Apple Expands App Store Rules to Block AI Deepfakes

Apple has expanded its App Store moderation rules to cover AI-generated nonconsensual intimate imagery.

Developers must now proactively remove deepfakes. Sexually explicit content was already banned, but the updated 1.2 rule requires app developers to prevent and remove AI-generated deepfakes from their platforms. Apps that fail to act risk removal from the App Store.

The rule names AI content creation apps and could affect services such as xAI's Grok, MacRumors notes. Users have reported Grok being used to generate nude deepfakes, and earlier this year Apple warned xAI that the app would be removed if it continued to permit such content. xAI later added safeguards and received approval for an updated version.

Guideline 2.5.19 addresses hidden app features. Developers who add features or functionality that were not in the version submitted for review risk removal from the App Store and the Apple Developer Program. Apple ties the rule to nudification apps that create sexually explicit images of real people without consent, suggesting some have hidden their purpose during review and added the feature after approval.

Apps containing hidden or undocumented features are prohibited. The guideline explicitly bans bait-and-switch tactics where developers misrepresent an app's purpose during review and later reveal harmful capabilities.

Section 5 now covers alternative app distribution. Apple extended its existing ban on apps that facilitate human trafficking or child exploitation to cover alternative app distribution channels.

Apple states its enforcement position. The company's developer news post reads: "We do not hesitate to remove apps, and terminate the developers behind them, when they violate these guidelines." The full updated guidelines are available on Apple's website.