Angus Crawford and Tony Smith
BBCNews
Some pedophiles use Artificial Intelligence (AI) technology to create and sell material that simulates real-life situations of child sexual abuse, a researcher tells the BBC.
They gain access to the images by paying for account subscriptions on popular content-sharing sites like the Patreon platform.
The company stated that it has a “zero tolerance” policy against such images on its site.
The UK’s National Police Chiefs Council (NPCC) said it was “outrageous” that some platforms were making “huge profits” by failing to take “moral responsibility”.
For its part, the British intelligence and cybersecurity agency GCHQ responded to the report saying: “Child sexual abuse criminals embrace all technologies and some believe that the future of child sexual abuse material lies in AI-generated content.”
The producers of the abuse images are using an AI program called Stable Diffusion, whose original function was to generate images for art or graphic design.
AI enables computers to perform tasks that typically require human intelligence.
Stable Diffusion software allows users to describe the image they are looking for, through keywords, so that the program then generates that image.
But the BBC has found that it is being used to create realistic images of child sexual abuse, including the rape of babies and minors.
British police teams investigating online child abuse say they have already come across such content.
Freelance journalist Octavia Sheepshanks investigated the matter for several months. She contacted the BBC through the charity Society for the Prevention of Cruelty to Children (NSPCC) to make her work public.
“Since AI imaging became possible, there has been a huge influx…not just very young girls, it’s about [pedófilos] Speaking of minors,” he said.
A computer generated “pseudo image” showing child sexual abuse is considered in the UK to be a real image and is illegal to possess, post or transfer.
NPCC child protection director Ian Critchley said it would be wrong to claim that no harm is being done to anyone because no real children are depicted in such “synthetic” images.
Critchley warned that a pedophile could “progress along that scale of offense from thought, to synthetic, to actually abusing a real child.”
Abuse images are shared through a three-stage process:
- Pedophiles generate the images with AI software
- They promote the photos on platforms like Pixiv, a Japanese image-sharing site
- These accounts have links that direct customers to more explicit images, which they can pay to look at on accounts on sites like Patreon.
Some of the creators of the images are putting them up on a popular social media platform called Pixiv, which is mainly used by manga and anime share artists.
However, because the site is based in Japan, where the sharing of sexual cartoons and illustrations of minors is not illegal, the creators use it to promote their work in groups and via tags, which create topic indexes based on words. clue.
A Pixiv spokesperson noted that the site places a lot of emphasis on addressing this issue. He claimed that on May 31 he had banned any realistic depiction of sexual content involving minors.
The company said it had proactively strengthened its monitoring systems and was devoting substantial resources to counter AI-related issues.
Octavia Sheepshanks told the BBC that users appeared to be producing images of child sexual abuse on an industrial scale.
“The volume is immense, to the point where [los creadores] they say that ‘we aim to produce at least 1,000 images a month,’” the independent journalist reported.
User comments on individual images on Pixiv make it clear that they have a sexual interest in minors, with some of them even offering to contribute non-AI generated abuse images and videos.
Sheepshanks has been monitoring some of these groups on the platform.
“Between those groups, which will have 100 members, people will be sharing, ‘Oh look, here’s a link to the real thing,’” he says.
“The most horrifying things, I didn’t even know those words existed. [descripciones]”.
Prices
Many of the accounts on Pixiv include links to their bios that direct people to what they call their “uncensored content” on the US-based content sharing site Patreon.
Patreon is valued at $4 billion and claims to have over 250,000 creators, most of whom are legitimate accounts belonging to well-known celebrities, journalists, and writers.
Their followers can support the creators by signing up for a subscription that gives them access to blogs, podcasts, videos and images for a cost starting at US$3.85.
But our investigation found Patreon accounts offering to sell obscene photorealistic images of children, with different price levels depending on the type of material required.
One wrote on his account: “I train my girls on my PC,” adding that they display “submission.” For $8.30 a month, another user offered “uncensored exclusive art.”
The BBC sent Patreon an example, which the platform confirmed as “semi-realistic and violates our policies.” He said that the account had been immediately removed.
Patreon stated that it has a “zero tolerance” policy, insisting that: “Creators may not fund content dedicated to sexual themes involving minors.”
The company said the rise of harmful AI-generated content on the internet was “real and concerning”, adding that it had “identified and removed “increasing amounts” of such material.
“We already ban synthetic AI-generated child exploitation material,” he said, describing the site as “very proactive,” with teams, technology and partners dedicated to “protecting teens.”
The Stable Diffusion AI imager was created as an international collaboration between academics and a number of companies, led by the firm Stability AI.
Several versions have been released, with restrictions written into the coding that control the type of content that can be created.
But last year, a previous “open source” version was released publicly that allowed users to remove any filters and empower themselves to produce any image, including illegal ones.
Stability AI told the BBC that it “prohibits any misuse for an illegal or immoral purpose on our platform, and our policies make it clear that this includes [material de abuso sexual infantil]”.
“We strongly support law enforcement efforts against those who abuse our products for illegal or nefarious purposes.”
As AI develops rapidly, questions have been raised about the future risks it could pose to people’s privacy, human rights, and security.
Jo [un pseudónimo por razones de seguridad]director of GCHQ’s child sexual abuse mission, told the BBC: “GCHQ supports law enforcement agencies to stay ahead of emerging threats such as AI-generated content and to ensure there is no safe haven for criminals.” offenders”.
Ian Critchley of the NPCC stated that he was also concerned that the flood of realistic or “synthetic” AI images could hinder the process of identifying actual victims of abuse.
“It creates an additional requirement, in terms of police surveillance and law enforcement to identify where an actual child, wherever it is in the world, is being abused as opposed to an artificial or synthetic child,” he explained.
Critchley further noted that he believed it was a turning point for society.
“We can make sure that the internet and technology can create fantastic opportunities for young people, or make it a much more damaging place,” he concluded.
Remember that you can receive notifications from BBC News World. Download the latest version of our app and activate them so you don’t miss out on our best content.
Do you already know our YouTube channel? Subscribe!
See original article on BBC