The amount of AI-generated child sexual abuse material found online rose by 14% last year, with the Internet Watch Foundation (IWF) identifying 8,029 AI-made images and videos of realistic child sexual abuse material (CSAM) in 2025. The figures underpin new guidance issued jointly by the National Crime Agency (NCA) and the IWF, warning parents and carers to stop putting photographs of their children on public display online.
AI-generated child sex abuse material surges 14% in a year
The rise in AI-generated CSAM was most acute in video content. The IWF recorded 3,440 AI-generated child sexual abuse videos in 2025, compared with just 13 in 2024 – an increase of more than 260-fold. A significant proportion of these videos fell into Category A, the most severe legal classification under UK law, which encompasses offences including rape, sexual torture and bestiality. In 2025, 65% of AI-generated CSAM videos were Category A, compared with 43% of non-AI criminal videos.
AI tools are becoming more powerful, widely available and easier to use, enabling offenders to create highly realistic CSAM with minimal technical expertise, the NCA said. The realism makes it difficult to distinguish AI-generated material from genuine abuse imagery. Existing CSAM can also be used to train AI models, leading to the ongoing revictimisation of children who have already suffered abuse.
Parents urged to lock down online images
The NCA and the IWF have issued landmark guidance recommending that parents and guardians make their social media accounts private or share pictures of their children only through a “close friends” group. The guidance also advises auditing social media accounts for old pictures that could be used by predators, and revisiting photo consent agreements – for instance with schools or sports clubs – that may have been signed before recent breakthroughs in AI made image manipulation possible.
“We encourage parents and carers to take a few simple steps today,” said Tim Wright, a senior manager at the NCA. The guidance sets out a trio of actions: checking privacy settings on social media accounts; reviewing who can see images of children; and having open discussions about giving permission for people and organisations to publish images online. Lorna Sinclair, a child sexual abuse education manager at the NCA, said: “The average parent or carer does not post a picture of a child online thinking that it might be scraped to be turned into CSAM. There are lots of parents and carers who do not know that this problem exists.”
The guidance follows cases where UK school websites were targeted by blackmailers who scraped pictures of children, used AI tools to convert them into child sexual abuse material, and then threatened to publish the results. A UK advisory body on tackling online harms, the early warning working group (EWWG), whose members include the NCA and IWF, has recommended schools remove identifiable pictures of pupils’ faces from their websites and social media accounts. The children’s charity the NSPCC also recommends that under-18s keep their social media accounts on a private setting.
Videos released as part of the guidance show fictional scenarios of parents taking photos of their children in everyday circumstances, such as playing sport or standing at the school gates, and being reminded about the risks of sharing photos online. The NCA and IWF say they want to encourage parents and children to say “no” to sharing photos online if they are uncomfortable about it. The guidance on carrying out a social media audit includes users checking their own accounts to gauge whether a child’s “face, body or school uniform” can be seen, whether they are comfortable with the image still being online, and whether it can be deleted or made private. It also recommends checking whether friends or family have uploaded images of a parent’s child, including historic posts, and discussing “clearly and calmly” what action could be taken with those posts. The guidance also recommends reviewing consent forms signed by parents at school, nursery or clubs giving permission to use a child’s image, and considering whether they want to withdraw that consent. The IWF’s head of marketing, Tom Dyson, said: “If you want a photograph of your children to be taken off a website or social media, you are perfectly able to do that.”
How AI enables a new wave of abuse
The NCA said most parents and carers would not be aware that advances in AI technology had given criminals publicly available tools to create CSAM without needing to contact – or “groom” – victims directly. Instead, offenders can scrape innocent images from social media, school websites and other public sources, then use AI to manipulate them into extreme abuse material.
AI is being exploited in several ways. “Nudify” apps can digitally remove clothing from existing photos or videos of children. Deepfake technology can place a child’s face onto an adult’s body, or manipulate ordinary, fully clothed selfies into sexually explicit content. Generative AI can also create entirely new, realistic images and videos of children in abusive scenarios, sometimes depicting children who do not exist but trained on actual abuse imagery – meaning real victims are revictimised when their images are used to train AI models. The IWF’s confidential service Report Remove, which removes explicit images of under-18s taken without consent, has reported examples of normal selfies being converted into extreme pornography via AI.
In one case heard by the Childline service, a 15-year-old girl said a stranger had made a “really convincing” fake nude of her that used her face and bedroom, having apparently taken the source material from her Instagram account. The IWF has also been contacted by under-18s who have been blackmailed by extortionists after their images were “nudified” by AI.
Under UK law, creating, possessing or distributing CSAM – including AI-generated material – is already illegal under the Protection of Children Act 1978 and the Coroners and Justice Act 2009. New legislation being introduced through the Crime and Policing Bill will criminalise AI models specifically optimised to create CSAM and ban manuals instructing offenders how to use AI for abuse. The UK has become the first country to ban AI tools designed explicitly for generating CSAM. The Online Safety Act is also being used to combat the threat.
Dan Sexton, the IWF’s chief technology officer, said he was “very uncomfortable” about telling parents not to put pictures of children on public display but felt there was no other option. “I don’t know what else to say to parents,” he said. “I would be very cautious [about putting pictures of children online] because there is no protection.”
