News

The deepfake image threat: Schools urged to audit their websites and social media use

Safeguarding Artificial intelligence Pastoral issues
Schools are being reminded about their duties to protect students and staff from the deepfake manipulation of images and videos used on school websites or social media feeds.
Emerging threat: The Internet Watch Foundation has documented the horrifying ways that AI is being used to create deepfake images and videos of child sexual abuse using real images trawled from the web - Adobe Stock

The rapid development of artificial intelligence technologies has given rise to the creation of deepfake images and videos, often made using real images found online.

In 2023 and 2024, reports from the Internet Watch Foundation documented horrifying ways that AI is being used to create deepfake images – and now videos – of child sexual abuse using real images trawled from the web. The research uncovered thousands of images of abuse on just one single dark web forum.

Given that many schools use images of pupils on both school websites and social media feeds, new guidance published by the UK Safer Internet Centre is seeking to help schools to manage this emerging and increasing threat and to ensure the proper management of image and video use online.

Social media platforms in particular can be vulnerable to “web-scraping”, which is used to collect public content at scale.

The guidance covers key areas including auditing online content, the appropriate use of student images in line with data protection policies, and restricting public access.

When it comes to student images, it advises schools to apply a “names-no-images and images-no-names” policy. It also recommends removing metadata and using lower resolution files to make manipulation more difficult.

The emerging threat of deepfakes has become a key safeguarding issue for schools. In a recent article discussing 10 ways schools can protect children from this phenomenon, Headteacher Update’s resident safeguarding writer Elizabeth Rose said there was a clear need for CPD for school staff on the deepfake threat and she urges schools to address the issue with students in the curriculum, too.

The recent government AI guidance for schools is lacking when it comes to this issue, focusing more on use of AI in resource-creation in the classroom.

Separate government guidance – entitled Sharing nudes and semi-nudes: Advice for education settings working with children and young people (DfE, 2024) – was updated in March 2024 to include greater reference to deepfakes, with overarching advice about responding in cases where these kinds of images have been shared.

The UK Safer Internet Centre guidance does acknowledge that schools are “typically strong in their approach to managing images and videos” but says there is no room for complacency.

The guidance states: “The misuse of publicly shared images and videos presents risks such as data privacy breaches and unauthorised distribution. Schools and colleges must remain vigilant in their approach to managing media content online, particularly on publicly accessible platforms such as school websites and social media accounts.

“Any image or video content shared on a social media platform set to public can be accessed by anyone. Schools should exercise great caution sharing images or videos of learners publicly on social media platforms due to the potential risk of the content being misused. Social media platforms can be vulnerable to web-scraping which facilitates public content to be collected at scale. Loss of control of images, for example, can expose learners to potential data and privacy risks.”

The guidance also urges staff training in “secure image management” and says schools should have clear guidelines for the use of student and staff images.

It adds: “Schools should engage with parents and carers to communicate safeguarding policies and set expectations for responsible image sharing.”