Does guideline 4.2.6 ban AI app builders?

No. It says apps created from a commercialized template or app generation service will be rejected unless the provider of the app's content submits them directly, and that those services should not submit on their clients' behalf. It is a rule about the submitting account. Ship from your own Apple Developer Program account and you are on the right side of it.

Do I have to tell Apple that AI wrote my app?

There is no field for it and no guideline that asks. What you do have to declare is different. Guideline 2.3.1 asks you to describe new features and product changes with specificity in the Notes for Review and forbids hidden, dormant or undocumented features. Guideline 5.1.2 asks you to disclose personal data going to third parties, including a third-party AI.

Why do AI generated apps get rejected then?

For ordinary reasons. Apple's App Review page says over 40 percent of unresolved issues relate to guideline 2.1, app completeness, which covers crashes, placeholder content and incomplete information. After that comes guideline 4.2, an app too thin to justify being an app. A generated first draft is more likely to be both, which is a quality problem rather than a policy one.

Can I publish twenty generated apps to the App Store?

Usually no. Guideline 4.3(b) rejects apps that are indistinguishable from what is already widely available, singles out categories like flashlight, wallpaper, simple timer and fortune telling apps, and warns that repeated low effort submissions may lead to removal from the Apple Developer Program. One app with content only you have is a different proposition from twenty variations of a template.

Does Google Play accept AI generated apps?

Yes, and Google publishes a policy that is genuinely about AI. Apps that generate content from user prompts have to prevent restricted content and must include in-app reporting or flagging so users can report offensive output without exiting the app. Like Apple's rules, it governs what the app produces rather than what produced the app.

Articles · GuidesUpdated October 2026

Does Apple accept AI generated apps? Yes, and the clause everyone quotes as a ban is about who presses submit.

Yes, Apple accepts AI generated apps, and the App Review Guidelines contain no clause about how an app was written. We read the current version, dated 8 June 2026, today. AI is named once in the entire document, in guideline 5.1.2(i), and that line is about disclosing when your app sends personal data to a third-party AI. There is no app store ai policy that bans generated apps, and nothing in App Store Connect asks what produced the code. Your app is judged on the same list every other app is judged on, which is what Apple asks for.

The clause people quote as a ban is guideline 4.2.6, and it is about who submits, not about which tool wrote the project. Two other rules do reject generated apps and deserve the attention instead: minimum functionality, and the spam rule about apps indistinguishable from what is already on the store. That is also where the answer splits. One finished app for your own idea usually goes through. Twenty near identical ones usually do not.

Check what the rules actually say

The short version

Nothing in the guidelines is about your tool, and two of them are about your app.

So the useful question is not whether AI wrote your code. It is whether the thing you are submitting is finished: real content instead of placeholders, screens that do not crash, a working demo account if there is a login, a support URL and a privacy policy that resolve. Apple's own App Review page puts over 40 percent of unresolved issues under guideline 2.1, app completeness. A generated first draft fails there more often than a hand written one, which is a base rate, not a rule.

The uncomfortable half of the answer: if your plan is to generate many apps quickly and let volume do the work, it is usually no, and the thing at risk is the developer account all of them ship from rather than any single app. That is guideline 4.3(b), and it was written years before AI builders existed.

What the guidelines say about AI, in full

The App Review Guidelines describe a finished app: what it does, what it collects, what it charges for, how it behaves on a phone. The way it was authored is not in scope anywhere in the document. A reviewer sees a binary, your metadata and your review notes. No field asks which editor, framework or model produced the code, and no rule would use the answer if you gave it.

The single mention of AI sits in guideline 5.1.2, on data use and sharing. It says you must clearly disclose where personal data will be shared with third parties, including with third-party AI, and get explicit permission before doing so. That one does land on plenty of generated apps, because an app that pipes whatever a user types into a hosted model is sharing personal data with a third party. It is a disclosure and consent job, not a ban, and it applies the same way to code a person wrote by hand.

Which makes the popular version of this question wrong in a specific way. Apple does not reject apps for being AI generated. Apple rejects apps that crash, that ship placeholder text, that cannot be signed into, that are a website in a shell, or that already exist twenty times over. An ai app rejection is nearly always one of those, and the fix is the fix any developer would make.

Apple, App Review Guidelines, guidelines 1.2, 2.1, 4.2, 4.2.6 and 5.1.2

Guideline 4.2.6 is about who submits, not what you used

Nearly every argument about apple ai app rules ends at this clause, so here is what it does. Guideline 4.2.6 rejects apps created from a commercialized template or app generation service unless the provider of the app's content submits them directly. The next sentence explains the intent: those services should not submit apps on behalf of their clients, and should offer tools that let clients create customized apps of their own rather than copies of one another.

Read the subject of each sentence. The rule governs the submitting party and the account, not the software that generated the project. It was aimed at template mills that pushed hundreds of near identical restaurant and event apps through a single account, and it even offers those providers a legitimate route: one binary that hosts all the client content in an aggregated or picker model. Nothing in it turns on whether a human or a model typed the code.

In practice it means the app ships from your own Apple Developer Program account, in your name, because you are the provider of the content. A builder can generate the project and upload the build; the submission has to be yours. Newly is built that way: it uploads to App Store Connect and TestFlight, and it does not submit for review, so you press that button yourself.

