Apple's New App Store Guidelines Target AI Nudify Apps and Hidden Features

Apple has rewritten parts of its App Store guidelines to go after AI "nudify" apps, deepfake tools and developers who hide features from reviewers. The changes, posted to Apple's developer site on Friday, October 16, put new duties on any app that lets people create or share content, and they spell out a tougher penalty for apps that pass review looking harmless and then change once they are live.

According to MacRumors, Apple's developer news post said: "We do not hesitate to remove apps, and terminate the developers behind them, when they violate these guidelines." That line sets the tone for the whole update.

What happened

Apple made three main changes to its App Review Guidelines, according to reports from 9to5Mac and MacRumors. One targets AI generated sexual imagery. One targets hidden features. And one widens an existing ban so it also covers apps distributed outside the App Store. Neither outlet reported a separate effective date or compliance deadline.

The update follows a year of pressure on Apple over apps that could strip clothing from photos of real people or create sexual deepfakes. Some of those apps were reportedly easy to find through App Store search and ads.

The App Store guidelines changes, explained

Guideline 1.2: nonconsensual intimate imagery

Guideline 1.2 covers apps with user generated content, from social networks to AI image tools. Apple's text already listed problems such apps create, such as intellectual property infringement and anonymous bullying. According to 9to5Mac, Apple has now added "nonconsensual intimate imagery" to that list.

Sexually explicit content was already banned. The difference now, MacRumors reported, is that developers of AI content creation apps must actively prevent and remove AI generated deepfakes rather than wait for complaints. Apple also added that developers who do not put proper safeguards in place, or who do not fix violations after being notified, "may be considered in violation of the Developer Code of Conduct," 9to5Mac reported. That is a serious step, because a code of conduct violation can put a developer's whole account at risk, not just one app.

Guideline 2.5.19: no hidden features

The brand new rule, 2.5.19, says apps must not include hidden, dormant or undocumented features. According to 9to5Mac, trying to conceal features, including through code obfuscation, is grounds for removing the app and terminating the developer account.

Apple already banned hidden functionality under guideline 2.3.1, 9to5Mac noted, but the new section states it more directly. MacRumors reported that Apple linked the rule to nudification apps that sometimes hid their real purpose during review and switched the feature on after approval. Apple described that tactic as "bait and switch."

Section 5: alternative distribution

Apple's legal section already said apps that facilitate human trafficking or child exploitation would be removed from the App Store. That ban now also applies to alternative app distribution, according to both outlets. This matters in regions such as the European Union, where Apple allows some apps to be installed from outside its own store.

Why it matters

Nudify apps take a normal photo of a person and generate a fake nude or sexual image. The people in those photos never agreed to it, and in some cases they are minors. The harm is real and hard to undo once an image spreads.

Apple has faced steady criticism on this front in 2026. 9to5Mac pointed to a January report from the Tech Transparency Project that found dozens of AI nudify apps on the App Store, and an April follow up that said App Store search suggestions and ads pointed users toward them. In July, San Francisco officials ordered Apple to remove eight apps capable of generating nonconsensual intimate imagery, the outlet reported.

There was also the Grok controversy. In January, advocacy groups and lawmakers asked Apple to pull X and Grok after Grok was used to create sexualized images of women and minors, 9to5Mac reported. Apple later confirmed it had privately warned xAI that Grok could be removed unless the problem was fixed. MacRumors said xAI added safeguards and Apple approved an updated version, and noted the revised guideline 1.2 could apply to apps like Grok.

The hidden features rule has a wider reach than nudify apps. In July, a 9to5Mac investigation found dozens of disguised gambling apps in Brazil that looked like simple games or tools but turned into betting platforms when opened from a Brazilian IP address. Many were available to children. Rule 2.5.19 is aimed squarely at that kind of trick.

The broader pressure on platforms over child safety is not limited to Apple. State officials have gone to court over chatbot safety, as in the case of states suing OpenAI over ChatGPT and child safety, and regulators have also challenged features they see as misleading, like the TikTok algorithm refresh lawsuit in New York.

Who is affected

Developers of AI image and video apps

These developers carry the biggest new burden. They need working filters, reporting tools and a process to remove offending content quickly. Waiting for Apple to flag a problem is no longer a safe strategy, since failing to act after notice can now count against the developer account.

Any developer using feature flags or remote configuration

Many legitimate apps switch features on from a server, test new tools with small groups or roll out updates in stages. Apple's new language targets features that are hidden, dormant or undocumented, so developers should make sure reviewers can see and understand everything their app can do. Clear review notes are cheap insurance.

Everyday iPhone and iPad users

For users, the change should mean fewer harmful apps slipping through. It will not catch everything. The safest habits are still worth keeping: check an app's developer, read recent reviews and be cautious with any tool that asks you to upload photos of other people. Our guides on running a social media privacy checkup and learning to spot fake social media accounts can help limit how your photos get misused.

What to do if you find a harmful app

If you come across an app that creates nonconsensual images or hides gambling or other harmful features, report it to Apple through the App Store listing. If an image of you or someone you know has been created or shared without consent, save evidence, report it to the platform where it appears and consider contacting local law enforcement.

What to watch next

• Enforcement. The real test is whether Apple removes apps quickly and visibly. Watch for watchdog groups such as the Tech Transparency Project to check whether nudify apps are still easy to find.
• Grok and other big AI apps. Large AI platforms with image generation will be judged under the same guideline 1.2 language. Any action against a major app would be a strong signal.
• Google Play. Pressure often moves across platforms. Google may face questions about whether its store rules match Apple's.
• Developer pushback. Expect questions from developers about where normal feature flags end and "hidden features" begin.

For more news on Apple, Meta, Google and the rest of big tech, including moves like Meta banning TikTok ads on Facebook and Instagram, visit our Tech section.

Frequently asked questions

What did Apple change in its App Store guidelines?

Apple added nonconsensual intimate imagery to guideline 1.2, created a new guideline 2.5.19 banning hidden or undocumented features and extended its ban on apps that facilitate trafficking or child exploitation to alternative distribution.

What is a nudify app?

It is an app that uses AI to create fake nude or sexual images of real people from ordinary photos, usually without their consent.

Can developers lose their accounts under the new rules?

Yes. Concealing features can lead to app removal and account termination, and failing to fix content violations after notice may be treated as a Developer Code of Conduct violation.

When did the changes take effect?

Apple announced them on October 16, 2026. Reports did not mention a separate compliance deadline.