What we know about action to crack down on 'sick predators' using AI for child sexual abuse
Promising action, the home secretary says AI is putting the issue 'on steroids' and 'escalating and industrialising' the scale of the problem.
The UK is set to become the first country in the world to make it illegal to own or use artificial intelligence (AI) tools that create child sexual abuse images.
New measures will see people jailed for up to five years for exploiting AI to create this content with even tougher 10-year sentences for people who run websites to share child sexual abuse images.
The government says the aim is for the law to catch up with rapidly developing technology that is being exploited by criminals and abusers.
Here, Yahoo News explains the four new laws being brought in to crack down on AI-assisted online child abuse, and the incoming Online Safety Act.
How is AI being used for child sexual abuse?
Today, according to Technology Secretary Peter Kyle, it is actually legal to possess a "paedophile manual" which gives guidance on how to use AI to sexually abuse children.
While it has been illegal to download or possess so-called "grooming manuals" since 2014, the Internet Watch Foundation (IWF) warned last year that a loophole in the law still allows people to create instructions on how to generate AI "deepfake" images of children.
"Nudeifying" real-life images of children, or stitching their faces on to existing images of abuse, are among the ways AI is being used by abusers, the Home Office said.
In a 2024 report, the IWF said: "AI-generated imagery of child sexual abuse has progressed at such an accelerated rate that the IWF is now seeing the first realistic examples of AI videos depicting the sexual abuse of children.
"Perpetrators can legally download everything they need to generate these images, then can produce as many images as they want - offline, with no opportunity for detection. Various tools exist for improving and editing generated images until they look exactly like the perpetrator wants."
The foundation said most of this AI-generated child sexual abuse material is now realistic enough to be treated as "real", with the most convincing images "visually indistinguishable" even for trained analysts.
It warns of a "noticeable increase" in AI-generated child sexual abuse imagery on the clear web (rather than the dark web) including on commercial sites, and that AI is offering more opportunities to make money off of these images.
Fake images are also being used to blackmail children and force them to livestream further abuse. Ministers believe that this online abuse is encouraging viewers to go out and offend in real life.
How many people are affected by it?
Appearing on the BBC's Sunday With Laura Kuenssberg programme, Home Secretary Yvette Cooper said up to 800,000 people are viewing this kind of material.
While there already is a legal framework in place for online child abuse, Cooper said AI is "now putting online child abuse on steroids" and is "escalating and industrialising" the scale of it.
A report by the IWF in October 2023 found over 20,000 AI-generated images on a dark web forum in one month, with more than 3,000 depicting criminal child sexual abuse activities.
In a subsequent report published in July 2024, the charity found over 3,500 new AI-generated child sexual abuse images had been uploaded onto the same forum.
The charity's latest data shows reports of AI-generated child sexual abuse images have risen by 380%, with 245 confirmed reports in 2024, compared with 51 in 2023. Each of these reports can contain thousands of images.
Cooper told Kuenssberg that half a million children every year experience some form of child sexual abuse, although this also includes abuse carried out offline.
What is the government doing about it?
Under new measures being brought forward in the Crime and Policing Bill, it will be illegal to own artificial intelligence tools designed to make images of child sexual abuse.
This new offence will be punishable by up to five years in prison.
Those who have been found to own AI "paedophile manuals" could be jailed for up to three years.
The Bill will also introduce a specific offence for paedophiles who run websites to share child sex abuse, which could carry a 10-year prison sentence.
New powers will be given to the Border Force to help prevent the spread of child sexual abuse images from abroad, including by allowing officers to call for individuals suspected of posing a risk to children to give up their phones for inspection.
When does the Online Safety Act come into force?
Another piece of legislation, the Online Safety Act, aims to protect both children and adults online by putting a range of new duties on social media companies and search services.
Platforms will be made to implement systems and processes to reduce risks their services are used for illegal activity, and to take down illegal content when it does appear.
They will be required to prevent children from accessing harmful and age-inappropriate content and provide parents and children with clear and accessible ways to report problems online when they do arise.
The act will also compel major platforms to be more transparent about which kinds of potentially harmful content they allow, and give people more control over the types of content they want to see.
Regulator Ofcom will have the power to fine firms who don't follow the rules up to £18m or 10% of their qualifying global turnover - whichever is greater - and in very serious cases can apply for sites to be blocked in the UK.
The Act passed into law on 26 October 2023 and is set to be implemented from 17 March this year.