We are not going to tell you that a named competitor breaches 4.2.6, because we have not established that about any of them, and it depends on how each product is configured rather than on how it markets itself. Three things you can check in an afternoon instead: whose developer account the app lands in, whose name appears as the seller on the listing, and whether you can open the build in App Store Connect yourself. If the answer is you all three times, 4.2.6 is not your problem. The order of operations is in publish an ai built app.

The two rules that do reject generated apps

Guideline 4.2, minimum functionality, is the first wall. Apple asks that an app include features, content and UI that elevate it beyond a repackaged website, and 4.2.2 adds that an app should not primarily be marketing material, a web clipping, a content aggregator or a collection of links. Generated apps hit this when the prompt was thin: four screens, a feed pulled from a site, nothing that needs a phone. The fix is product work rather than tooling. Give it one thing a browser tab cannot do, such as the camera, notifications, offline data or a sensor.

Guideline 4.3(b), the spam rule, is the second. It rejects apps indistinguishable from what is already widely available, and names categories Apple will not take new submissions in without a meaningfully different or improved experience: dating, flashlight, sound effects, wallpaper, simple timers and fortune telling. It also warns that repeated submissions of that kind may lead to removal from the Apple Developer Program. Generating an app in an hour makes this much easier to walk into than it used to be.

So the honest answer differs by reader. Building one app for your own business or your own idea, with content only you have? Apple's position on AI is a non-event and your only job is finishing the app properly. Planning a catalogue of generated apps as an income stream? Usually no, and the loss is not one rejection, it is the account. Whether the output is worth submitting at all is a different question, and we took it on in are ai built apps good enough.

Check it

Rule or rumour

Pick something you have read about Apple and AI built apps. The answer is the guideline that actually covers it.

False

No guideline covers how an app was written. The one mention of AI in the whole document is guideline 5.1.2(i), about disclosing personal data shared with a third-party AI.

The store rule that is genuinely about AI lives on Google Play

If you want an explicit AI policy from a store, Google has one. Google Play's AI-Generated Content policy covers apps that create content from user prompts, and names the two obvious shapes: a text to text chatbot where talking to it is a central feature of the app, and images or video generated from text, image or voice prompts. Those apps must comply with the existing developer policies, which includes prohibiting and preventing the generation of restricted content, and they must carry in-app reporting or flagging so a user can report offensive output without leaving the app.

Notice what the policy is about. It regulates what the app generates for its users, not what generated the app. Apple's nearest equivalent is guideline 1.2 on user-generated content, which asks for a way to filter objectionable material, a mechanism to report it, the ability to block abusive users and published contact information. If your app has a prompt box or a chat screen, that list is yours whether you typed the code or described it in a sentence.

Put together, the platform position is consistent and not very dramatic. Neither store asks how the project was written. Both care what it does with a model once real people are typing into it. Add the disclosure guideline 5.1.2 wants, the report and block controls, and a privacy policy that matches what you actually send, and the AI specific part of your submission is finished. The mechanics of the submission itself, the screenshots, the age rating and the review notes, are getting it through and a walkthrough of their own.

Google Play Console Help, AI-Generated Content policy

What actually decides whether Apple accepts your generated app

What you are submitting	The guideline that decides it	Whose account submits	Likely first outcome	What to change first
One finished app, your own content, a reason to be on a phone	2.1, completeness	Yours	Accepted like any other app	Nothing to do with AI
A generated app that is mostly your website in a shell	4.2 and 4.2.2	Yours	Rejected	Add something a browser cannot do
Another timer, wallpaper or fortune telling app	4.3(b), already widely available	Yours	Rejected, and repeats risk the account	Build a different app
A chat or image generator sitting on a model API	1.2 and 5.1.2(i)	Yours	Held for the missing controls	Reporting, blocking, data disclosure
A build the tool submits from its own account	4.2.6	The tool's	Rejected	Move it to your own account

Building an app that passes on its own merits

Everything above points one way. The outcome is decided by what you ship, so the leverage is in finishing rather than in disclosing. Real copy, no dead links, a demo account in the review notes if there is a login, and one capability that explains why this is an app instead of a page. None of that has anything to do with how the code was produced, which is the whole answer to the question in the title.

Newly is an AI app builder. You describe the app in plain English and it writes a real React Native and TypeScript project you own, runs it on cloud iOS and Android simulators while it builds, then ships it to TestFlight and to Google Play internal testing. It is $25 a month and there is no free plan. New apps start local-first, keeping their data on the phone with no accounts and no server, which keeps the privacy section of your submission short. The agent adds a hosted backend, with sign-in, an API service, a Postgres database per environment and file storage, only when the app needs accounts, shared data, syncing or payments. That is the point where the data questions get longer.

Ownership matters here for a practical reason. A rejection is a fix and a resubmission, so you want the code in your hands: it comes out as a ZIP from project Settings, through the two-way GitHub sync under Deploy, or with the CLI. And the submission itself stays yours, which is the part guideline 4.2.6 actually cares about.

Questions people ask about Apple and AI generated apps

Yes. The App Review Guidelines have no clause about how an app was written, and nothing in the submission asks. AI is named once in the document, in guideline 5.1.2(i), which is about disclosing personal data shared with a third-party AI. Your app is judged on what it does, the same as every other app on the store.

Describe the app, then finish it before you submit

Write down the one thing your app does that a web page cannot, the content that will be real on day one, and the login you will hand to review as a demo account. Then build it, and press submit yourself.

Start building