Artificial intelligence (AI) is driving a surge in cyberbullying, acting as a catalyst for online violence against women, according to Belgium’s Institute for Equality between Women and Men (IEFH) on International Women’s Day.
The institute highlighted the rise of “deepnudes” and “deepfakes,” hyper-realistic fake images that are often sexualised, as the most visible forms of this trend. Generative AI tools now make it possible to create such content in just a few clicks using profile pictures from social media.
These technologies primarily target women, particularly those active in public spheres, including politicians, journalists, and experts. IEFH spokesperson Véronique De Baets warned that such harassment could silence women’s voices in democratic debate, as many victims resort to self-censorship or withdraw from online platforms and public life.
Women who have previously faced online sexism remain the most targeted—especially those vocal on polarising topics or gender equality issues. According to De Baets, AI amplifies the scale, speed, and intensity of attacks, enabling them to be created and shared en masse.
Young people are also highly affected. A 2024 IEFH report revealed that one in five youths had encountered fake, intimate images generated by AI. De Baets believes this figure has likely increased since the data was collected.
Beyond deepfakes and deepnudes, AI is facilitating new forms of digital violence, including intimate partner tracking, automated smear campaigns, and sextortion using artificially created images. While sexually explicit deepfakes tend to target women, other crimes, such as sextortion, disproportionately affect men.
Gender-based violence online impacts women, LGBTQIA+ individuals, young adults, and other vulnerable groups, particularly those active on social media, the institute stated. Reports of online sexual violence received by the IEFH—including non-consensual sharing of intimate images, sextortion, or voyeurism—rose by 25% in 2024 and 75% in 2025.
The institute noted, however, that the true extent of the problem remains unclear due to significant underreporting. Victims often refrain from lodging complaints due to shame, fear of disbelief, or technical challenges, such as proving an AI-generated image is fake, tracking anonymous perpetrators, and dealing with content hosted on overseas servers.
Although legal frameworks have improved in recent years, effective implementation presents ongoing challenges. De Baets called for increased resources, enhanced psychological and legal support for victims, and stronger preventative measures. She also emphasised the role of sex and relationship education programmes in raising awareness among young people about online consent and AI-generated content risks.
The IEFH recommended integrating gender considerations into AI development to mitigate the reproduction of gender biases. A February study by the institute revealed that 74% of Belgian recruiters use AI tools in hiring processes, often unaware of the risks of discrimination.
To address digital violence, De Baets advocated for specialised support centres offering psychological, legal, technical, and police assistance to victims, similar to existing sexual violence response centres.
