They are part of a growing trend of accounts on Instagram and TikTok that has been criticised as racist, exploitative and misleading because of racial tropes and language used.

The images and videos were generated by AI but not labelled as such, in apparent breach of the platforms' guidelines.

Nearly all the accounts were on Instagram and about a third also had versions on TikTok.

Instagram's parent company Meta told the BBC it was investigating, but did not say it had taken any action.

Account names include terms such as "black", "noir", "dark" and "ebony". Several include comments about white males in their posts, such as "loves white men" and "why I need a white guy in my life". Many of the accounts follow or like each other.

The BBC, working in collaboration with analysts Jeremy Carrasco and Angel Nulani from Riddance, has identified 60 such accounts, mainly on Instagram, that have carried links, or chains of links, to paid-for sexually explicit content on third-party sites.

The sites labelled the imagery as AI-generated, but the Instagram accounts did not.

The research also identified many more accounts on both Instagram and TikTok with similar AI-generated avatars that did not link to paid content.

One of the accounts shut down by TikTok - though still operating on Instagram at the time of publication - has caused further anger by stealing videos from real people.

It is presented as the account of a striking AI-generated character and had amassed three million followers within a few weeks of its creation in December.

"I was angry," Riya told the BBC.

"Of course my videos are all out there… It doesn't mean that you can just take it and steal it and post it as your own. "

While the three videos that clearly match Riya's content are not sexual, other videos on the AI account using the same digitally created character show it in revealing clothing or performing provocative actions.

A chain of links from the account leads to paid-for adult content. "I'm not sure if I'm more concerned about them taking my video to promote their explicit content or [that] people actually believe in that," she says.

It is becoming harder for users to tell whether content is real and "people keep on falling for these AI models", she adds.

Many viewers appear to treat the avatars as real, despite their unrealistic features. In posts or Instagram stories, some of the accounts deny using AI, including the one that took Riya's content.

Riya says she reported the account to both platforms multiple times but, at the time, the content was not removed.

TikTok banned it after the BBC approached it for comment.

"I believe these accounts are racist because their existence perpetuates a long history of the exploitation of black people," says Nulani, one of the researchers.

"Their use of caricatures, race-play terminology and unrealistic depictions of black women prove they're not concerned with our safety or wellbeing, but our ability to be capitalised as part of the online porn machine," she adds.

Carrasco, who critiques AI trends and techniques on his social media accounts, says "the new thing is the quantity of shameless, racist depictions of extremely black people".

While "that fetish" may have existed in the past, AI "gives it new purchase", he says.

He explains that AI makes it easier to remove undertones in images and videos in order to create dark skin tones that are not natural, and to create effects that would previously have required animation or skin painting.

Also, he says, there are no social consequences for an avatar: "There's no shame… that's something AI uniquely exploits. "

It says it wants users to know when they are looking at posts that have been made with AI, and that it has policies in place for the labelling of AI-generated content.

Nine Instagram accounts the BBC has tracked appear to no longer exist

Highlighted sentences link to their corresponding claims. Click any highlighted sentence to jump to its detailed analysis.
Highlight Colors Indicate Claim Quality:
✓ Healthy Claim - No fallacies or contradictions detected
⚠️ Minor Issues - Has contradictions or minor fallacies
🚨 Serious Issues - Multiple contradictions or severe fallacies
Quality Criteria: Claims are evaluated for logical fallacies and contradictions with other news sources. Green highlights indicate healthy claims suitable for reference.
Source