Legal experts have reported a “significant spike” in concerns from schools about the risk that pupil photos on their websites could be used to create AI deepfake content. In May, the Internet Watch Foundation (IWF) reported the attempted blackmail of a UK school using website images that had been turned into child sexual abuse material. Browne Jacobson, which provides legal advice to more than 1,300 schools across the UK, said this had led to experts being “absolutely inundated” with requests from schools for more information about the issue. The Department for Education (DfE) quietly issued new guidance on the “significant and evolving” matter last month, and said leaders should consider limiting public visibility of images of their pupils. Claire Archibald, legal director specialising in AI at Browne Jacobson, said there should not be a “sweeping removal of all imagery”. But schools should “ensure there is genuine strategic thinking, rather than habit, behind how they use pupil photographs in the digital space”, she added. This month, the IWF issued a report on the rising threat following the experiences of one UK secondary school, which has not been named. Blackmailers sent the abuse images to the school and threatened to publish them online if they did not receive money. The IWF – part of the UK online harms early warnings working group – used a digital tool to prevent around 150 of the blackmail images from being uploaded. What can schools do? In a recently released training module on the safe use of AI in schools, the DfE said: “Automated tools can harvest photographs from public online platforms, such as school websites or social media. “These scraped images can then be used without consent, with the possibility of them being used for the creation of AI-generated media or deepfake abuse materials, sometimes leading to extortion, posing a threat to child protection. “Education leaders should consider their approach to publicly sharing images, considering the opinions of the whole school community, and staff should ensure they follow guidance issued by their organisation.” Guidance from the UK online harms early warning working group said schools can safeguard pupils by ensuring that images do not contain “identifiable information” which could be used to harm or blackmail an individual. Schools should use images that are harder to misuse or abuse, for example those taken from a distance or blurred, the guidance said. Leaders could also apply privacy settings and remove metadata, as well as train staff on image security awareness. Schools take action Archibald said the warnings from the IWF had led to a “significant spike in schools enquiring about how to protect themselves”, though the firm has not yet had to support any schools at the receiving end of a threat. She said that, “rather than reacting hastily, school leaders would be well served by treating this moment as an opportunity to establish, or revisit, a clear strategy for the use of pupil imagery across all external-facing materials”. While Archibald did not call for the “sweeping removal of all imagery”, she said schools should “ensure there is genuine strategic thinking, rather than habit, behind how they use pupil photographs in the digital space”. Helen Masters, director of SEND and safeguarding at Chiltern Learning Trust said in the DfE’s AI training module that her trust had “considered very carefully” what it is doing with images, why it is using them, and which platforms they are being posted on. Lewis Duncan, marketing officer at Cedars Upper School in Bedfordshire, said that for images of younger pupils they “tend to just do work-focused shots from behind, or over-shoulder shots of them doing those activities”. The UK online harms early working group said schools should always report threats to publish sexual imagery of children to the police, as the creation of such material is illegal regardless of whether it is AI-generated or not. Jess Phillips, the former minister for safeguarding and violence against women and girls, said at the time of the report that the attempted blackmailing of schools was a “deeply worrying emerging threat”, and that laws on the use of AI to create explicit images would be updated if necessary. Schools Week has approached the Home Office and DfE for more details.