AI image generation has created a genuinely new safeguarding risk, and it's one DSLs need to understand before an incident rather than during one. Here's what matters and how to respond.
The risk in plain terms
Free tools can now generate convincing images of real people from a single photograph. In a school context that means a pupil's face — taken from social media, a school website or a classmate's phone — can be used to create fake images, including sexual ones. This is happening in UK schools, and pupils are both perpetrators and victims.
The most important point for DSLs: a deepfake nude of a child is child sexual abuse material. It does not matter that the image is synthetic, that "no real photo was taken", or that a pupil created it "as a joke". Treat it with exactly the same seriousness as any other CSAM incident, and follow your existing child protection procedures.
What DSLs should understand
- The barrier to creating these is now very low — no technical skill and no money required
- Victims experience it as real — the distress is not lessened by the image being fake
- It spreads like any other image — screenshots, group chats, and beyond your reach quickly
- Pupils often don't grasp the seriousness — many genuinely believe a fake image isn't a real offence
- It disproportionately targets girls, as with other forms of image-based abuse
Responding to an incident
- Follow child protection procedures — this is a safeguarding incident, not an IT one
- Do not ask to see, forward or store the image — follow your usual protocol for indecent images
- Support the victim first — their distress is real regardless of the image's origin
- Refer as you would for any CSAM incident, including to police where appropriate
- Consider the perpetrator's needs too — often a child who has not understood what they've done
- Record thoroughly and review what the incident says about your prevention work
Your existing KCSiE-aligned procedures cover this — see updating your KCSiE policy for AI.
Prevention
Education is the main lever. Pupils need to understand that creating these images is a criminal offence, that it causes real harm, and that "it wasn't a real photo" is no defence. This belongs in your RSHE and online safety curriculum now, not next year. Staff also need briefing so they respond correctly rather than improvising.
Talk to parents about it
Parents are often unaware this is possible. A calm, factual message explaining the risk, encouraging them to talk to their children, and making clear how the school will respond does a great deal — both preventatively and for trust. See communicating with parents about AI.
Don't wait for an incident
The schools that handle this best have already briefed staff, taught pupils, told parents and checked their procedures cover it. That preparation takes an hour or two; managing an unprepared incident takes weeks.
Need staff briefed on AI safeguarding risks?
We run sessions that get DSLs and staff ready before an incident, not after. AskColin gives schools low-cost monthly support with one-to-one help whenever you need it — so when you're stuck on a real job at 8am, there's someone to ask. Practical, jargon-free, built around your team.
Get one-to-one AI support for your schoolFrequently asked questions
Is a deepfake nude of a child illegal?
Yes — a synthetic sexual image of a child is child sexual abuse material. It doesn't matter that the image is AI-generated, that no real photograph was taken, or that a pupil created it as a joke. Treat it with the same seriousness as any other CSAM incident.
How should a DSL respond to an AI-generated image incident?
Follow child protection procedures — this is a safeguarding incident, not an IT one. Don't ask to see, forward or store the image, support the victim first, refer as you would for any CSAM incident, consider the perpetrator's needs, and record thoroughly.
Why do pupils create deepfake images?
Often because they genuinely don't understand the seriousness — many believe a fake image isn't a real offence. The barrier to creating them is now very low, requiring no technical skill or money, which is why education about the criminal and personal consequences matters urgently.
How can schools prevent AI image abuse?
Education is the main lever — pupils need to understand it's a criminal offence causing real harm, and that 'it wasn't a real photo' is no defence. Brief staff so they respond correctly, teach it in RSHE and online safety now, and tell parents proactively.