Safety by Design: Reclaiming the Digital World for Women

Belamy Boson at the FemTech Academy on Digital Rights hosted by Equality Vanguard.

By Belamy Boson, Communications Strategist, SafeOnline Women Kenya (SOW-Kenya)

The Algorithmic Pipeline of Hate

Research conducted by University College London (UCL) and the University of Kent found that TikTok’s algorithm increased misogynistic content on users' feeds from 13 per cent to 56 per cent Nobody had to search for it. The platform delivered it automatically.

This was a key conversation had during the FemTech Academy on Digital Rights convened by Equality Vanguard. Online harm against women is not a side effect of the internet. It is a feature of how it was built. Every platform runs on one thing , keeping people online as long as possible. The easiest way to do that is emotion. Anger, outrage, and fear keep people scrolling. And content that targets and dehumanises women is extraordinarily good at generating exactly those emotions. So, the platform surfaces it. Recommends it and pushes it to more people.

The Rise of the Manosphere

The manosphere and redpill content did not just appear. The manosphere is a loose network of online communities that claim to address men’s struggles but often promote harmful advice and attitudes, united by opposition to feminism and the narrative that men are victims of the current social and political climate. It includes incels — involuntary celibates — men who blame women for their romantic failures and men’s rights activists who frame any gain in women’s rights as a loss for men. All this is tied together by red pill ideology — the belief that society is set up to be anti-men and that women and feminism are to blame.

None of this went mainstream by accident.It was algorithmically amplified, one recommended video at a time, targeting vulnerabilities like loneliness and loss of control, and walking users steadily toward more extreme content.And this is not a distant, Western phenomenon. In Kenya, data shows that manosphere language soared from near anonymity five years ago to tens of thousands of mentions annually, with the Covid-19 pandemic serving as an inflection point. African countries have their own versions of the manosphere, with content creators espousing red pill ideology that frames feminism as an oppressive system men must push back against. The algorithm does not recognise borders. And neither does the harm.

The AI Diversity Gap

Then there is the AI problem, which sits underneath all of it. AI systems learn from data, and that data reflects a world that was never built around women. According to UNESCO, women represent only 20 per cent of employees in technical roles in major machine learning companies, 12 per cent of AI researchers, and 6 per cent of professional software developers.

When the people building these systems do not reflect the people using them, the gaps become inevitable. An AI trained on incomplete data will reproduce that incompleteness at scale flagging feminist content as harmful, missing abuse directed at women, getting it wrong in ways that consistently disadvantage the same people.

This is why spaces like the FemTech Academy matter. And it is why Safe Online Women Kenya built SafeHer — a platform created not in response to women’s exclusion from the digital world, but as proof that inclusion is possible when women are part of building it from the start. The digital world is not going anywhere. The only question is who it will be built for.