Meta Platforms Inc's Facebook and Instagram served thousands of advertisements for artificial intelligence applications designed to transpose women's faces onto nude bodies and fabricate sexualised imagery, according to an investigation by the Tech Transparency Project, a nonprofit research organisation focused on holding major technology companies accountable. The discovery highlights a significant gap between Meta's stated policies and actual enforcement, particularly regarding content that originates from China and targets international audiences through intermediary advertising agencies.
The investigation traced the campaign to Beijing-based GatherOne Inc, which functions as a critical bridge connecting Chinese businesses to Meta's advertising infrastructure. Between April and June of this year, GatherOne facilitated approximately 7,600 advertisements promoting mobile applications capable of creating synthetic intimate imagery. More troublingly, the watchdog identified one such app that enabled users to share what some flagged as child sexual abuse material, raising alarms about how thoroughly Meta screens its advertising partners and the content they promote.
Meta's own advertising policies explicitly forbid sexually suggestive promotions and advertisements for applications that purport to remove clothing from images or create non-consensual sexual content. The prolific nature of these violations—numbering in the thousands—suggests either inadequate screening mechanisms or insufficient consequences for repeat violations by partners. Katie Paul, director of the Tech Transparency Project, characterised Meta's oversight as granting "a free pass to one of its top Chinese advertising partners when it comes to nudify ads," implying a deliberate choice to overlook infractions rather than technical negligence.
The revelation proves particularly significant given Meta's financial dependence on Chinese advertising revenue. The company generated US$18.4 billion (RM75.2 billion) from China-based advertisers in 2024, representing 11 percent of its global advertising income and more than double the figure from 2022. This dependency creates structural incentives to accommodate Chinese advertising partners even when they violate content policies, particularly when enforcement would directly reduce profitable revenue streams. The financial stakes help explain why Meta might prioritise commercial relationships over strict compliance verification.
One specific case underscores the severity of the issue. An application called BAfter, developed in China's Guangxi province, was promoted through GatherOne's advertising network. Upon inspection, the app contained a "ShareZone" section offering pornographic face-swap options and videos depicting women in sexual situations. More alarming still, multiple user reviews on Google Play flagged the presence of illegal child sexual abuse material and synthetic videos of minors, yet the application was classified as suitable for all ages. The app reportedly accumulated more than 2.5 million downloads globally, suggesting the advertising campaigns successfully drove substantial user adoption.
GatherOne's response, following media inquiries, involved suspending new advertising accounts for AI nudification and face-swap services and promising a "comprehensive compliance upgrade." However, this reactive posture emerged only after the Tech Transparency Project's findings became public, raising questions about whether voluntary action would have occurred absent external scrutiny. The advertising intermediary maintains its website lists it as an official Meta partner in China, operating alongside ten other agents, indicating Meta's continued reliance on such intermediaries despite documented compliance failures.
Meta's official statements attempt to reconcile the contradiction between policy and practice. Cindy Southworth, the company's Head of Women's Safety Policy, asserted that Meta "does not allow non-consensual intimate imagery or nudify apps" and takes "aggressive steps to combat them." A company spokesperson added that all advertisers must comply with advertising standards and that violations can trigger account termination and financial penalties. Yet these declarations ring hollow against the documented reality of thousands of ads running across multiple months, suggesting either enforcement mechanisms are ineffective or enforcement actions follow discovery rather than preceding violations through proactive auditing.
The pattern reflects a broader institutional challenge within Meta's operations. A Reuters investigation in December documented that internal Meta analyses concluded approximately 19 percent of its China revenues derived from advertisements for prohibited content including pornography, scams, and illegal gambling. Similarly, a former Meta executive turned whistleblower, Sarah Wynn-Williams, testified before Congress that the company has historically offered censorship tools to China and facilitated its advancement in artificial intelligence development, demonstrating a willingness to compromise policy principles for market access and revenue expansion.
Beyond Meta's enforcement failures, the investigation exposes vulnerabilities across the entire technology ecosystem. Google, which operates the Android operating system and Google Play store where these apps proliferated, declined to comment when queried about the matter. The delayed removal of BAfter from Google Play—occurring only on July 21, after the investigation identified problems—indicates reactive rather than preventive content moderation. Similarly, Apple Inc has been identified in earlier Tech Transparency Project research as failing to systematically prevent nudification apps from its ecosystem, suggesting this represents an industry-wide challenge rather than Meta-specific negligence.
The implications extend into regulatory and legislative domains particularly relevant for Malaysia and Southeast Asia. A recent survey of 557 American teenagers revealed that more than half had used nudification tools to create imagery of themselves or others, while over one-third reported being victims of such applications. This adolescent exposure to synthetic intimate imagery creates psychological harms and potential legal jeopardy for young users who may be creating or sharing such content without understanding its implications. The United States has responded through the Take It Down Act, which criminalises publishing non-consensual sexual content and requires social media platforms to remove such material, with the Federal Trade Commission beginning enforcement in May.
Southeast Asian jurisdictions, including Malaysia, should examine whether their regulatory frameworks adequately address deepfake pornography and synthetic intimate imagery. As technological capabilities advance and advertising networks like GatherOne facilitate distribution across borders, enforcement within individual countries becomes insufficient. The Malaysian government, through the Communications and Multimedia Act and related legislation, should consider whether existing provisions adequately criminalise the creation and distribution of non-consensual deepfake intimate imagery, particularly when created by or distributed through international technology platforms.
The GatherOne case demonstrates how Meta's business model—deriving substantial and growing revenue from advertising intermediaries operating in jurisdictions where its core services remain blocked—creates inherent tension between compliance and profitability. When revenue streams depend on advertising partners in countries where content moderation oversight is more difficult and enforcement mechanisms weaker, the economic incentives systematically favour overlooking violations. Until regulatory frameworks impose penalties sufficient to outweigh forgone revenue, or until advertisers themselves face prosecution for facilitating prohibited content distribution, such gaps between stated policy and actual enforcement are likely to persist.
