Quarter of teachers report pupils use AI to generate sexual images of classmates

Almost one in four secondary teachers say a pupil at their school has created an AI-generated sexual image of a classmate, a survey suggests.

From September, new keeping children safe in education guidance says schools must treat incidents involving AI-generated sexual images of children as a safeguarding incident. It brings such sexual “deepfakes” in line with the rules for real photographs.

The recent crime and policing act also states that distributing and possessing AI-manipulated child sexual abuse imagery is now a criminal offence.

Teacher Tapp data shared with Schools Week shows 23 per cent of secondary teachers polled were aware of a pupil creating a sexual image or video of another, with 19 per cent reporting the image was shared with other classmates.

Twelve per cent of secondary teachers also said pupils had targeted a member of staff.

But the issue is not restricted to secondary schools, the data suggests.

Three per cent of primary teachers polled said a pupil had targeted another pupil, while 2 per cent reported a staff member had been targeted.

Analysis by the Internet Watch Foundation found 3,440 AI-generated videos of child sexual abuse in 2025, compared with 13 in 2024.

Dai Durbridge, a partner specialising in safeguarding at the legal firm Browne Jacobson, said pupils creating AI-manipulated content would become more prevalent as it became easier for children to access free or low-cost AI tools.

‘Consistent message’

Leaders at the Chiltern Learning Trust have become aware of an “increasing number of cases” of pupils creating AI-images of each other.

While teaching about the dangers of AI is required from September in the new relationships and sexual health education guidance, the trust is looking to embed learning across its school curriculum, in lessons such as science, computing and technology.

Christian Turton, the trust’s edtech lead, said: “We’re trying to triangulate those different bits together to get a consistent message about things.

“We’re taking it very seriously…we just need to be prepared to do the best we can to support young people.”

Durbridge said the “dangerous trend” could cause serious behavioural issues among offending pupils and their victims who could face “emotional distress due to bullying”.

Pepe Di’Iasio, the general secretary of the Association of School and College Leaders, said incidents could be distressing for everyone involved.

“While these issues can be challenging to deal with, schools will have robust safeguarding in place and be teaching students about the harms and risks of misusing social media and other online tools.

“We would like to see the strongest possible legislative action taken against the providers of any apps used for this purpose.”

Staff images

School staff are also finding themselves the victims of non-sexual deepfake images created by pupils.

Thomas Michael, a deputy designated safeguarding lead in the West Midlands, discovered a deepfake video of him earlier this year.

His school had been experiencing a significant number of AI-generated “memes” that were posted on an anonymous TikTok account.

Pupils, Michael claimed, would send in “requests“ for posts. The account would be deactivated once teachers became aware, only to be re-created under a new name.

The video showed him at a wedding throwing a bunch of flowers, only for no woman to catch it. The caption suggested he was instead “talking to year 12 girls”.

Michael said pupils made other videos about teachers that were “antisemitic in nature”, or “accusatory in terms of conduct”.

“Depending on the nature of what it was, some can just brush it off. Others have been upset by it. It’s more not knowing who it is, so then you’re quite wary about what you’re doing, what you’re saying, who you’re interacting with.”

The school leaders’ union NAHT passed a motion at its 2025 conference calling for the creation and distribution of such images to be criminalised.

The motion said teachers were “increasingly vulnerable” to AI manipulation, and that those affected should receive “immediate access to legal assistance, mental health support and crisis intervention services”.

School images at risk

Criminals outside school communities could also pose a risk.

The UK online harms working group recently urged schools to consider removing identifiable pictures of pupils from their websites and social media accounts.

According to the Internet Watch Foundation, criminals blackmailed a secondary school when they “scraped’ images off its website to create child sexual abuse material.

The National Crime Agency has also encouraged parents to limit their child’s presence on their social media.