Child sexual exploitation crisis: AI-generated abuse material a ‘diabolical scenario’ for authorities

Australia’s top cop is warning that children are no longer safe at home if they have unsupervised access to social media and revealed police now face the “diabolical scenario” of spending years investigating crimes against kids who don’t exist.

In a speech at the National Press Club on Wednesday, Australian Federal Police Commissioner Reece Kershaw expressed a “growing sense of dissatisfaction with social media companies” and billionaire tech bosses like Elon Musk and Mark Zuckerberg whose decisions are making it exponentially harder to keep kids safe online.

“Many of us grew up at a time when our parents knew we were safe if the front door was locked but now, because of technology, a locked front door is no longer a barrier for criminals,” he said.

“Virtual hands can reach through almost any device, and the consequences of those actions can be as damaging as if they happened in the real world.

“If it used to take a village to raise a child, constant advances in technology now means it takes a country, global law enforcement and the private sector to help keep them safe.”

Mr Kershaw also called out social media bosses for creating “criminal safe havens” on their platforms.

“Numerous law enforcement agencies, including the AFP, have appealed to social media companies and other electronic service providers to work with us to keep our kids safe,” he said.

We’re seeing AI (child abuse material) images using the faces of known, real victims.

“That includes not transitioning to end-to-end encryption until they can ensure their technology protects against online crime rather than enabling it.

“We recognise the role that technologies like end-to-end encryption play in protecting personal data, privacy and cyber security, but there is no absolute right to privacy.

“People have the right to privacy just like they have the right not to be harmed.

“People expect to have their privacy protected just like they expect police to do their job once a crime has been committed against them, or a loved one.

“That expectation includes being able to respond and bring offenders before the justice system.”

In the 2018-2019 financial year, the AFP-led Australian Centre to Counter Child Exploitation received more than 14,000 reports of online child sexual exploitation. In 2022-2023, it received more than 40,000 reports.

“Nearing the end of this financial year, we have already exceeded the past financial year’s figures,” Mr Kershaw said.

“I think the internet, the dark web, has brought more and more of the paedophiles together.

“I think what we’re seeing, sadly, is … that some people have a persuasion for wanting to sexually assault children.

“It’s a real concern and the other piece of that is that they normally re-offend… It’s a crime type that’s quite unique in the sense that I believe you’re hardwired that way and you can’t change.”

In the candid address, he revealed police were staring down the barrel of a “tsunami of AI-generated child abuse material” which will present a “diabolical scenario” for investigators.

A recent report from the UK-based Internet Watch Foundation states that within just six months, AI-generated child exploitation material has become so lifelike that even highly-trained analysts are struggling to determine if the victims are real.

As a result, police could spend capability, resources and years investigating heinous crimes before realising the children they thought were being abused do not actually exist.

In dark web forums, perpetrators have expressed jubilation that their perverted “fantasies can now be made to order” thanks to artificial intelligence.

“All you need is the language to tell the software what you want to see,” Mr Kershaw said.

“We’re seeing AI (child abuse material) images using the faces of known, real victims.

“We’re seeing how technology is nudifying children whose clothed images have been uploaded online for perfectly legitimate reasons.

“Together, we must find solutions to these diabolical challenges but we also must be proactive about our own safety.”

A joint declaration by 32 European police chiefs, agreed to on Sunday, stated that two key capabilities were crucial to supporting online safety.

First, the ability of technology companies to provide law enforcement investigations – in response to a lawful authority with strong safeguards and oversight – the data of suspected criminals on their service. This is known as lawful access.

Second, the ability of technology companies proactively to identify illegal and harmful activity on their platforms.

Police chiefs globally are deeply concerned that end-to-end encryption – which provides criminals with an “instant invisibility cloak” – is being rolled out in a way that will undermine both of these capabilities.

“The AFP echoes concerns raised in this declaration but I want to forewarn criminals that we will still identify them, although we acknowledge some of our investigations may be more resource-intensive,” Mr Kershsaw said.

“I want to warn criminals that despite tech advancements, we can still identify them and bring them to justice.

“As Commissioner, I am focused on ensuring there are no criminal safe havens.”

Meta, which owns Facebook, is about halfway through its end-to-end encryption rollout.

“Our fear is that they may still refer matters to us, but we won’t be able to take action and we won’t know who the end receiver is,” Kershaw said.

“And so that’s going to probably cause more stress in our system of not being able to do anything with it, which I think is a disgrace.”

The career cop, who has worked in child protection, also urged parents to protect their kids’ identities.

“A simple measure is locking down settings on social media accounts to make it harder for others to access images and then use AI to create abuse material,” he said.

“Think about it like this. You probably wouldn’t give a stranger a photo album of your kids. That’s essentially what can happen if privacy settings are not locked down.”

Another growing crime type impacting young people online is financial sexual extortion.

“Criminals, pretending to be someone else, use social media to trick youth into sending intimate images of themselves, and then blackmail them for money,” he said.

“Fearing their images will be sent to loved ones, young people have taken their lives.”

At least five Australian children have recently committed suicide after being ‘sextorted’.

“We need to constantly reinforce that people are not always who they claim to be online; and that also applies to images and information.

“Right now, I know there are silent victims feeling like their world is closing in.

“Please, if this is your situation, do not be embarrassed. Tell your parents, someone you trust and law enforcement.

“We will protect you and make it stop.”

He urged the public to report sextortion and any form of online child sexual exploitation material to the ACCCE www.accce.gov.au/report.

LIFELINE 13 11 14