TikTok has deleted 20 accounts after a BBC investigation that uncovered a troubling trend of AI-generated black female avatars deployed to encourage users to engage with sexually explicit content. The platform took action after the BBC and experts at the independent AI publication Riddance discovered numerous accounts across Instagram and TikTok featuring heavily sexualised digital characters with exaggerated proportions and darkened skin. The accounts, which were unlabelled as artificially created in apparent breach of platform guidelines, employed racist tropes in their naming practices and marketing copy. Whilst TikTok has acted swiftly, Meta, Instagram’s parent company, said it was looking into but has not confirmed taking similar action against the accounts active on its platform.
The Discovery: Many Misleading Profiles
The BBC’s inquiry, conducted in collaboration with Riddance researchers Jeremy Carrasco and Angel Nulani, revealed a sophisticated operation spanning multiple platforms. The team discovered 60 accounts mainly found on Instagram that included links directing guiding users towards commercial adult content on third-party websites. Notably, whilst these third-party sites marked the content as AI-generated, the Instagram accounts in turn provided no such information, resulting in calculated dishonesty for unsuspecting users. The research revealed an even larger ecosystem of similar accounts across both Instagram and TikTok that failed to direct users to paid content, suggesting the problem goes well past profit-driven abuse.
The accounts utilised a deliberate approach to evade detection and expand reach quickly. The vast majority of accounts were focused on Instagram, with around one in three also maintaining versions on TikTok. Account names specifically drew upon racially charged language, featuring words like “black”, “noir”, “dark” and “ebony”, paired with content advancing racial stereotypes and sexualisation. Numerous accounts maintained connections to one another, establishing linked systems that amplified their reach and exposure. This coordinated approach suggests organised activity rather than isolated incidents, raising serious questions about the scale and sophistication of the campaign.
- AI avatars featured exaggerated body shapes and heavily tanned skin tones
- Account names employed racially charged language and stereotypical depictions about white men
- Videos were unlabelled as AI-generated, contravening platform guidelines
- Many accounts were interconnected, establishing interconnected networks for amplification
Abuse Via Artificial Creation
Unauthorised Material and Digital Manipulation
The investigation exposed a particularly troubling facet to the enterprise: the systematic appropriation of creative work from real creators. One profile that gathered three million users within weeks after its December release had systematically appropriated content from actual creators, most notably Malaysian influencer Riya Ulan. The offenders superimposed the AI-generated avatar’s face—displaying an darkened skin tone—onto Riya’s form, carefully copying her actions, attire and backdrop. This flagrant act of identity theft compounded the original exploitation, converting real creative work into material for misleading accounts.
Riya’s involvement illustrated the breach of privacy fundamental to such methods. Upon discovering her material was taken and repurposed, she expressed her distress to the BBC: “I was frustrated. Of course my videos are widely distributed… It doesn’t mean that you can grab it and steal it and post it as your own.” Her distress reveals a major weakness in safeguarding systems, where creators miss adequate safeguards against having their likenesses appropriated for fraudulent accounts. The case reveals how machine learning can magnify present dangers, permitting fraudsters to scale their exploitation across numerous creators in parallel.
The misuse surpassed basic content piracy. Accounts intentionally created artificial personas with exaggerated bodily features and digitally altered complexions to produce eye-catching, eye-catching content crafted to generate interaction. These algorithmic creations were presented as real personalities, featuring false narratives and character traits, whilst simultaneously perpetuating harmful ethnic prejudices and fetishisation tropes. The complexity of the scheme meant many users engaged with the posts believing they were connected to actual persons, not algorithmic constructs engineered to funnel engagement toward exploitative paid content.
- Avatar’s face overlaid onto stolen videos from genuine content creators
- Digitally darkened skin tones and pronounced characteristics designed for engagement
- Accounts marketed as authentic influencers with false backgrounds and personas
Racial Stereotypes and Damaging Stereotypes
The accounts identified by the BBC and Riddance researchers exhibited a troubling pattern of racist abuse through their deliberate naming conventions and content. Account identifiers included terms such as “black”, “noir”, “dark” and “ebony”, whilst posts regularly contained race-based terminology and fetishisation commentary. Many featured comments such as “loves white men” and “why I need a white guy in my life”, perpetuating harmful stereotypes and portraying Black women to sexual commodities. This communicative structure transformed the avatars into dehumanized depictions, reinforcing centuries-old racist tropes that commodify and dehumanise Black femininity.
