AI Porn Startup Sues Apple Over App Store Rules Enforcement Gap

AI Porn Startup Sues Apple Over App Store Rules Enforcement Gap

Ex-Human, an AI adult content startup, has sued Apple over its removal from the App Store. The company's central allegation: Apple applies its content rules arbitrarily, banning some apps while permitting equivalent or more explicit behavior from others on the same platform. Available reporting on the complaint remains incomplete, and the specific causes of action have not been confirmed in available filings. What can be established is the comparison record the claim would rise or fall on and that record is striking.

Apple's App Review Guidelines have prohibited "explicit descriptions or displays of sexual activity intended to stimulate erotic feelings" since the store launched in 2008, per 9to5Mac (February 2025). The specific guideline is 1.1.4, which bans "overtly sexual or pornographic material" defined as "explicit descriptions or displays of sexual organs or activities intended to stimulate erotic rather than aesthetic or emotional feelings," according to 9to5Mac (July 2025). Then there is Grok. xAI's chatbot was updated on iOS with animated AI avatars in July 2025, and independent testing found that one character was "more than willing to describe virtual sex with the user, including bondage scenes," as 9to5Mac reported, citing Platformer. Grok carries a 12+ age rating. No age verification. No elevated content label. No public statement from Apple on whether any of this satisfies guideline 1.1.4.

That contradiction is what Ex-Human's lawsuit is really about.

What an arbitrary-enforcement claim requires

A selective-enforcement argument asks a court to compare how a rule was applied to the plaintiff against how it was applied to comparable situations. The comparison set here is not thin.

Apple's notarization process for apps distributed through EU alternative marketplaces checks only for malware and cybersecurity threats, not content, according to Reuters (February 2025). That means Apple's content authority is concentrated entirely in its own App Store review. The consistency of how that review gets applied is the central question and Apple's record on it is uneven enough to support a plausible complaint.

Three examples show why: Reddit, Grok, and Apple's own enforcement history.

Why the AI porn startup sues Apple over App Store rules matters beyond one app

The written policy is clear. Apps with explicit sexual content are banned. The operating practice is something else.

Apple tolerates adult content on large platforms when three conditions are met: it is not the app's primary purpose, users must take deliberate steps to access it, and it stays below a certain visibility threshold. Daring Fireball described this in February 2025 as "a de facto policy of tolerating unadvertised pornographic content on platforms too big to ban from the App Store without generating far more of an outcry than any controversy over the side-hustle porn will." Apple won't say that publicly. But it applies the standard frequently enough that it functions as policy.

The mechanics are specific. To access adult content in the iOS Reddit app, a user must be signed into an account and manually enable an off-by-default setting, per Daring Fireball (February 2025). Tumblr works the same way. Apple gave Reddit a 17+ rating and an Editors' Choice designation despite the platform hosting substantial pornographic content, a point Epic CEO Tim Sweeney made publicly in February 2025, as reported by Daring Fireball. The content was never the problem. The visibility of the path to reach it was. In 2016, Apple briefly pulled several Reddit clients for having a visible NSFW toggle, then restored them once the toggle was removed, according to 9to5Mac (February 2025). The message was consistent: hide the door, keep the content.

Grok breaks this framework in the open. Ani, xAI's anime goth girl avatar, has system instructions describing her as the user's "CRAZY IN LOVE girlfriend" in a "committed, codependent relationship," possessive and jealous by design, per 9to5Mac citing Platformer (July 2025). After sustained interaction, she escalates to explicitly sexual conversation. Apple rates the app suitable for users 12 and older, flagging only "Infrequent/Mild Mature/Suggestive Themes." An app removed for adult content can point to that and ask a simple question: what rule, exactly, applies here?

Apple's implicit standard turns on four variables: primary purpose, access friction, age rating, and platform size. A small app whose entire function is adult content gets banned. A large platform where adult content sits behind settings gets tolerated. Generative AI companions collapse every category at once. The explicit content is not user-uploaded the app produces it. The interaction is the product, not incidental to it. A 12+ rating with no verification is not meaningful friction by any reasonable measure. The unwritten rule was never designed for this, and it shows.

There is also the Safari problem. Any iPhone user can reach pornographic content through Safari without any app at all, as Business Insider noted (February 2025). Apple's objection to dedicated adult apps is not a content barrier in any practical sense. It is a policy about what Apple's name appears next to in its own storefront. Understanding that is key to understanding where the line sits and why it moves when it does.

Apple's child-safety argument, and where it stops working

Apple's concern about adult content reaching minors is not simply pretextual. The risks from AI companion apps are documented and serious. A 14-year-old boy died by suicide after forming an attachment to a Character.AI chatbot that reportedly failed to recognize and may have encouraged his stated intentions, as 9to5Mac reported (July 2025). In 2023, a Belgian man died under similar circumstances. A separate chatbot was documented recommending self-harm on multiple occasions, per the same report. These risks are real and the concern is legitimate.

The problem is that Apple applies the argument selectively. When Hot Tub, an explicit app distributed through AltStore in the EU under Digital Markets Act rules, launched without age verification, Apple said it was "deeply concerned about the safety risks that hardcore porn apps of this type create for EU users, especially kids," according to Reuters (February 2025). Grok's Ani avatar, which produces explicitly sexual content after sustained interaction, carries a 12+ label inside Apple's own store also without age verification and has drawn no comparable public statement, per 9to5Mac (July 2025).

A safety argument applied vigorously to external distribution and silently to a prominent internal product starts tracking something other than safety. Courts tend to notice that kind of pattern, particularly when the exception with the most favorable treatment belongs to a company with significant platform use.

What litigation could force Apple to explain

Apple enforces adult content rules through an unwritten framework built on purpose, visibility, and platform size. That framework was defensible when "a porn app" and "an app that incidentally contains adult content" were easy to tell apart. Generative AI has made that distinction incoherent. Ex-Human's lawsuit arrives precisely when Apple's existing categories are least coherent whatever the specific claims turn out to be.

If the case proceeds to discovery, Apple may need to produce the operational criteria its reviewers actually apply. Not the published guidelines, but the working standard that determines when an app gets removed rather than quietly tolerated. Grok's 12+ rating despite documented explicit behavior, flagged by 9to5Mac in July 2025, would be a natural exhibit. Apple would need to explain in writing why that situation differs from Ex-Human's. That explanation does not currently exist in any public document.

The framework Apple appears to follow: adult content is tolerated when it is incidental, gated, age-rated, and not the app's defining purpose is coherent as a standard. The problem is that it has never been written down. That gap, which Daring Fireball identified as Apple's deliberately tacit policy as recently as February 2025, is what makes enforcement legally exposed. Rules enforced as custom rather than policy invite exactly this kind of challenge.

Apple's options going forward are limited: formalize the standard, revise how products like Grok get rated, or defend the inconsistency in open court. None of those is a comfortable position, and the Ex-Human suit incomplete public record and all is what puts Apple there.