Kenya's internet reaches the world through Mombasa. Submarine cables landing there carry more than 95% of the country's international traffic, according to ATC News, which counts seven operational systems today and expects two more by 2027. Seven sounds like resilience. The geography says otherwise, because several of those systems, including SEACOM, which launched in 2009, reach Europe by running north through the Red Sea.
The news: a war zone on the main route
The Red Sea's cables enter through the Bab al-Mandab strait, less than 32 km wide. Rest of World reported on 7 October 2026 that more than 90% of data capacity between Europe and Asia passes under the Red Sea. It also reported that the Houthis seized Yemen's side of the strait in September, and that Saudi-backed forces claimed on 5 October to have retaken key positions. The Houthis denied that claim, so control remains contested.
The commercial reaction is already visible. Rest of World reports that Microsoft said on 23 September it would invest more than $400 million in subsea and land links across the Middle East by 2030. It also reports that Google bought fibre strands along Turkish state pipelines at two to three times the roughly €6 million that newer lines there are expected to cost. Google and Meta have begun sending some live traffic over a land route through Iraq. These are hyperscalers hedging with their own money.
The repair risk is the part Kenyan policymakers should dwell on. Doug Madory of Infoblox told Rest of World that a repair can take one to two weeks if a ship is available. But past cuts in these waters took months, because repair crews needed assurances they would not be fired on. Three Red Sea cables damaged in 2024 took months to fix, largely over permit delays.
The strongest case for a state-led response
There is a serious argument for the state to act directly. Resilience is a public good. A private operator prices in its own outage losses, not the cost to a Kenyan clinic, a mobile-money agent or a county government that loses service when a strait closes. Hyperscalers can buy Turkish fibre; a small operator cannot. On this view, government should fund or mandate route diversity, perhaps through a state-owned cable or minimum-redundancy obligations on landing licensees.
Why mandates would backfire
The diagnosis is right, but the prescription is where it goes wrong. The binding constraint on Kenyan resilience is not a shortage of state-owned capacity. Capacity is arriving: ATC News says the Daraja and Africa-1 systems, due by 2027, would push Kenya's international bandwidth past 28 Tbps. The constraint is what the routes have in common. More cables that share the same chokepoint add capacity without adding independence.
A mandate to build or buy specific cables would also lock in today's threat map. Nobody can say with confidence which corridor will be calm in 2030. Private operators with their own capital at risk adjust faster than a procurement cycle does. The Microsoft and Google moves show that firms with exposure will pay for diversity without being told to.
So the policy job is to remove friction from private diversification. The Communications Authority of Kenya, which describes itself as the regulatory authority for the ICT industry in Kenya, publishes a model submarine cable landing rights licence. Landing approval is the gate through which every new route must pass. Three changes would help most.
- Fast-track licences for routes that avoid the Red Sea. Cables that approach East Africa via the southern Africa corridor, or that link to the Gulf and onward by land, should get priority processing and published decision deadlines. Delay is the cheapest way to make a resilient route uneconomic.
- Open access to landing stations and terrestrial backhaul. A cable is only as useful as the path from the beach to the customer. Cost-oriented, non-discriminatory access to Mombasa landing stations lets smaller carriers buy diverse capacity rather than depend on one incumbent's lit fibre.
- A pre-agreed emergency protocol. The 2024 experience shows repair delays are as much about permits and security assurances as ships. Kenya cannot grant safe passage in Yemeni waters, but it can pre-clear its own permits and coordinate with regional partners so a repair ship loses no days on paperwork where Kenya has a say.
What not to do
The reflex in a crisis is to reach for localisation or sovereignty rules, such as requirements to keep traffic on state-controlled routes or to restrict foreign-owned capacity. Those would reduce choice just when operators need every option. The risk here is physical, not jurisdictional, and a data-residency rule does nothing for a cut cable.
The point also applies to how Kenya talks about the problem. Resilience should be measured by independence: how many distinct physical corridors can carry Kenyan traffic if one closes. Counting cable systems tells you little if they share the same strait.
The bottom line
Fighting at Bab al-Mandab has not cut Kenya off, and this analysis does not claim any Kenyan outage has occurred. But the exposure is real, and the repair timeline in a contested war zone is measured in months. Kenya's best response is not to out-build the hyperscalers. It is to make the licensing and access regime fast, open and predictable enough that private capital can build the diverse routes the market is already starting to pay for.