“There are no guardrails”, says Karen Davila about popular social media platforms. “AI, like anything, can be used for good – but it can also be abused.”
Davila, an award-winning broadcast journalist and UN Women National Goodwill Ambassador for the Philippines, describes her experience coming face-to-anonymous-faces with online abusers in sophisticated – and disturbing – digital abuse campaigns.
UN Women National Goodwill Ambassador Karen Davila moderates a panel during the 70th Session of the Commission on the Status of Women (CSW70) at United Nations Headquarters. March 2026. Photo: UN Women/Ryan Brown
Immediately following her moderation of a 2016 presidential debate in the Philippines, in a deliberate campaign to bully the press, tens of thousands of vulgar and threatening comments appeared in Davila’s social media feeds – every hour. Although it was clear the commenters were bots, the misogynistic attacks were designed by humans to damage her credibility and her sense of safety.
The campaign of online violence was a message to the country and to journalists. “They want to push you to silence so that you don’t criticize the actions of the administration”, says Davila.
“What’s important to me is my integrity, something I’ve fought for all my life”, she reflects. “Being a journalist, you’re built tough. But you start to question your own sense of integrity, because that’s what trolling and harassment make you do.”
UN Women is tracking the impacts of digital abuse
Cases like Davila’s have a potential chilling effect on women pursuing careers in journalism – and on freedom of the press itself. UN Women is working with gender equality advocates like Davila to expose the reality of online harassment and the cost it exacts on women in media and the public sphere.
In our latest Tipping Point report, UN Women, TheNerve and partners uncover the growing and increasingly sophisticated forms of online violence faced by women in the public sphere.
In a survey for the report, 45 per cent of women journalists and media professionals reported self-censoring on social media to avoid abuse – a 50 per cent increase since 2020. And almost 22 per cent say they are self-censoring at work.
Generative AI has increased the speed, scale and anonymity of online attacks
“Journalists have always been targeted”, explains Francesca Donner, founder and editor of The Persistent. “The job of a journalist is to uncover, reveal, go deep into uncomfortable places, and put that out there for all the world to see. And certain people don’t want that.”
“Women journalists have been targeted in very specific ways. And whether you’re a journalist, a politician, an activist, or anyone that’s visible, the motivation [of the attacks] is always the same: Get them to stop writing, talking, running for office, doing their activist work. The goal is silence”, says Donner.
With deepfake photos, videos and nudification apps, generative AI has taken online violence from text comments and memes a few years ago, to ultra-realistic photos and videos today. “All a bad actor needs is a photo”, says Donner. “[AI-assisted attacks] are virtually impossible to put back in the bottle – and this has always been the way with online violence.”
As Donner explains, when these online attacks start, “It’s like someone lighting a match and trying to get others to pile on. It can seem so minor in the grand scheme of things, but it’s done with ill will, with the intent to troll.”
And then it escalates: “The unpleasant and unkind things said; violent things said; your photo posted; your address posted; [images of] your children, your family, all of the things that people can possibly go after. It spills over into real life.”
