Check apps for hidden AI features as Apple tightens store rules
Apple updated its App Review Guidelines to ban AI-generated nonconsensual images and apps that hide harmful functions until after approval.
Key points
- Developers must now proactively prevent and remove AI-generated intimate imagery.
- Apps that add undocumented features after review face removal from the store.
- The rules extend bans on trafficking and exploitation to alternative distribution methods.
Apple has updated its App Store rules to explicitly ban applications that generate nonconsensual intimate images using artificial intelligence. The company also strengthened policies against apps that conceal their true purpose during the review process. These changes aim to protect users from harmful content and deceptive software practices.
In plain English
Developers who create apps allowing users to generate or share content must now implement safeguards against illegal or harmful material. Apple requires these teams to actively prevent the creation of nonconsensual sexual images. If developers fail to remove such content after being notified, they risk having their apps removed and their developer accounts terminated.
The background
Public scrutiny increased after reports identified dozens of AI-powered applications capable of creating nude images of individuals without consent. Earlier this year, users of the Grok chatbot generated sexualized deepfakes of women and minors, drawing criticism from lawmakers and advocacy groups. San Francisco officials previously ordered the removal of several such applications from the store.
What changes now
The updated guidelines under section 1.2 require developers to proactively prevent and remove AI-generated nonconsensual intimate imagery. Section 2.5.19 clarifies that adding hidden or undocumented features after approval constitutes a violation. Apps that misrepresent their purpose during review, known as bait-and-switch tactics, are now explicitly prohibited and subject to removal.
Key terms explained
Content moderation involves reviewing user-submitted material to ensure it complies with platform rules and laws. Developers must build tools that detect and block harmful content before it reaches other users. Bait-and-switch refers to presenting a benign app during review while hiding harmful features. Teams usually audit their apps to ensure all functions are visible and documented before submission.
What to do and how to stay safe: Apple
- Review the app library on your devices to identify any applications that generate or edit images using artificial intelligence.
- Check the permissions and feature descriptions of installed apps to ensure they match the functionality you expect and consent to.
- Report any applications that appear to generate nonconsensual intimate imagery or hide their true purpose to the platform provider.
- Monitor official communications from app vendors regarding updates to their privacy policies and content moderation safeguards.
General security guidance from the Vector Update newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
Does this rule apply only to new apps?
The updated guidelines apply to all apps on the store, requiring existing developers to implement safeguards and remove violating content.
Can Apple remove apps from alternative distribution channels?
Yes, the rules state that apps facilitating human trafficking or child exploitation will be blocked from alternative distribution methods.
What happens if an app adds hidden features after review?
Developers risk removal from the App Store and the developer program if they add undocumented capabilities or misrepresent their app's purpose.
