Apple has removed the AI-powered 'Kromix' app from its App Store after disturbing reports revealed that sexually explicit deepfake images, including one depicting a female US politician, appeared in advertisements running across Meta's platforms. The incident has intensified scrutiny on both app store content policies and advertising moderation.
Meta Ads Promote Explicit AI Content
According to a report by Wired, Meta platforms, including Facebook and Instagram, ran advertisements for Kromix, an AI tool marketed as an image editor but capable of generating explicit content. One ad reportedly featured a sexually explicit AI-generated image of a female US politician, raising serious questions about Meta’s enforcement of its own policies against sexual material.
The advertising campaign reportedly consisted of 32 separate ads, primarily targeting male users, and ran for varying durations between five and 46 hours before being taken down by Meta.
Apple's Stance and Previous Removals
Apple confirmed its action, stating, "We have always strictly prohibited apps designed to generate, distribute, or consume pornography." The company emphasized that "‘Nudification’ apps are against our App Review Guidelines, and we aggressively work to identify and remove apps—as well as the developer accounts behind them—that attempt to evade our processes and policies."
This removal of the AI nudify app, Kromix, is not an isolated event. Apple has recently taken down several other apps with similar capabilities, including 'MaskAI,' which allowed users to face-swap individuals into sexual scenarios. These actions often follow media inquiries highlighting such problematic applications.
Growing Threat of Nonconsensual Deepfakes
The proliferation of AI tools capable of generating nonconsensual intimate imagery poses a significant threat, particularly to public figures and women. In January, AI-generated images of women and girls, reportedly numbering around 3 million, led to investigations and regulatory pressure on platforms. Social media companies are increasingly under pressure to develop robust mechanisms to detect and remove such harmful content, with some, like X, restricting AI's ability to create suggestive or explicit depictions of real people.