their own pass, since they ask for broader access than apps do, and that checklist is in vetting an AI browser extension. A wider view of what can go wrong for a small team is in the practical guide to AI risks.

FAQ

Does being in the store mean an app was checked?

It means it passed a review process that both platforms describe publicly and that both acknowledge is imperfect, since each maintains removal policies for apps that get through. Review is a filter, not a guarantee.

Is a paid app safer than a free one?

Not reliably. The subscription itself is the business model for many fakes, so paying is sometimes the harm rather than protection from it.

What if the app has thousands of good ratings?

Look at the dates and the text. Ratings bought in bulk cluster in time and read interchangeably. An app with two years of mixed reviews is harder to fake than one with ten thousand ratings in a fortnight.

Are open-source AI apps safer?

They are more auditable, which is not the same as audited. The same checks apply, plus one: confirm the store listing is published by the project rather than by a repackager.