Deepfake kids: Child exploitation cops facing AI minefield

AI-generated deepfake pornography could tie up law enforcement agencies as they investigate crimes against fake children.

Experts fear a tsunami of AI-generated deepfake pornography and exploitation material will tie up law enforcement agencies who could waste their time and resources investigating crimes against fake children.

Detective Inspector Glen Donaldson, from the Queensland Police Task Force Argos team,

said he was concerned about the potential of artificial intelligence technology.

“We have concerns about material that can be generated that is not a real child, but looks like a real child,” he said.

’The risk is that law enforcement somewhere in the world will be spending their time and resources trying to rescue what we believe is a victim when it’s just a fake child.”

He said law enforcement was racing to catch up developments in technology, including deepfakes in which a person’s face was superimposed into an image.

“Again, the technology is getting better and better,” he said.

“Our primary focus is not arresting offenders, our primary focus is identifying children and removing them from harm and obviously arresting offenders is a by-product of that but we are a victim centric unit and most police agencies have the same mentality.”

Detective Inspector Glen Donaldson. Picture: Liam Kidston

Insp Donaldson said social media giants had a responsibility to help prevent child exploitation.

“I think it’s clear that they have a role to play in this, I fully agree with the e-safety commissioner’s principle of safety by design … these tech companies, it is their platform that the offending is happening on, where children are able to be contacted by someone anywhere in the world,” he said.

“There is a real obligation on them to do that and I fully support the e-safety commissioner’s moves to assist us with this because law enforcement alone cannot solve this problem.”

Insp Donaldson said Argos had identified sex offenders from all different age groups including children.

He said offenders’ motivations were sexually based and increasingly financially based.

“The sexual-based motivation is something that has been around since Argos first started, where they will socially engineer a child into believing that they are talking to someone who they are not,” Insp Donaldson said.

Former Victorian police officer turned cyber safety expert Susan McLean said deepfake porn was no longer discernible to the naked eye.

“(But) it doesn’t undo the harm to the child who thinks that a nude of themselves is being circulated, and we know that it is leading to suicide,” she said.