Australian schools warned photos they post are…

Australian schools have been issued an urgent warning regarding the risks of posting photographs of students and staff online, as these images are increasingly being harvested for AI manipulation. The eSafety Commissioner reported a significant rise in incidents where school-sourced imagery was utilized to create harmful deepfake content, including child exploitation material.

What happened

Between January and March 2024, the eSafety Commissioner received over 100 reports involving anonymous accounts targeting schools and staff. These incidents involved the use of AI to create face swaps, memes, fabricated stories about teachers and principals, and AI-generated dance videos. In the most severe cases, imagery was manipulated to remove clothing, resulting in the creation of child exploitation material. The source of this content is almost exclusively imagery harvested directly from school websites and social media pages.

In response to the threat, a new advisory released on May 14, 2024, urges schools to implement stronger parental consent protocols and review existing online content. Cyber safety educator Paul Litherland has advised schools to reconsider the volume of content posted and the identifiability of students in those images. Some educational institutions have already begun shifting their practices, such as opting not to show faces in photographs, utilizing AI to generate fictional pupils for advertisements, and transitioning school newsletters to password-protected digital formats or physical printouts. Additionally, the eSafety Commissioner has reminded social media and AI companies of their legal obligations, noting that firms can face fines of up to $54 million if they fail to adequately prevent or address the presence of illegal AI-generated content on their platforms.