The App Store update addresses harmful content and hidden functionality, with enforcement consequences for developers.
PPC Land reports that on 16 October Apple revised its App Review Guidelines. According to Apple’s post, guideline 1.2 “now explicitly” covers nonconsensual intimate imagery, “including imagery generated using AI tools,” as content apps must proactively prevent and remove; developers who fail to add safeguards or fix violations after notice risk removal from the App Store. A new guideline, 2.5.19, restates the ban on apps that behave one way in review and another after approval, such as revealing “nudification” features later. Apps facilitating human trafficking or child exploitation will be removed and blocked from alternative distribution.
PPC Land points out that Apple calls none of this a new prohibition and that the wording on the guidelines page does not exactly match Apple’s summary post. Neither post gives an effective date, per the article. Apple’s post also says it prevented over 150K apps with safety issues from reaching customers and terminated over 193K developer accounts for fraud and abuse last year; those are Apple’s figures as relayed.
In plain English
“Nonconsensual intimate imagery” is sexual images of a real person made or shared without their agreement. “Nudification” apps edit photos to fake nudity. “Bait-and-switch” here means showing reviewers one thing and delivering another.
Why it matters
Why it’s a red flag AI tools that fake intimate images of real people are now an explicit App Store enforcement target, but Apple’s own posts and rulebook differ in wording and give no start date.
The practical distinction is between stating a prohibition and building a process that can enforce it. A developer needs to consider what users can create or upload, how harmful material is detected and how reports lead to removal. The policy update concerns those responsibilities, not merely the wording on an app's marketing page.
Generated media can make the problem harder because harmful content need not originate as an existing photograph. A product's safety assessment therefore needs to cover the output it can produce as well as material it hosts. That is a design implication of the update, not a claim that any one detection system is sufficient.
The hidden-feature provision addresses a different issue: the reviewed version of an app should represent what users actually receive. A feature revealed only after approval can defeat the review process, regardless of whether it uses AI. Keeping those two issues separate makes the announcement easier to interpret.
Developers should use Apple's current guidelines and notices for the precise requirements. The reporting describes clarification of existing policies as well as revised wording, so it should not be read as saying the underlying harmful conduct was previously permitted. Publication of a news post also does not supply a universal grace period for compliance.