This story written by Jetlir Buja originally appeared on Global Voices on August 20, 2026.

In May 2026, I launched the first phase of a project aimed at understanding how sexist hate speech operates in Kosovo’s online public sphere with the help of AI tools. While discussions about online violence against women have become increasingly common, there is still limited local research examining the actual language used in these attacks and the social narratives they reproduce. 

The project was motivated by a simple but important question: What role does sexist hate speech play in public discourse? Is it merely a collection of offensive comments posted by anonymous users, or does it serve a broader discursive function? 

Understanding these dynamics is important because online discourse increasingly shapes public attitudes, political participation, and perceptions of who belongs in public life. If misogynistic narratives become normalized online, they can discourage women from participating in public debate and reinforce existing inequalities.

See also: Could Social Media Affect Who We Are?

A disturbing picture

After putting together a team to facilitate the project, we began monitoring social media comments published on 22 major Kosovo-based Facebook and YouTube pages. We collected and analyzed 625 comments across 67 social media posts published between May 1 and 31, 2026. 

These posts were made by well-established media outlets that report on national news. 

The monitoring period coincided with the local election campaign and a period of intensive political debate online.

The findings paint a disturbing picture of how women are discussed in Kosovo’s digital public sphere — one that is even more disturbing than many might imagine.  

Delegitimization of women

The most common form of sexism was the systematic delegitimization of women in public life. Two-thirds of all sexist comments sought to undermine women’s participation in politics, media, or public debate. 

Women were routinely mocked for their appearance, intelligence, profession, or gender rather than criticized for their views or actions. 

A striking linguistic pattern emerged across the dataset. Women were frequently described through animal metaphors and degrading labels. Terms such as “lopë” (cow), “sorrë” (crow), “laraskë” (magpie), and “shtrigë” (witch) appeared repeatedly. In other cases, women were reduced to sexualized insults such as “kurvë” (whore/slut), “lavire” (slut/prostitute), or “rrospi” (bitch). 

The purpose of these comments was not merely to insult. They functioned as a way of stripping women of legitimacy and credibility in public spaces. 

Another recurring pattern concerns cleanliness. Sexist commenters portray women through the lens of contamination, depicting them as physically and/or morally dirty. 

The hostility was not limited to women who voluntarily participated in the public debate.

Among the monitored comments were attacks directed at women who had been reported as victims of domestic violence. Rather than expressing solidarity with victims, many commenters justified violence, blamed women for provoking abuse, or argued that men had a right to “discipline” their wives through physical or emotional abuse. 

In these discussions, violence was often framed not as a crime but as a consequence of women failing to comply with traditional expectations.

Equally disturbing was the treatment of women who had been reported missing. In several cases, commenters responded to reports of missing women not with concern or empathy but with speculation, ridicule, and sexualized language. The fact that a woman was potentially in danger did not shield her from misogynistic abuse.

The monitoring also documented a case in which sexist and sexualized language was directed at a minor girl. This finding is significant because it challenges a common assumption about online misogyny. 

Women who enter politics, journalism, or other public-facing professions are often expected to face disproportionate levels of abuse. However, the dataset suggests that sexist hostility can affect virtually any woman or girl who becomes visible in the online public sphere.

This clearly shows that the underlying driver of many sexist comments is not opposition to particular views or actions, but hostility directed at women and girls as a social group. 

Whether a woman appears in the public sphere as a politician, a journalist, a victim of violence, a missing person, or even a child, she can become the target of the same degrading narratives, stereotypes, and forms of verbal abuse.

Recurring narratives

The dataset reveals that misogyny in Kosovo’s social media environment operates through several recurring narratives.

First, women are portrayed as inherently less competent and less deserving of participation in public life. 

Second, traditional gender roles are presented as natural and unquestionable, with women expected to remain primarily within domestic and caregiving roles. 

Third, women are judged through standards of sexual morality that are rarely applied to men. 

Finally, violence and male authority are often normalized through language that portrays women as responsible for the abuse they experience.

See also: Social Media Overtakes Traditional Media as Top News Source, Reuters Study Finds

Against this backdrop, anti-sexist voices were rare but significant. They actually exist — and that is the good news that emerges from this sample. Of the 625 comments opposing sexism, 32 tended to focus on condemning violence against women, challenging double standards, and calling for respect and equality. 

Their language differed markedly from that of sexist commenters. Whereas misogynistic comments relied heavily on insults, dehumanization, and ridicule, anti-sexist comments invoked concepts such as dignity, fairness, and human rights.

This article represents the first phase of a broader research effort. Through manual analysis, the research has identified recurring forms of misogynistic discourse, including the delegitimization of women in public life, the normalization of gender stereotypes, the sexualization of women, and the justification of violence. 

The most important lesson from this phase is the value of actively monitoring sexist hate speech in the online public sphere. Without systematic observation, many of these patterns remain invisible. 

At the same time, monitoring is not only a research exercise. It can also serve a public-interest function. Throughout the project, we encountered comments containing threats, the justification of violence, and sexualized language directed at vulnerable groups. 

Such cases highlight the need for stronger moderation practices. Indeed, one of the findings of this project is that many of the monitored pages appeared to exercise little or no moderation over sexist and abusive comments. They also highlight the need for more responsible responses from media organizations that host public discussions on their platforms.

[KS]

Suggested reading: