Skip to content
Somalia is missing from the global digital future episode artwork

Sep 20, 2026 5 min

Somalia is missing from the global digital future

from Somali Waa kuma Podcast Hosted by MI

MI’s 5 Mins of Fame - week 38 of 2026 presents… Watching the Ai revolution from the sidelines, while powerful countries debate how Ai should be controlled, Somalia risks becoming a consumer of systems designed and governed elsewhere (In Somali Language see https://tinyurl.com/yx53k9m5). I still begin most days with printed newspapers. The Age, The Australian and The Australian Financial Review remain part of my daily reading. This week the intensity of the Ai debate was difficult to miss in any of them. The AFR’s coverage is the most useful specimen, because its density shows how the question is currently being framed. Of the 34 AFR articles I collected between Monday and Friday, close to thirty dealt substantially with artificial intelligence: safety, employment, copyright, data centres, energy, investment, sovereign capability and the technological contest between Washington and Beijing. Against that volume, the absence of Africa, and of Somalia in particular, was conspicuous. My interest in this debate is not new. I approach AI as a technologist, researcher and former Somali government minister responsible for telecommunications. In earlier writing on LinkedIn, Ayuub.org and MI’s 5 Minutes of Fame, I have repeatedly returned to the place of Africa in technological change. One case has stayed with me. Kenyan workers contracted through the outsourcing firm Sama helped make ChatGPT safer by labelling deeply disturbing material. According to a TIME investigation, some earned between US$1.32 and US$2 per hour for exposure to accounts of violence, sexual abuse and other traumatic content. Africa was present at the making of the Ai revolution, at its lowest and least protected tier. The week’s coverage therefore raises a question of standing rather than of technology. The United States, China, Europe and Australia are arguing over how Ai should be controlled, regulated and owned. Africa is seldom a party to that argument, and Somalia is not a party at all. The debate belongs to those who already hold power The AFR examined autonomous systems, cybersecurity, copyright, employment, data centres, energy, corporate concentration and the rivalry between the United States and China. Several articles reported warnings from senior technology executives that advanced systems are becoming more capable, more autonomous and harder to supervise. Others interrogated the motives behind those warnings. Regulation may protect the public; it may equally entrench the firms that already dominate the market by raising the cost of entry for smaller competitors. The copyright dispute poses a related question of consent, namely whether models may lawfully be trained on the work of writers, journalists and artists without permission or payment. What gives those warnings their force is their source. Sam Altman of OpenAI, Demis Hassabis of Google DeepMind, Dario Amodei of Anthropic and Geoffrey Hinton, who left Google in order to speak without corporate constraint, have all endorsed the proposition that reducing the risk of extinction from artificial intelligence should rank as a global priority alongside pandemics and nuclear war. The men who built these systems, and who continue to profit from them, describe an invention they believe capable of ending human civilisation as we have known it. That apocalyptic register now sits inside ordinary business journalism, printed beside earnings results and property advertisements. I am reading Toby Walsh's God AI: Boom or Doom? as this argument unfolds, and his framing is the sharper one: whether we are building a successor or a deity, and whether a species should build either before it knows how to switch it off. Australia is also debating sovereign capability. Kevin Rudd, the former prime minister, has argued that the country should develop its own AI capacity rather than rely wholly on Washington or Silicon Valley. The argument is only nominally technical. It concerns national security, economic independence and a state’s capacity to defend its own infrastructure. Ai has become so powerful that even kings are now worried about saving humanity. When royalty starts warning us about machines taking over, perhaps the rest of us should stop asking AI to write our emails and start preparing for the revolution. These are serious debates, but they are conducted among states that already command capital, infrastructure, trained researchers and regulatory leverage. Australia can negotiate with global technology firms, establish an AI safety institute, regulate data centres and weigh the merits of sovereign models. Most African states are not calibrating the speed of the frontier. They are trying to reach it. Africa is present, but rarely powerful Africa is not entirely silent. The African Union adopted a Continental Artificial Intelligence Strategy in 2024. Countries including Kenya, Rwanda, Nigeria, Egypt, Morocco and South Africa have introduced policies, research initiatives or national discussions. The difficulty is structural. Much of the continent still lacks reliable electricity, affordable internet access, advanced computing capacity and sufficient investment in research. An IMF assessment reported in 2026 found that Ai could raise Sub-Saharan Africa’s economic output by around 4 per cent over the next decade. Without improvements in electricity, connectivity and digital skills, however, the gain could be almost negligible. Only about 38 per cent of Africans used the internet in 2024, compared with 68 per cent globally. Africa also holds only a small share of global data-centre capacity. This creates a new form of dependency. African countries may use Ai extensively without owning the models, computing infrastructure or intellectual property behind it. They may provide data and low-paid labour while the financial and strategic value is created elsewhere. This is not inclusion in any meaningful sense. It is the old economic arrangement placed inside a new technology. There is also the question of knowledge. Ai models are trained mainly on material produced in dominant languages and wealthier societies. Somali language, history, culture and social realities remain poorly represented. A system may produce confident answers about Somalia while understanding very little of the country. Somalia cannot remain indifferent Somalia has demonstrated an impressive ability to adopt telecommunications and mobile money. Yet success with one generation of technology does not guarantee preparedness for the next. The country still lacks a publicly visible and comprehensive national Ai strategy, a dedicated legal framework, a clear institutional lead and a serious public debate about Ai in education, government, finance, media and security. There are emerging academic and institutional initiatives, but they do not yet amount to a coordinated national response. This leaves Somalia exposed. AI-generated disinformation could inflame political and communal tensions. Imported systems could influence policing, education, banking and public services without adequate scrutiny. Somali data and cultural materials could be extracted without consent. Decisions affecting citizens could be made by systems that local institutions cannot explain, examine or challenge. Somalia does not need to build its own version of OpenAI. It needs a realistic national programme. This should begin with an AI-readiness assessment covering infrastructure, education, cybersecurity, data protection and employment. A national taskforce should bring together government, universities, telecommunications companies, financial institutions, civil society and diaspora expertise. Universities should develop Somali-language datasets and conduct applied research. Government procurement should require transparency, human oversight and the ability to suspend systems that cause harm. The greatest Ai danger facing Somalia may not be an … This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit somalia.substack.com

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

Somalia is missing from the global digital future

0:00 5:06

No transcript yet.

Requests help us decide what to transcribe next.

Frequently asked questions

How long is this episode of Somali Waa kuma Podcast?

This episode is 5 minutes long.

When was this Somali Waa kuma Podcast episode published?

This episode was published on September 20, 2026.

Can I download this Somali Waa kuma Podcast episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.