Social media app investigated over child abuse material
Zac de Silva |
An Elon-Musk-backed AI chatbot that allows users to request explicit images of real people is under investigation by Australia’s online safety watchdog for generating child exploitation material.
Grok, the artificial intelligence program linked with social media platform X, has attracted a flurry of criticism for allowing users to generate sexualised images of people based on real photos.
Some users have also complained about the tool being used to create pictures of children in compromising positions.
eSafety Commissioner Julie Inman Grant said her agency had two active investigations into the AI app’s role in creating child exploitation material.
“We have two investigations, one into X for hosting child sexual abuse material,” she told reporters in Brisbane on Friday.
“The second that we issued yesterday is around our designated internet standard, which prevents high-impact AI generative services from creating child sexual abuse.
“We’re taking and removing content that’s non-consensual, intimate imagery of adults.
“But I think we all agree that the generation of (child sexual abuse material) is heinous and this needs to be more closely looked at,” Ms Inman Grant said.
X AI responded to AAP’s request for comment with what appeared to be an auto-generated email reading “Legacy Media Lies”.
The social media platform says it has implemented new measures preventing the public Grok account from creating pictures of real people in revealing clothing, after countries including the UK threatened tougher laws.
X has also limited image creation to paid subscribers only.
But AAP was still able to generate near-nude images of a real person using a free account connected to the standalone Grok website.
The federal government has promised to outlaw apps that allow users to create deepfake nude pictures but has yet to introduce legislation to do so.
It expects to receive advice from departmental officials in coming months and has met with Apple, Google and Microsoft to discuss ways to restrict access to nudifying apps.
Communications Minister Anika Wells said she was speaking with the eSafety commission about the issue.
She said a digital duty of care – a separate policy the government plans to introduce – would crack down on “abhorrent” behaviour online, and the image creation tools currently provided by Grok would be a clear breach of those proposed rules.
“A digital duty of care asks Australians to be clear about what these platforms owe us, what duty of care they owe us if these platforms seek to conduct business on Australian shores,” Ms Wells told Nine’s Today Show on Friday.
Opposition communications spokeswoman Melissa McIntosh said the government needed to move faster to crack down on nudifying apps.
“It’s absolutely unacceptable that nude images can be created of an individual, whether it’s a child, a woman, anyone … and everything should be thrown at this particular issue,” Ms McIntosh said.
AAP
