Images generated by artificial intelligence (AI) and depicting real victims of child sexual abuse are now so realistic they would breach UK law, an internet watchdog has warned.
The Internet Watch Foundation (IWF) says AI-generated child abuse images are threatening to “overwhelm” the internet.
IWF said it found 11,108 AI images had been shared on a single dark web child abuse forum in one month, and it said 2,978 of these were images which were in breach of UK law and 2,562 were so realistic they would have to be treated on a par with real images of abuse.
The most convincing imagery would be difficult even for trained analysts to distinguish from actual photographs, said the IWF.
Home Secretary Suella Braverman issued a joint statement in May with U.S. Homeland Security Secretary Alejandro Mayorkas in which they committed to tackle the “alarming rise in despicable AI-generated images of children being sexually exploited by paedophiles.”
‘Our Worst Nightmares Have Come True’
On Wednesday, Susie Hargreaves, the chief executive of the IWF, said: “Our worst nightmares have come true. Earlier this year, we warned AI imagery could soon become indistinguishable from real pictures of children suffering sexual abuse, and that we could start to see this imagery proliferating in much greater numbers. We have now passed that point.”“Chillingly, we are seeing criminals deliberately training their AI on real victims’ images who have already suffered abuse,” she said.
The IWF said criminals were now using AI to manipulate and clone the faces and bodies of real children who have appeared in real child abuse images.
They also found evidence of images of child actors being manipulated to make them appear involved in sexual abuse.
The IWF researchers also found images of famous singers and actors, including one famous female celebrity, whose images had been digitally manipulated to make them look like children who were involved in sexual scenarios.
They also found AI had been used to “nudify” images of child models and other infants whose clothed images had been uploaded for legitimate reasons.
In one example, hundreds of images of two girls taken from a legitimate modelling agency photoshoot had been manipulated and inserted into Category A—the highest and most obscene category—sexual abuse scenes.
Dangers of XR Technology
In its global threat assessment, the alliance also warned of the dangers of paedophiles using extended reality (XR) technology.The report says: “The global market for XR is forecasted to surpass $1.1 trillion by 2030. It is likely offenders will increasingly exploit XR technologies as they become more accessible and affordable.”
The NSPCC chief executive Sir Peter Wanless said the Online Safety Bill, which is due to get Royal Assent this week, would be crucial to help stop online abuse.
Sir Peter said: “With the eye-watering scale of online sexual abuse, it is crucial we do not forget the impact it is having on children and commit to working together to prevent it.”
The two-day event will be the first major global summit on AI safety. Attendants reportedly will include Canadian Prime Minister Justin Trudeau, French President Emmanuel Macron, U.S. Vice President Kamala Harris, and Ursula von der Leyen, the president of the European Commission.
