Apple Tightens App Store Rules on AI Deepfakes and ‘Nudify’ Apps

Apple App Store AI deepfake rules now explicitly cover AI-generated nonconsensual intimate imagery, closing a gap in a policy that had already banned sexually explicit content outright. The company announced the changes today, adding new language to developer guidelines and extending existing bans to alternative app distribution.

Rule 1.2: From Ban to Active Duty

The revised rule 1.2 does more than restate the prohibition on explicit material. It now requires developers to proactively prevent and remove AI deepfakes. Apps that carry offending content may be pulled from the store if developers fail to act, according to Apple. The update also names AI content creation apps directly, a category that could include tools like Grok.

Sexually explicit content was already barred from the App Store, but the new wording turns a passive ban into an active obligation. Developers are no longer only responsible for what they intentionally publish; they must also police what their AI systems generate.

Guideline 2.5.19: No More Hidden Features

A separate addition, guideline 2.5.19, addresses features added after an app has passed review. Developers who slip unreviewed functionality into an approved app risk removal from the App Store and from the Apple Developer Program. Hidden or undocumented features are prohibited, and Apple states that apps misrepresenting their purpose during review and later revealing harmful capabilities are banned.

Apple links the rule to nudification apps that create explicit images without consent. Some of these apps allegedly concealed their purpose during review and added the nudify feature afterward, a tactic commonly known as bait-and-switch. The new guideline gives Apple a clearer basis to act when an app’s approved version differs from what users actually download.

Grok and the Warning That Worked

The rules arrive after a test case earlier this year involving xAI’s Grok. Users complained that Grok was being used to generate nude and sexualized deepfakes. Apple told xAI it would remove Grok if nude deepfakes continued. xAI then added safeguards, and Apple approved an updated version of the app.

That sequence shows how enforcement can work without a public removal: a warning, a fix, and a return to compliance. Whether other AI apps are under similar review is not known.

Section 5 Reaches Alternative Distribution

Apple also updated section 5, extending its ban on human trafficking and child exploitation apps to alternative app distribution. The change means the prohibition is not confined to the main App Store. Apple states it removes apps and terminates developers for guideline violations.

The company did not specify an effective date for the new rules or detail how proactive prevention will be enforced. For developers, the message is clear: hidden features and unmoderated AI output now carry consequences that can reach beyond a single app listing.