When your own face becomes the weapon: Kenyan women confront online abuse

This story has significance for readers across Kenya and beyond.

What you need to know:

- Kenyan women are warning that artificial intelligence is making online gender-based violence more sophisticated, exposing victims to deepfakes, impersonation and cyber harassment.

- Survivors of technology-facilitated gender-based violence are calling for stronger digital safety laws and tougher action against perpetrators exploiting artificial intelligence.

A photograph can build a career. It can also be stolen and turned into something its owner never agreed to, and for a growing number of women, that risk is real. Experts call it technology-facilitated gender-based violence (TFGBV), and as artificial intelligence (AI) makes manipulation both easier to pull off and harder to catch, more women are finding their identities, images and voices used against them without ever giving consent. Valerie Kulola and Millicent Atieno have both lived through it, and have chosen to speak out.

Every photograph Valerie posted online represented a dream taking shape. As a young commercial model, each image built her portfolio, expanded her reach and opened doors to new opportunities. She was only 19 when those same photos were turned into weapons by people she had never met.

Someone stole her identity and used it to set up a fake Tinder profile. Not long after, her face was lifted and altered into an AI-generated romantic video that hundreds of people ended up watching, without her ever knowing it existed.

What started as a chase for opportunity became a hard lesson on TFGBV, though Valerie refused to let it push her off the internet altogether. She fought back instead, and now speaks out to educate others, push back against destructive online behaviour, and campaign for digital spaces where women and girls can thrive without fear of exploitation.

Social media was never meant to be a battlefield, she says. The Egerton University Gender Development graduate and commercial model got onto these platforms for the same reason many young people do: to share her life, build a modelling portfolio, and keep in touch with friends. Through carefully curated photographs and content, her audience grew, and along with it came opportunities to work with brands and make a living.

“I started as a commercial model and content creator while in university. I loved photography, and social media became a place where I could showcase my work,” she says in an interview.

Like many young women, Valerie treated the internet as simply an extension of her social life, never imagining the photos helping build her career could one day be turned against her.

Her first brush with TFGBV came in 2023. A friend messaged her asking if she had joined Tinder, and when Valerie said she had not, the friend sent back a screenshot that left her stunned. Someone had lifted her photographs and built a fake profile on the dating app using her identity.

“I wasn't even on Tinder, and even worse, they had used my photos on an LGBTQ profile. My biggest fear wasn't that people would think I was on a dating site. I was worried someone was using my image to scam or catfish innocent people,” she recalls.

She grew anxious about relatives stumbling on the account and assuming it belonged to her, and worse, imagined strangers confronting her over things done by whoever was hiding behind her face. At the time, she did not even think to call it a cybercrime.

“I thought cyber violence was people insulting you in comment sections. Online harassment, I understood, but I didn't know impersonation was also a form of TFGBV,” she says.

Because she had no intention of actually joining Tinder just to report the account, she asked friends already on the app to flag it for her, while telling her own followers across every platform that the profile was fake.

“I wanted to take back my voice. Someone else was trying to tell my story, so I had to reclaim my narrative,” Valerie notes.

She thought that ordeal was behind her until January this year, when a former classmate sent her a video that looked, at first glance, like Valerie embracing a man while a love song played in the background. It was not real. Someone had taken one of her professional modelling photographs and used artificial intelligence to turn it into a convincing deepfake.

“I took that photo in July 2025, and it was uploaded to Pinterest by the photographer, since we were promoting a weave. It looked like me, but it wasn't me. The picture had never been a video. The manipulated clip, uploaded on TikTok and YouTube by a content creator, had already attracted hundreds of views, hundreds of shares and countless comments before I knew it existed,” she explains.

As a commercial model, she says, your image is your brand, and anyone searching for you online should find your work, not fake content casting doubt on your integrity.

This violation played out on a platform she actually used, which made it more difficult to deal with. TikTok’s reporting tools exist for exactly this kind of harmful material, but Valerie's complaints kept getting rejected because, as she understood it, the platform simply was not built to recognise an AI-generated video wearing her face as a form of impersonation.

“It kept telling me my report was not valid. How do you explain that it is your face, but not your video?” she says, still visibly frustrated by it.

She eventually found the creator behind the video and confronted him directly. He claimed to have downloaded the image from Pinterest, where the photographer who shot her original portfolio had posted it, and agreed to take the clip down once he learned she had never consented to any of it. By then, though, the damage had already spread far beyond his reach.

