UK Issues Urgent AI Safety Guidance as Threat to Children's Images Grows

Parents and carers across the UK are being urged to rethink how they share photos of their children online, following new guidance from the National Crime Agency (NCA) and the Internet Watch Foundation (IWF) warning that everyday family pictures are increasingly being stolen and manipulated using artificial intelligence.

A Rapidly Growing Problem

The scale of the issue has escalated sharply. <cite index="1-1">The IWF identified 3,440 AI-generated videos of child sexual abuse material in 2025, compared with just 13 the year before</cite>, and <cite index="6-1">the organization has also reported a 14 percent rise in AI-generated abuse material overall, with more than 8,000 confirmed images and videos found in 2025 alone</cite>.

Officials say the tools behind this surge are becoming both more capable and more accessible. <cite index="4-1">According to the NCA, criminals are now able to turn ordinary photographs into explicit content without ever making contact with or grooming a child directly</cite>.

How the Images Are Being Exploited

Investigators have documented specific cases illustrating the threat. In one instance, <cite index="1-1">a criminal gang pulled images of pupils from a UK school's website and used AI to generate more than 100 sexualized images of the children, then attempted to blackmail the school to prevent the material from being published</cite>. The IWF says it acted quickly to stop those particular images from being uploaded, though concern remains that other schools could be targeted in similar ways.

Officials note that the source material for these fakes is often mundane. <cite index="7-1">Agencies have pointed out that even fully clothed children shown in completely non-sexual settings can be digitally "nudified" or otherwise altered into abusive content</cite>, meaning the risk isn't limited to any particular type of photo.

What the Guidance Recommends

Rather than telling parents to stop sharing pictures of their children altogether, <cite index="4-1">the guidance stops short of that and instead stresses that greater awareness is essential as AI-driven image manipulation grows more sophisticated</cite>. The core recommendations are straightforward:

• Review privacy settings on social media accounts to control who can see photos of children.
• Limit sharing to trusted circles, such as a "close friends" list, rather than posting publicly.
• Audit older posts and other accounts — including those of relatives or friends — where images of a child may already be visible.
• Talk openly with children about AI, deepfakes, and image consent, so they understand the risks and know what to do if they're targeted.
• Revisit consent agreements with schools, nurseries, and clubs, many of which were signed before AI manipulation tools became so widely available.

Why Awareness Is the Focus

Officials involved in the campaign say the biggest obstacle isn't a lack of available safety tools — it's that many parents simply don't realize the risk exists. <cite index="4-1">One NCA official noted that most parents do not upload family photos believing they could later be collected and transformed into abusive content, and that public awareness of the threat remains limited</cite>.

The IWF has acknowledged the discomfort inherent in this message, effectively asking families to weigh the risks of a practice most consider completely normal. The organization has framed it as wanting to give parents an informed choice rather than a blanket rule against sharing.

A Broader Pattern

This guidance builds on earlier efforts. <cite index="1-1">It follows similar advice the IWF and NCA gave to education professionals the previous year, aimed at helping schools protect student images from AI manipulation</cite>, and a separate practitioner-facing guide has already reached tens of thousands of professionals who work with children.

If a Child Is Targeted

The agencies have also issued guidance for what to do if a family discovers a child has already been targeted. <cite index="3-1">Recommended steps include not paying any blackmail demands, avoiding contact with whoever is making threats, preserving evidence such as messages and images rather than deleting it, and reporting the incident to police or, for under-18s, directly to the NCA's CEOP Safety Centre</cite>. Reports can also be made to the platform where the material appeared, and to Report Remove, a joint service run by Childline and the IWF.

This article is based on publicly available guidance issued by the UK's National Crime Agency and Internet Watch Foundation. Readers can find the full parent guide at iwf.org.uk.