Cyber criminals are blackmailing schools with sexually explicit AI deepfakes of pupils created using photos taken from school websites, experts have warned.
Research published on Monday found that there is widespread awareness of the issue among UK schools, though few protections are in place. Of 250 teachers surveyed, nearly a quarter said that their school has already been affected.
The Internet Watch Foundation (IWF) said criminals had targeted schools by creating child sexual abuse material (CSAM) of real pupils using AI tools. They then threatened to publish the images if they did not receive money.
The UK’s National Crime Agency (NCA) issued a recommendation to educational institutions to remove any identifiable pictures of children from their websites and social media accounts.
“It is absolutely vital that schools nationwide rethink their policy around online images of their pupils,” child psychotherapist Dr Catherine Knibbs told The Independent.
“I have seen first-hand the impact this sort of crime can have on children. If you run a school, now is the time to act – rethink your photography policy before it is too late. And while you do so, use it as an opportunity to remind parents and guardians of the risks too.”
Some schools have already taken action by blurring faces of pupils, while a UK start-up unveiled a new safeguarding tool on Monday as a countermeasure to the emerging trend.
The platform, developed by child safety advocates at Aidos, uses artificial intelligence to make pupils in school photographs unidentifiable when posted online.
Each image of a face is replaced with an AI-generated one that matches the lighting, skin tone and context of the original, while remaining realistic.
“The online sexual abuse of children is reaching pandemic levels, and the emergence of AI is fueling the demand,” said Simon Bailey, the former national policing lead for child protection and adviser to Aidos.
“Aidos can prevent this form of sexual extortion in its tracks. I would urge every school leader in the country to act now. The technology criminals are using is already here, the question is whether schools are protected against it.”
Read MoreChina unveils humanoid robot for Moon mission
Why trademarking your face isn’t a ‘silver bullet’ against deepfakes
‘AI gave me your number’: AI doxxing turning ChatGPT hallucinations into harassment
