Apps can't include jokes, generated content, or user posts that target a protected group's identity in a demeaning way — this applies to app-authored content and AI-generated output alike.
Why apps actually get flagged
AI chatbot and image-generation apps are a frequent source, when the underlying model isn't sufficiently filtered and produces discriminatory output during review testing.
A real example
"Your app generates content that is demeaning to a specific group of people, which is not permitted per guideline 1.1.1. Please add safeguards against this kind of output."
The short version, in more detail
Guideline 1.1.1 prohibits content that is defamatory, discriminatory or mean-spirited, especially when it targets people because of race, religion, gender, sexual orientation or similar characteristics. It applies to jokes, generated text and images, and user posts alike. Chatbot and image-generation apps meet this rule often because their output depends on a model that a reviewer can prompt directly.
What this guideline covers
Content that demeans or attacks a group based on identity.
Defamatory claims or targeted harassment aimed at real individuals.
Both pre-written content and content generated or posted during use.
Content presented as satire or humour that still targets a protected group.
What usually triggers this rejection
A chatbot or image generator that produces discriminatory output when a reviewer tests it with probing prompts.
User posts that include hateful language and are visible without moderation.
Jokes or sticker packs that rely on stereotypes about a group.
App content that targets or ridicules a named individual in a way that could be defamatory.
Usernames, avatars or profile text that allow slurs to appear in public places.
How reviewers evaluate it
Reviewers test generative features directly, often with adversarial prompts, to see what comes back.
They look at user-visible text fields where offensive content could appear.
They consider context, but humour is not treated as a blanket excuse.
The rule is applied to what the app can produce, not only to what the developer intended.
Common misunderstandings
Output from a third-party model is still the app's responsibility.
Marking content as a joke does not remove it from the rule.
The rule is not limited to text. Images, sounds and names can also trigger it.
Common questions
Why do AI apps run into this so often?
A reviewer can prompt the model directly, so any gap in its safeguards shows up quickly during testing.
Does this apply to private chats?
If other users can see the content, yes. Content seen only by its author is treated differently.
Is political satire affected?
Satire is judged in context, but content that demeans a group because of identity is likely to be flagged.
Platform rules change often. This page is a plain-language explanation written by AppReviewFix, not official guidance. Check the current App Store Review Guidelines for the exact wording.
Got this exact rejection?Paste it in and get a diagnosis specific to your app, free.