In September 2025, during the UN General Assembly in New York, I sat in side event after side event about artificial intelligence. At one of them, Semafor’s The Next 3 Billion, chief executives took turns describing how AI and new technology would connect the 3 billion people still offline and open the digital economy to them.

Many of those 3 billion people are African. Very few of them, and very few people who know their lives, were on the stage.

That is the problem with the AI and digital rights debate as it stands. It is happening about Africa, often for Africa, but rarely with Africans setting the terms. And if we are not in the room now, the gap between those who build this technology and those who simply receive it will be wider than any we have seen before.

Two days earlier I was in a different room in the same city: the African Peer Review Mechanism’s Forward Africa Leaders Symposium at Nasdaq. There, African leaders were on the stage, on a panel called “Investing in the Future: People, Infrastructure and AI”. A GSMA presentation put the stakes in one line: “Leaving no one behind is leaving no one offline.” Its figures showed only a quarter of people in sub-Saharan Africa using mobile internet. Rooms like that one exist. They are still the exception.

Built in the West, for the West

Every technological revolution has been led by the countries with the capital and the infrastructure to drive it. That was true of the Industrial Revolution, and it is true of the fourth one we are living through now.

What is built in the West, for the West, by the West will leave out the Global South. Not out of malice, but because nobody designs for a life they have never seen. The engineers are not thinking about a farmer outside Kabala, a market woman in Makeni or a student in Bo with no laptop and an unreliable signal.

Africa’s median age is about 19. We are the youngest continent on earth, and millions of our young people are still off the grid entirely: no electricity, no smartphone, no internet. If the people designing AI policy and products never hear from someone who understands that last mile, the inequality will widen twice over. First between the West and Africa, and then inside our own countries, between the capital cities and everywhere else.

No one is bringing AI for your development

This year my team supported Wadhwani AI Global, which works with African governments on AI for health, education and agriculture. Our projects were mainly in health and education. The work taught me how “AI for development” gets talked about, and I came away convinced we need to retire the phrase.

Too often, AI for development is discussed the way aid is discussed: as something that will be brought to African governments, who wait for it to arrive. But look at who is actually building this technology. Nvidia, OpenAI, Google, Meta. These are for-profit companies. This is capitalism, and capitalism does not deliver anything as a favour.

No one is going to bring AI for your development. African leaders and African private enterprise have to build it and invest in it, and they needed to start yesterday. So let us stop talking about AI for development and start talking about how we design and shape AI for African realities, at the scale our population demands.

What Sierra Leone stands to lose

People ask what is at stake if Sierra Leoneans are left out of this conversation. The honest answer is that we are already left out.

The AI jobs that have come to Africa so far have gone elsewhere. In Kenya, workers label data and help train models for some of the biggest AI companies in the world. Our universities are not yet preparing graduates for the AI age. A young person who is left out of this shift becomes less marketable, earns less and produces less value. Multiply that across a generation and you have a country falling further behind every year.

It is also a question of language and information. If our languages are not in the datasets, the tools will not speak to our people. If our realities are not in the models, the misinformation that spreads on our phones will be easier to make and harder to catch. Left out, Sierra Leoneans become consumers of AI. Never its designers, and never the ones who set its agenda.

Six African voices the world should be quoting

The good news is that the expertise already exists. Here are six people from West, East and North Africa whose names should be on every serious AI panel and in every reporter’s contacts.

• Dr. Paul Azunre, Ghana. An MIT-trained computer scientist who co-founded Ghana NLP, an open-source effort building language technology for Ghanaian and other African languages. He is proof that the language gap is a choice, not a law of nature. He also performs as the musician Dr. Pushkin, which tells you something about the range of talent we are overlooking.
• Gbenga Sesan, Nigeria. Executive Director of Paradigm Initiative, which has defended digital rights and pushed for digital inclusion across Africa from Lagos since 2007. He sits on the UN Secretary-General’s Internet Governance Forum Leadership Panel, and his Digital Rights and Inclusion Forum has been setting the continent’s agenda for more than a decade.
• Dr. Shikoh Gitau, Kenya. CEO of Qhala, the Nairobi company she founded in 2020 to drive Africa’s digital transformation from within. She led Safaricom’s Alpha innovation unit and sits on the steering committee of the Lacuna Fund, which pays for the datasets AI in our regions has been missing. When Wadhwani AI Global asked us to find AI leaders in Nairobi, my research led to her. I had never met her. I sent a cold email, she answered, and the introduction became a partnership between Qhala and Wadhwani AI Global.
• Dr. Seydina Moussa Ndiaye, Senegal. An AI researcher at Senegal’s Cheikh Hamidou Kane Digital University, where he leads the ALIVE Lab on responsible AI. He sat on the UN Secretary-General’s High-level Advisory Body on AI, advises the African Union on AI policy, and helped bring Senegal its Taouey supercomputer. His argument is simple: Africa should build AI that creates value for its own people, not just consume what is made elsewhere.
• Karim Beguir, Tunisia. Co-founded InstaDeep in Tunis in 2014 with Zohra Slim, two computers and $2,000. BioNTech bought it for a reported $700 million. He argues AI built by Africans for African communities is less likely to carry Silicon Valley’s biases, and that the continent could produce “10 InstaDeeps in 10 years.”
• Dr. Timnit Gebru, Ethiopia. Born in Addis Ababa, she co-founded Black in AI and in 2021 launched the Distributed AI Research Institute (DAIR) to study how AI affects marginalized communities, with a focus on Africa. Her work on the risks of large language models changed how the industry talks about harm.

None of them waited to be invited. Each built something, said something new, or both.

How African experts get into the room

I have spent my career on both sides of the microphone, as a journalist and as a publicist. Here is what gets an expert booked or quoted, and it is not a long CV.

1. Bring new context. International editors are hungry for what they cannot get from their usual sources. Tell them what the global story looks like from Freetown, Accra or Nairobi.
2. Name the thing. Coin a term or a framing that captures what others are circling. A good phrase travels further than a good paper.
3. Find the missed angle. If every panel is about AI safety, ask who the model is safe for. If every report is about access, ask who owns the data once people are connected.
4. Show up while the story is breaking. When a global AI conversation is happening on X or Threads, join it that day, with substance. That is how African voices end up in the CNN and New York Times roundups: a reporter finds them in the conversation, not on a list.

How I use AI, and where it stops

I use AI every day at VR&C. I always write the first draft myself. Then I bring in Claude, my favourite, and Gemini as editors and as thinking and strategy partners. At journalism school I learned how much a good editor is worth, and now I have one on call at any hour from my desk in Bowie, Maryland.

But the limit is clear. There is no formula and no algorithm for the life you have lived. The Sierra Leonean-ness of things is not in the datasets, because Sierra Leoneans are not designing them and are barely in them. The tools can sharpen my sentences. They cannot supply the context, the memory or the relationships that make the work mine.

That originality is my value, and it is the value every African expert brings to this debate. It is exactly what the models are missing.

What global organizations should do differently

When a global organization builds an AI panel or a report on Africa, it usually looks to the giants: Kenya, Egypt, Tunisia, South Africa. That is a good start. It is not the whole continent.

Widen the circle. Invite the voices from countries like Sierra Leone, Liberia and Guinea, where AI infrastructure and skills are still thin. And change the frame when you do. The absence of infrastructure is not only a deficit to report. It is an opportunity: a chance to build right the first time, for people who have been designed out of every revolution before this one.

Africa will be home to a quarter of humanity within a generation. The AI being built today will shape how those people learn, work, vote and are seen. They deserve to be in the room where it is decided, and so do the experts who know their lives. We are ready. Pull up a chair.