App stores rate companion apps mostly on self-declared answers

What happened

As retrieved on 16 September 2026, Apple's App Store Review Guidelines and Google's Play Console content-rating help page set out two different systems for how companion and chatbot apps get their age labels. Neither document mentions companion apps by name; both describe general rules that apply to any app offering chat-based or user-generated content. This is a living-document comparison, not a record of a single dated event.

What the documents show

Apple's guideline 4.7.1 requires apps that embed chatbots or similar non-binary software to include a filtering mechanism for objectionable material, a way to report content, and the ability to block abusive users, and guideline 1.2 separately requires published contact information and timely responses to complaints for any app with user-generated content. Apple's age-rating guideline, 2.3.6, places the burden on the developer to 'answer the age rating questions in App Store Connect honestly.' Google's help page describes a different process: developers complete an International Age Rating Coalition questionnaire, and independent bodies such as ESRB, PEGI or GRAC calculate a rating from the answers given, each using its own methodology; misrepresenting the questionnaire, the page states, 'may result in its removal or suspension.'

The mechanism

Both systems rest on the same underlying mechanism: self-certification. Apple's chatbot-specific rules require certain safety features to exist, but do not describe an independent test of whether a companion character actually stays within a stated age rating during a real conversation; Google's IARC questionnaire computes a rating purely from a developer's own answers, with no mention in the retrieved page of chatbot behaviour, romantic role-play or AI-generated content as a distinct category at all. In both cases, the platform verifies that required disclosures and mechanisms exist, not that a chatbot's actual output matches its declared rating.

What it leaves open

Neither document specifies a distinct review process for generative AI or companion personas as opposed to static app content, meaning an app's effective age-appropriateness depends heavily on how accurately, and how conservatively, its developer answered the relevant questions. Editorially, this is the same self-declaration gap outside evidence from Common Sense Media's testing has separately described as bypassable; these store policies do not resolve that gap on their own terms.

• Does an app's store rating reflect an independent test of its chatbot's output, or only a developer's declared answers?
• Does the store guideline require ongoing monitoring after launch, or only a one-time filing?
• If a chatbot can be steered into content beyond its stated rating, does that count as a policy violation under either store's stated rules?

Read side by side, the two frameworks show two large platforms placing the substantive judgment about a companion app's suitability for minors on the developer's own disclosure, with the platform acting mainly to check that a disclosure was made at all.