Apps that enable AI image abuse face more explicit prevention and removal duties under Apple’s revised App Store rules. The changes address both nonconsensual intimate imagery and developers who hide harmful capabilities until after an app passes review.
Apple announced the revisions on October 16, describing them as a strengthening of existing protections rather than a first-time ban. Guideline 1.2 now expressly includes nonconsensual intimate imagery generated with AI among the material apps must proactively prevent and remove.
That wording puts responsibility on the platform, not solely on the person targeted to complain after an image circulates. Apple says inadequate safeguards, or a failure to remedy violations after notice, can lead to removal from the App Store under its Developer Code of Conduct.
The change is not a blanket prohibition on AI photo editing. It expressly addresses sexualized deepfakes made without consent, while Apple’s other content restrictions still apply. A developer cannot treat an AI-generation feature as exempt from moderation merely because users, rather than staff, create the output.
A separate rule, 2.5.19, addresses hidden, dormant or undocumented functionality. Apple says an app’s purpose must be clear to users and reviewers. It identifies the bait-and-switch tactic of presenting one set of capabilities during review and revealing harmful functions after approval.
The revised legal section also has a narrower distribution consequence. Section 5 says apps facilitating human trafficking or child exploitation will be removed from the App Store and blocked from alternative distribution. That provision should not be presented as a blanket alternative-distribution ban covering every app accused of generating nonconsensual imagery.
The Verge reported the update as a move against so-called nudify apps. Apple’s policy announcement does not identify a new list of apps removed under the revision; a rule change and a completed enforcement action are not the same thing.
Alongside privacy scrutiny and deepfake litigation, the revision makes the developer’s obligations clearer. It does not establish how often review catches concealed capabilities, how quickly abusive images are removed, or how effectively the new wording will be enforced.
