Awareness and Prevalence of Deepfake Cyberbullying: A Study of Polish Students Aged 12‒15

Introduction: One in five teenagers has experienced cyberbullying. However, their families were often not aware of the problem, with 75% of parents believing that their children had not experienced it. Many studies have been devoted to cyberbullying and its devastating psychological consequences, yet few researchers have investigated the potential role of AI in school cyberbullying. This pilot study, part of a larger doctoral dissertation project, aimed to: (1) determine the level of awareness of deepfake cyberbullying among students and teachers; (2) estimate the prevalence of the phenomenon in 12–15-year-olds; and (3) assess their opinions about its consequences and potential countermeasures. Method: The study was conducted in two stages among primary school students in the Trzebnica and Milicz communes. In the first stage, 192 students in grades 7–8 (aged 12–15) were surveyed about their experiences with cyberbullying. In the second stage, an online survey was administered to 36 teachers. Results: A significant number of students (38.5%) and teachers (44.4%) reported firsthand encounters with AI-generated bullying, indicating that a substantial portion of the surveyed population has been exposed to harmful content created with AI. The surveyed students reported that AI was used to alter existing photos (30.2%), generate new images (15.6%), create short videos or other types of animations (17.7%), and generate audio recordings simulating the victim’s voice (15.1%). Moreover, only one-third of students were confident in their ability to recognize such content. Conclusions: These results confirm that AI is already being used for cyberbullying and poses a growing challenge. It provides perpetrators with new audiovisual tools that leverage the same digital media formats popular among students.