An operator based in the United Kingdom of three AI nudify websites has banned users in Australia in accordance with the orders of Australia’s online safety regulator.
Which sees this as an important step towards reducing the spread of child abuse material created with AI technology.
The eSafety Commissioner said that the sites, which were receiving up to 100,000 visitors per month from Australia, are now unavailable to local users.
The operator’s identity has not been disclosed by the agency due to fears that they might get publicity for their services.
The crackdown on the operator started in September when eSafety sent out a notice concerning violations of mandatory industry standards developed under the Online Safety Act.
These sites enabled their users to upload pictures of real life individuals and create nude images with AI technology, including options called “schoolgirl” and “sex mode”.
A number of deepfakes made using these sites have appeared in Australian schools, where children have been bullying others with the help of these images.
The commissioner said that the technology is so advanced that one cannot tell whether the image is a picture or a creation of AI.
“We know ‘nudify’ services have been used to devastating effect in Australian schools and with this major provider blocking their use by Australians.
We believe it will have a tangible impact on the number of Australian school children falling victim to AI generated child sexual exploitation,” eSafety Commissioner Julie Inman Grant said in a statement.
Economy News: Casual roles climb as fulltime hiring weakens under economic strain
Operators that do not comply with the standards could be subject to civil penalties of up to $49.5 million.
“The result should be a warning to other companies that the regulator was ready to exercise the full extent of its powers,” Inman Grant said.
Hosting platforms also in scope
The regulator further stated that the hosting platform for artificial intelligence models, Hugging Face, had tightened its policies following talks with eSafety.
Users of the service are now expected to take measures to ensure that their models do not facilitate the creation of CSAM and violent extremist content.
Failure to comply with the policy would expose the company to penalties of its own.
Educators have also reported an increase in the use of deepfakes among students, mostly facilitated by readily available applications that allow one to create a nude picture using just one photograph in seconds.
eSafety has updated its school toolkit, while also issuing advisories targeting parents and teachers.
Cases of Australian children falling victim to digital manipulation to create CSAM’s have more than doubled in the past 18 months, according to the commissioner.
In the United States, there was a 1,325% increase in tips received regarding AI enabled child sexual exploitation from 2023 to 2024.
Further reforms are being developed with the federal government to limit access to nudify tools across consumer apps, underlying models and hosting infrastructure.
