Microsoft Corporation [NASDAQ: MSFT] and its partner G42, the Abu Dhabi-based technology conglomerate, have hit a major obstacle in their planned $1 billion data centre project in Kenya after negotiations with the Kenyan government broke down over guaranteed payment demands.

Microsoft and G42 had asked the government to commit to purchasing a fixed amount of computing capacity annually, but talks collapsed when the government was unable to meet the threshold the companies required.

The project, first announced in 2024, had envisioned a geothermal-powered campus starting at around 100 megawatts with a long-term target of scaling up to 1 gigawatt, and the first phase had been expected to become operational this year.

No construction has started, and the original timeline has already passed, with the delay potentially slowing what would have been the largest proposed private-sector digital investment in Kenyan history.

Kenya’s principal secretary at the Ministry of Information, John Tanui, told Bloomberg that discussions remain active and that the project has not been called off, adding that “the scale of the data center they wanted to do still requires some structuring.”

President William Ruto also flagged energy concerns at a recent state event in Nairobi, warning that powering a facility of that size would put severe strain on the country’s electricity grid.

Philip Thigo, Kenya’s special envoy for technology, later clarified that Ruto was highlighting an infrastructure challenge rather than signalling the project had been abandoned.

A smaller, 60-megawatt alternative facility with local developer EcoCloud remains under discussion as a possible reduced-scope option.

The Kenya project was the first joint venture between Microsoft and G42 following Microsoft’s $1.5 billion investment in the Abu Dhabi firm, which also required G42 to exit its Chinese investments and remove Chinese-made equipment from its systems.

The episode illustrates a broader challenge for hyperscale cloud expansion in emerging markets, where the scale of AI infrastructure projects increasingly requires sovereign finance commitments that many governments are not positioned to provide.