Inteligencia Artificial (IA)
How to Protect Children from Online Violence and the Challenges of AI
Paloma Firgaira
2025-09-18
5 min read
Promoters of social, educational, and cultural projects demand greater institutional support and funding, while those responsible for aid lines acknowledge the lack of a culture of trust and awareness regarding digital risks.
In Romania, media literacy classes for teachers and families; in Spain, participatory theater; in Ukraine, reporting lines and cyber police. These are just some of the responses that organizations and institutions are implementing in the face of increasing forms of digital violence affecting children and adolescents.
The rise of cyberbullying, exposure to harmful content, hate speech, and online child sexual abuse is alarming. The rapid evolution of artificial intelligence adds new threats, such as deepfakes, which exacerbate the emotional and psychological risks associated with excessive screen use.
According to Eurobarometer Flash 2023, 82% of Europeans believe that parental controls are no longer sufficient to protect minors online. Schools are just beginning to address the complexity of digital violence, while authorities and tech companies move slowly and with little transparency.
In light of this situation, what tools and measures are truly effective? How can we build safer and more supportive digital environments for children and adolescents?
**Persistent Risks and New Threats**
Real cases illustrate the severity of the problem: in Romania, a 12-year-old boy was a victim of sextortion by peers; in Spain, 16 adolescents suffered the dissemination of nude images generated by AI; in Ukraine, a minor was coerced into posting intimate photos on Telegram. These increasingly frequent situations show that digital violence is ongoing and knows no hours.
Early access to devices and the expansion of online time—up to 6 hours daily in some countries—increases exposure to risks. 17% of European minors have experienced online harassment, and in Spain, the prevalence of cyberbullying reaches 57.5%. Exposure to pornography and violent content is also concerning, with serious psychological consequences.
Digital sexual abuse, especially through self-generated images and deepfakes, has become an epidemic. Most victims are minors who rarely receive education on how to protect themselves online.
**Solutions from Culture, Education, and Technology**
In Spain, participatory theater has established itself as an educational and preventive tool. Plays like "Aulas" or "Girls Like That" allow young people to reflect and empathize with victims, complemented by educational materials and subsequent discussions. Paratheatrical techniques and psychodrama, promoted by associations like AEPAE, have proven to improve minors' self-esteem and coping skills.
Other cultural initiatives include the use of series, novels, and guides to foster critical thinking and responsible technology use. Programs like E-tic, Campus Fad, and Te pongo un reto offer practical training on digital risks to students, families, and teachers. However, the lack of curricular integration and institutional support limits their reach.
Artificial intelligence, while part of the problem, can also be an ally. Chatbots like Sof-IA from Fad Juventud and assistants like DEL.IA help young people identify and report situations of digital violence. Additionally, AI literacy is promoted so that minors and adults learn to distinguish between real and manipulated content.
**The Role of Families and Schools**
Many parents feel overwhelmed and fear imposing excessive limits. Projects like the workshops of the Zi de Bine Association in Romania provide concrete tools and "digital detox" camps for adolescents. Media literacy, driven by the Center for Independent Journalism, has trained thousands of teachers and students, although cultural and victimization barriers persist.
In rural areas, World Vision Romania's programs bring digital education to vulnerable communities, although the disciplinary approach and lack of resources hinder mindset change. The role of the school counselor is key, but workload overload limits their effectiveness.
Organizations like Salvați Copiii (Save the Children) in Romania and the Internet Watch Foundation in the UK manage reporting lines and platforms to report abuses, collaborating with authorities to remove content and support victims. In Ukraine, projects like StopCrime and the Safe Internet Center, in partnership with the cyber police, have managed to block domains and apprehend criminals, in addition to conducting educational campaigns and providing resources for parents and teachers.
**Regulation, Prevention, and Future Challenges**
The ban on mobile phones in classrooms is a debated measure. Countries like Portugal and France have implemented restrictions, and in Spain, some autonomous communities have observed a reduction in cyberbullying following their application. However, experts warn that digital education is more effective than a total ban.
Legislatively, the European Union is advancing with the Digital Services Act, which requires large platforms to protect minors and implement age verification systems, although doubts about privacy and the effectiveness of these tools persist. Spain, Romania, and Ukraine have adapted their legal frameworks to criminalize new offenses, such as the creation of child pornography through AI, but practical enforcement remains a challenge.
The digital environment evolves faster than legislation and institutional responses. Cooperation between governments, the private sector, NGOs, and civil society is essential to build a culture of trust, prevention, and support. Digital literacy, emotional education, and the active involvement of young people, families, and teachers are crucial to reducing risks and protecting the most vulnerable.
Ultimately, the fight against digital violence requires a comprehensive approach: education, culture, technology, legislation, and, above all, a committed community to ensure the safety and well-being of children and adolescents in the digital world.