Images of Australian children are being used to create powerful artificial intelligence (AI) tools without parental consent or knowledge, according to Human Rights Watch.
The organisation says photos are being scraped off the internet and placed in large data sets that companies can use to then train generative AI systems.
From there, others can use the tools to create deepfake images, putting children at risk of exploitation.
“Children should not have to live in fear that their photos might be stolen and weaponised against them,” Human Rights Watch researcher Hye Jung Han said in a statement.
“The Australian government should urgently adopt laws to protect children’s data from AI-fueled misuse.”
Human Rights Watch analysed a fraction of the 5.85 billion images and captions in the LAION-5B training data set for AI, and found 190 photos of children from Australia, ranging from newborn babies, to school students in Book Week costumes.
Australian data scientist Ian Oppermann said the use of the images was the exact kind of “collateral damage” associated with what he termed “blunt and careless” harvesting of data for AI creation.
“It is not very hard to think about consequential issues like this before unleashing data harvesting applications,” he told The Epoch Times. “It also requires us to think more carefully about what we voluntarily put online.”
Meanwhile, Human Rights Watch’s Ms. Han said the discovery of the Australian images meant both parents and politicians should be taking note.
“Children have the right to safety where they live, learn and play, including online,” she said.
“To prevent and address such violations in the future, changes are urgently needed in legislation, regulation and industry practice.”
The federal government is expected to introduce Privacy Act reforms to Parliament in August, including the Children’s Online Privacy Code.
It follows the introduction of laws in Australia last month to criminalise the creation and sharing of non-consensual sexually explicit deepfake images and videos.