“He told me people had already downloaded and shared it. That's when I realised once something is online, it's almost impossible to erase it,” she notes.

Fixing the problem

Despite the violations, Valerie still would not back down from the digital spaces, choosing instead to become part of fixing the problem, speaking now at awareness forums on TFGBV and sharing her own experience so other women can learn to spot online abuse, document it, and know where to turn for help.

Survivors of cyberstalking, impersonation and online harassment have started coming to her directly, and through those conversations she has come to see how many victims stay silent, whether out of fear of ridicule or simply because they do not know how to report what happened to them.

Technology companies need to strengthen their reporting systems, Valerie believes, but she is equally convinced that society has to change how seriously it takes online violence in the first place.

“When I tell my story, it helps me heal. More importantly, it lets other people know they are not alone. People still don't take these crimes seriously. Sometimes when you tell your story, people laugh because they think online violence isn't harmful, but the emotional and reputational damage is real,” she notes.

Her call now is for governments, law enforcement agencies and technology companies to work together on tighter digital safety policies, better reporting systems, and genuine accountability for perpetrators, especially as AI-generated content keeps getting more convincing.

“We all have a responsibility to use technology responsibly. Just because you can manipulate someone's image doesn't mean you should,” she advises.

Do not let online abuse silence you; that is her advice to women. “I am choosing to reclaim my voice instead of retreating into silence, becoming exactly what many survivors need most, a champion reminding women that their identity, dignity and story belong to them alone,” she concludes.

Millicent's fight has taken a different shape. Known online as Tindi Mwenyewe, the Nakuru-based content creator has had to deal with body-shaming and colourism instead.

Life before content creation meant casual jobs and wages too small to properly support her family, she says, recalling her upbringing in Kivumbini slum. Now a grandmother of seven, she credits content creation with changing the course of her life. She joined TikTok at 54, at the height of the Covid-19 pandemic, and used the platform to sensitise people to basic hygiene.

“People at the Wakulima market in Nakuru encouraged me to join TikTok to help sensitise a bigger audience to washing hands. On the first day, I gained 2,000 followers, something that motivated me. I was helped to create the account,” she notes.

She began collaborating with other creators, and her following grew steadily, but so did the abuse that came with it. First came the comments that she was too old to be making content, then the attacks turned to her physical appearance.

“I would make content, and in the comments section, some people would say I am ugly, shaming me for my body size. This would make me cry; however, I did not let them weigh me down enough for me to exit the platform,” she explains.

Young men have made sexual advances towards her during live sessions, Millicent says, and when she turns them down, they respond with insults, prompting her own children to step in and defend her.

She is aware of how to report the harassment but has never done it, choosing instead to block and delete the comments as they come.

“It got to an extent my mother did not sit right with what I was experiencing online; she was ashamed. I have challenges with reading, which makes it hard for me to report violations I experience online,” Millicent reveals.

Figures from UN Women show that while online violence reaches broadly across the population, female students bear a disproportionate share of it: 64.4 per cent report experiencing at least one form of online violence, against 35.5 per cent of their male counterparts.

Cyberbullying ranks as the second most common form of TFGBV, at 19.1 per cent, followed by non-consensual pornography at 17.8 per cent, cyberstalking at 14.6 per cent, online impersonation at 13.3 per cent, online sexual exploitation at 8.4 per cent, sextortion at 3.3 per cent, and doxing at 1.6 per cent. Online defamation tops the list overall, at 21.9 per cent.

Millicent wants tech platforms to take local languages seriously, arguing that most of the abuse Kenyans face online happens in vernacular languages and Kiswahili, territory these platforms are poorly equipped to monitor.

“For instance, TikTok does not use Swahili, so when people want to insult you, they switch from English to Swahili or their native language. It is important that techies ensure people can navigate online spaces without fear of abuse,” she explains.

She hopes the government follows through this time, enforcing laws such as the Cybersecurity and Computer Misuse Act and actually prosecuting perpetrators of online violence.

“I still encourage women to take up all the space they want to be in, and not allow trolls to stop them from realising their dreams. It can happen if perpetrators are punished and women continue to soldier on despite the hardships they experience online,” says Millicent.

Reporting originally appeared via Nation Africa. Read the full source for additional context.