The graphic depiction intensified these linguistic harms. The avatars were regularly shown in exposed swimming attire and skimpy clothing, their bodies algorithmically altered into exaggerated proportions designed to maximise engagement through sexualisation. Skin tones were artificially darkened to create an artificial, near monstrous appearance that had scant similarity to authentic human variation. This combination of exaggerated features, revealing attire and artificially manipulated characteristics created algorithmic stereotypes that functioned as contemporary racist performance, converting discriminatory visuals into algorithmic content optimised for widespread distribution and financial gain.
Why AI Magnifies the Problem
Artificial intelligence technology has significantly changed the scale and sophistication of racist exploitation online. Previously, such harmful content needed substantial investment and manual effort to produce; AI generation enables widespread production of racist caricatures, enabling bad actors to produce hundreds of convincing fake accounts at minimal cost. The technology’s capability for producing photorealistic imagery provides misleading legitimacy to racist tropes, making them appear credible to unsuspecting users. This technological advantage converts what could be fringe exploitation into a expandable, revenue-generating enterprise.
The computational nature of social media platforms amplifies these harms exponentially. AI-generated content designed to maximise engagement—particularly content that exploits racial and sexual stereotypes—proliferates quickly through algorithmic distribution mechanisms designed to prioritise user interaction. Platforms face challenges moderating the vast quantity of AI-generated content, whilst the absence of required transparency markers means users are unable to differentiate authentic creators from computational entities. This establishes conditions where racist stereotypes proliferate unchecked, obscured by the veneer of digital advancement and presented as entertainment rather than exploitation.
Platform Accountability and Response
| Platform | Action Taken |
|---|---|
| TikTok | Banned 20 accounts following BBC investigation; removed AI-generated black female avatars driving users to sexually explicit content sites |
| Parent company Meta stated it was investigating but had not confirmed taking action at time of BBC publication; dozens of accounts remained operational | |
| Both platforms | Failed to implement mandatory AI disclosure labels, breaching their own guidelines requiring identification of artificially generated content |
TikTok’s swift action in removing 20 accounts constitutes a rare example of platform action, yet it underscores the reactive approach of content management in the digital age. The bans came only after ongoing media investigation and public backlash, suggesting that absent external scrutiny, these accounts would have remained active unchecked. Notably, TikTok’s response was restricted to one platform, whilst numerous problematic accounts maintained active presences on Instagram, where they continued harvesting followers and sending users to exploitative third-party sites. This piecemeal strategy highlights the insufficiency of isolated platform measures to widespread challenges.
Meta’s claim to be investigating, coupled with its failure to announce specific steps, reveals the disconnect between corporate responsibility messaging and genuine enforcement. The company’s slow reaction differs markedly from the pressing need to tackle swiftly spreading damaging material. Both platforms have established guidelines requiring the revelation of machine-generated visual content, yet these regulations went unapplied across numerous profiles. This implementation shortfall indicates that without legislative action and statutory obligations, platforms will continue prioritising user engagement and ad income over platform security and the safeguarding of vulnerable groups from coordinated abuse and discriminatory content.
Extended Implications for Digital Authenticity
The rapid growth of unlabelled AI-generated content on major social media platforms raises key questions about online integrity and audience confidence. As artificial intelligence becomes increasingly sophisticated, telling apart real content creators and synthetic avatars grows exponentially more difficult for ordinary users. This erosion of authenticity threatens the core assumption upon which social media communities are built—the assumption that profiles show actual individuals posting authentic content. When vast numbers of people interact without realising with fabricated personas created to manipulate them, the networks’ trustworthiness as reliable platforms for interaction declines sharply. The lack of required transparency measures deliberately misleads audiences and compromises informed consent.
Beyond personal deception, the widespread deployment of AI-generated black female avatars represents a especially pernicious exploitation of racial identity in online environments. These artificial characters weaponise harmful stereotypes and sexualised racial imagery whilst simultaneously appropriating real creators’ content and labour. The practice reinforces the commodification of black femininity within algorithmic systems designed to increase user engagement and financial returns. This algorithmic discrimination operates extensively, affecting vast populations whilst staying largely undetected to content moderation departments. Without comprehensive regulatory frameworks requiring openness, authentication standards, and substantive penalties for violations, digital spaces will keep enabling widespread damage against underrepresented groups.
- AI transparency obligations must be binding in law across all platforms globally
- Authentication systems should verify the creator’s identity before monetisation or follower accumulation
- Racial stereotyping in artificially created material necessitates explicit prohibition with continuous identification
- Platforms must implement immediate content review with consequences matching content severity
