Instagram Investigating AI Profiles ‘Fetishising’ Disabled People

Instagram’s parent company, Meta, is investigating AI-generated social media accounts that sexualise disabled people appearing on its platform.

It comes after the BBC flagged dozens of profiles showing AI-generated images of women with disabilities, including Down’s syndrome or vitiligo.

Some profiles post fake images and videos of women with missing limbs, visible scarring or in wheelchairs. Many are in sexualised positions, wearing revealing clothing.

Some accounts have amassed hundreds of thousands of followers in a matter of months. One profile, claiming to be conjoined twins, has about 400,000 followers, despite only joining Instagram in December 2025.

Kamran Mallick, chief executive of Disability Rights UK, said the emergence of “accounts that fetishise, mock, or monetise the identities of disabled people is nothing short of horrific”.

He added: “What we have here is technology weaponised to strip disabled people of their agency and dignity, turning our lived experiences into digital caricatures for the profit and titillation of others.”

Medical charities have also voiced concern. A spokesperson for Gemini Untwined, which funds specialist surgery for rare newborns joined by the head, said portraying conjoined twins as entertainment was “morally reprehensible”.

They added: “We witness first-hand the medical challenges that these children and their families endure, which makes the portrayal of conjoined twins as a form of entertainment or spectacle especially problematic.”

Dr Amy Gaeta, from the University of Cambridge, told the BBC she researches how AI shifts power in relation to gender and disability. She said the internet is “flooded” with free and low-cost generative AI image-making tools.

Generative AI is a type of software that creates new content based on patterns it has learned from existing data in response to a user’s prompt.

Gaeta said while some tools have restrictions on content, like banning sexually explicit prompts, others do not or can be easily bypassed.

“Sometimes, without my prompting or intent, hypersexualized images of disabled people will be generated. This clearly shows a bias in the datasets that these tools are trained upon,” she said.

An Ofcom spokesperson said it was “tracking how AI is evolving, the risks that may emerge, and what actions may be needed to address them”.

They said the watchdog’s online safety rules require tech firms to tackle illegal content and protect children from harmful material – including abusive or hateful content.

The Online Safety Act requires platforms like Instagram to apply terms of service consistently, including where content mocks people based on protected characteristics, like their disability.

The Equality and Human Rights Commission said the accounts flagged by the BBC were “deeply disturbing”, adding: “It’s vital there are robust regulatory powers in the digital space to protect people from harm.”

A spokesperson for Meta said it was investigating the content and that it removes material that promotes sexual exploitation or attacks people based on protected characteristics.

While these accounts are generating fake personas, there is serious concern over the commands for the generative tools behind them.

Alison Kerry, head of communications at disability equality charity Scope, said the practice amounted to “discrimination dressed up as content”.

She added: “These AI images don’t appear from nowhere – they’re built from real disabled people’s images, often without consent – and unmoderated comment threads turbocharge objectification and harassment.”

Gaeta said moderation on sites like Instagram is not strong enough.

“Even when safeguards are in place, it has been proven relatively easy to bypass these if someone is insistent enough on doing so,” she said.

“Big tech needs to be held accountable just as much as misogyny and ableism need to be tackled.”