Online violence against women journalists and public figures has surged in recent years, with AI tools increasingly aiding gendered abuse, according to multiple UN reports and independent research. Since 2020, reports of online violence against women journalists have doubled, while AI-assisted attacks contribute to a growing pattern of harassment, deepfakes, and manipulated sexual imagery targeting women.
UN Women and partners revealed in April 2026 that nearly one in four women human rights defenders, activists, and journalists surveyed in 2025 suffered AI-assisted online violence. The surge is backed by UNESCO’s global survey showing that 75% of women journalists experienced online abuse last year, with 42% linking these attacks to offline harassment and assault.
Research by the Oxford Internet Institute in May 2025 found that 96% of publicly downloadable deepfake image generators targeted identifiable women. These AI-powered tools make abusive imagery more accessible and destructive, intensifying threats to women in public life.
A UN Women statement quoted by UN News emphasised that “AI is making abuse easier and more damaging, and this is fuelling the erosion of hard-won rights in a context marked by democratic backsliding and networked misogyny.” This highlights the serious democratic and social consequences of AI-enabled harassment beyond individual harm.
UNESCO cautioned that AI applications like nudification, manipulated sexual images, and deepfake content are creating invasive new risks for women journalists. This reflects a disturbing trend where technology compounds longstanding patterns of gendered violence online.
The rise in reports to police of online violence against women journalists illustrates the increasing scale of the problem. In 2026, authorities saw a doubling of such complaints compared to 2020, underscoring the urgent need for effective measures to combat AI-driven abuse.
Experts warn of a chilling effect on free expression, as 45% of women journalists surveyed in 2025 admitted to self-censoring on social media to avoid harassment. This curtails diverse voices in media and public discourse, deepening inequalities.
While these findings come from global data, the implications resonate in South Africa’s media and public spheres, where gendered violence remains a pressing issue. AI’s role in automating misogyny presents a new layer of challenge for rights defenders, law enforcement, and digital platforms alike.
Further investigation is needed to track how AI-driven abuse spreads locally and to devise targeted interventions. Monitoring developments from SAPS, social media companies, and women’s organisations will be crucial to understanding and addressing this evolving threat.