Meta Ran Ads for Nudify Deepfake Apps: Moderation Failures and Legal Fallout (2026)

Meta Ran Ads for a Deepfake Nudify App Targeting Women

In August 2026, Meta ran paid advertisements on its platforms Facebook and Instagram for an app called Kromix, which explicitly promoted the ability to “nudify” real women without their consent. The ads, uncovered by WIRED, featured a deepfake image closely resembling a prominent female US politician. The app, which was available on Apple’s App Store, allowed users to upload a photo of any real person and generate a fake nude image using AI. This represents a direct failure of Meta’s ad review systems, which are supposed to block ads containing sexual content or promoting nonconsensual intimate imagery.

The Kromix App and Its Capabilities

Kromix marketed itself as an AI-powered tool for creating “undressing” images. The ads seen by Meta users explicitly targeted female politicians, with one ad showing a deepfake of a recognizable elected official. According to reports, the app was able to generate realistic-looking nude images of real people without their permission, a clear violation of both Meta’s advertising policies and Apple’s App Store guidelines. Times Now News reported that Kromix was designed to exploit women, particularly public figures, for harassment and blackmail. The app had been live on the App Store for an unknown period before its removal.

Meta’s response to the WIRED inquiry was to remove the ads, delete the associated Facebook pages, and block the URLs linked to the app. However, the ads had already run and been seen by an unknown number of users. Ars Technica noted that this was not an isolated incident, but part of a “series of failures by Meta to keep advertisements for tools that produce nonconsensual intimate imagery off its platforms.”

Apple Pulls Kromix from the App Store After Investigation

Following the WIRED investigation and a direct inquiry from the publication, Apple removed Kromix from its App Store. According to iThinkDiff, the app was pulled because it violated Apple’s guidelines against harmful content, including apps that facilitate nonconsensual pornography. Apple’s removal was reactive: it only acted after the media spotlight was turned on the app. This raises questions about the effectiveness of Apple’s own pre-release app review process, which is supposed to catch such violations before apps reach users.

Comparison of the Two Major Incidents

The controversy over Kromix is the most recent, but it is not the only instance of Meta failing to moderate ads for harmful deepfake tools. The table below compares the two most significant incidents reported in August 2026.

Incident	App / Tool	Platform	Action Taken by Meta	Action Taken by Apple	Source
Deepfake nudify ads for female politicians	Kromix	Facebook, Instagram	Removed ads and pages after WIRED contacted them	Removed app from App Store after investigation	WIRED
AI-generated child sexual abuse imagery ads	Unknown (multiple ads)	Facebook, Instagram	Ads contained CSAM imagery; Meta’s internal systems failed to block them	N/A (not an app)	WIRED

Both cases underscore a systemic failure in Meta’s automated and human moderation systems. The company claims to use advanced AI to detect harmful content, but in these instances, the AI either did not flag the ads or was bypassed by malicious actors.

Meta Also Ran Ads Containing AI-Generated Child Sexual Abuse Imagery

In a separate but related investigation also published in August 2026, WIRED revealed that Meta had run advertisements that contained AI-generated child sexual abuse imagery. These ads were discovered by researchers and internally reported to Meta. The fact that such imagery—which is illegal in virtually all jurisdictions—made it through Meta’s ad review process is a grave failure. The production and distribution of child sexual abuse material (CSAM) carries severe criminal penalties, and Meta’s inability to prevent these ads from appearing on its platforms opens the company up to legal liability.

Meta has long faced criticism for its content moderation practices. In 2025, the company relaxed its hate speech and harassment policies in an effort to “support free expression.” Critics argue that these policy shifts have created loopholes that enable harmful content, including nonconsensual deepfake ads and CSAM, to proliferate.

Broader Legal and Regulatory Consequences for Meta

The ad controversies come at a precarious time for Meta. The company is already facing an “astronomical” legal fight with multiple US state attorneys general, according to CNBC. The trial, which began in August 2026, centers on allegations that Meta knowingly harms young users by designing addictive features and failing to enforce safety measures. The Kromix and CSAM ad scandals add to the narrative of Meta’s inability to protect its users from exploitation.

Additionally, a US judge ruled that Meta “wiped key scam evidence” in a lawsuit filed by Australian businessman Andrew Forrest, as reported by the Australian Financial Review. The case involves cryptocurrency scam ads that used Forrest’s likeness without permission. The destruction of evidence could lead to severe penalties, including financial sanctions or adverse inferences at trial.

What This Means for Meta’s Reputation

From a trust and safety standpoint, Meta’s behavior in 2026 is catastrophic. The company spends billions of dollars annually on content moderation and AI safety research, yet it continues to allow ads that promote tools for nonconsensual deepfakes and even CSAM. Investors and regulators are watching closely. The US Senate has held multiple hearings on AI-generated exploitation material, and Meta’s CEO Mark Zuckerberg has been called to testify. The failure to prevent these ads may accelerate new legislation requiring platforms to verify the safety of AI tools advertised on their sites.

Conclusion: A Pattern of Proactive Negligence

Meta did not merely fail to catch harmful ads—it actively ran them. The Kromix app ads were paid advertisements, meaning Meta’s billing systems processed payments and its ad delivery systems served the ads to targeted users. The company only reacted after media inquiries. This pattern of reactive rather than proactive moderation is dangerous, especially as AI tools become more sophisticated. The same technology that generates nonconsensual deepfakes can be weaponized for blackmail, political sabotage, and exploitation. Until platforms like Meta invest in robust, real-time content detection that prioritizes safety over ad revenue, these incidents will continue to happen.

AI Weekly summarized the situation succinctly: “Meta’s direct complicity in promoting technology for nonconsensual deepfake creation highlights the ongoing struggle for app stores and social platforms to consistently enforce their guidelines.”

Users should be concerned about the safety of their platforms. Advertisers should be concerned about their brands being associated with such content. And regulators should accelerate enforcement against companies that profit from harm.