Apple has tightened its App Store rules around AI-generated content and apps that conceal harmful capabilities during review, according to an October 19 TechCrunch report. The changes target practices including applications that appear innocuous when submitted, then expose features that generate nonconsensual intimate imagery after approval.

AI creation apps face explicit moderation duties

Appleā€™s published App Review Guidelines now explicitly include ā€œAI content creation appsā€ in section 1.2, alongside social media and other user-generated content platforms. The section identifies nonconsensual intimate imagery, intellectual property infringement, and anonymous bullying as risks.

Covered apps must filter objectionable material, provide a reporting mechanism with timely responses, let users block abusive accounts, and publish contact information. Developers must remove violating content and address identified problems. Apple says egregious or repeated behavior can lead to immediate removal from both the App Store and its developer program.

Review must reflect the actual product

Section 2.5.19 separately prohibits hidden, dormant, or undocumented features. It says functionality must be clear to users and App Review, and identifies attempts to conceal features through code obfuscation or other methods as grounds for removal.

For developers shipping AI image tools, the practical implications extend beyond the modelā€™s output filters: reporting, abuse handling, and accurate disclosure of features are also distribution requirements. Appleā€™s submission guidance requires live backend services and reviewer access to app functionality.

The policy language establishes obligations and removal grounds. It does not, by itself, demonstrate that every offending app has been detected or removed.