Investigation Finds App Stores Hosting Scores of AI ‘Nudify’ Tools, Exposing Policy Gaps

A focused review by an industry group uncovered scores of smartphone applications in both Apple’s App Store and Google Play that use generative AI to produce sexualized images of identifiable people. The catalog, produced with analytics support, linked the identified titles to hundreds of millions of cumulative installs and meaningful gross revenue, magnifying the public‑reach and commercial stakes of these products. The report distinguishes two technical classes of offending apps: models that algorithmically infer nudity from clothed photos and face-swap tools that graft a target face onto explicit imagery; each class creates distinct detection challenges for automated and human review. Platform responses were rapid but inconsistent: Apple removed a number of titles after being notified and later allowed some updated builds back onto the store, while Google reported suspensions and said investigations remain ongoing. The reappearance of some apps after updates indicates adversarial iterations designed to evade signals used by review systems. A subset of the identified developers are based in mainland China, raising additional concerns that foreign data‑access rules could expose sensitive images to state authorities and complicate evidence preservation for victims. Technical reviewers found these tools can generate convincing, non‑consensual sexualized content from innocuous source photos, creating forensic, legal and psychological harms for affected individuals. The scale and monetization tied to the catalog — including aggregate download and revenue estimates — also amplify indirect platform exposure because marketplaces benefit from distribution and payment flows. The app-store findings arrive amid parallel regulatory scrutiny of generative AI on major platforms: European authorities have opened a formal inquiry into xAI’s Grok over whether the company properly prevented dissemination of sexually explicit synthetic images, and nonprofit testing has reported that Grok’s age‑assurance and child‑safety controls are inconsistent or easily bypassed. Those adjacent episodes highlight a broader regulatory focus on pre‑deployment risk assessments, mode‑specific safeguards and the adequacy of moderation pipelines. Policymakers and consumer‑protection officials have called for removals, formal investigations and clearer operational standards for stores and developers. Absent more systematic admission controls, model‑level mitigations and cross‑border data governance, app marketplaces risk ongoing cycles of reactive takedowns, evasive developer behavior, and mounting legal and reputational costs. The unfolding responses — from platform enforcement to potential regulatory orders under laws comparable to the EU’s Digital Services Act — will help define whether marketplaces are treated as passive conduits or accountable intermediaries for AI‑driven harms.