In February 2026, residents of Iskandar Puteri in Johor protested against the construction of a 300-megawatt data center complex. Their concerns ranged from falling water pressure and construction dust to the loss of the green landscape around their homes. Reuters later described it as Malaysia’s first public resistance to a data center project.
The protest arrived in the middle of an investment boom. Johor had attracted US$35 billion after Singapore’s earlier moratorium pushed capacity across the border. By July, however, the debate had spread to Selangor and become politically harder to ignore.
This is an early warning for the Global South. Countries are competing to host the infrastructure of the AI economy, yet the decisive resource may be their ability to supply vast amounts of electricity without making surrounding communities absorb the scarcity.
Who can turn low-carbon energy into reliable compute?
Stay ahead of the geopolitical week.
MD Briefing delivers expert analysis across five global fronts — the Indo-Pacific, energy, geoeconomics, European security, and the Middle East — every Monday morning. Free.
The Renewable-to-Compute Conversion Gap
Renewable abundance creates an advantage, but it does not create an AI hub by itself. Solar potential in a desert, hydropower in a mountain range, or geothermal reserves in a rift valley may sit far from fiber routes, urban demand, and transmission capacity. Electricity must be delivered continuously to facilities whose workloads cannot simply wait for better weather.
This is the renewable-to-compute conversion gap: the distance between possessing low-carbon energy and turning it into dependable, investable, and socially legitimate computing capacity.
Conversion readiness depends on four alignments. Clean energy must exist at sufficient scale. The grid must deliver it through transmission, storage, and firm capacity. Data centers need fiber, cooling, land, and access to technology. Institutions must make projects financeable while preserving public acceptance and credible local benefits.
Renewables matter beyond climate targets. Conflict around the Middle East and its shipping routes can raise energy and construction costs across importing economies. Buyers of AI services also face growing pressure to account for supply-chain emissions. Clean power can reduce one source of conflict with communities, although it cannot excuse excessive water use or privileged access to electricity.
The AI economy will therefore reward regions that can manage energy as a system. Installed megawatts reveal only the beginning of the story.
Latin America’s Renewable Lead
Latin America enters the competition with the strongest regional electricity base. The Latin American Energy Organization reported that renewables supplied 69 percent of electricity generation in 2024, driven mainly by hydropower, wind, and solar. The International Energy Agency estimated that clean-energy investment reached US$70 billion in 2025.
Brazil offers scale, an established data center market, and a comparatively clean grid. Chile has abundant solar resources and a mature renewable-investment environment. Colombia and Mexico add large markets and expanding digital demand. Together they give the region something many competitors still lack: several plausible clean-compute nodes rather than one national champion.
Yet the regional average conceals deep asymmetry. Caribbean economies remain heavily exposed to imported fossil fuels, while several mainland grids face drought risk, congestion, and weak cross-border integration. Hydropower has supported Latin America’s lead, but climate volatility can turn the same dependence into a reliability problem. Clean energy geography also does not always match the metropolitan areas where data centers, fiber, and customers are concentrated.
Latin America’s advantage is real. Converting it will require transmission and storage investment, more affordable finance, and corridors that connect renewable zones with digital clusters. Otherwise, the region may export green electrons and critical minerals while capturing too little of the intelligence built on top of them.
South Asia’s Scale Paradox
South Asia has a different advantage: demand. India combines a vast digital market, domestic technology capabilities, data center expansion, and a state willing to mobilize infrastructure at scale. By January 2026, non-fossil sources represented more than half of India’s installed power capacity.
Actual generation tells a harder story. Coal supplied 69.5 percent of electricity from April to June 2026 and around 75 percent during non-solar peak hours. The divergence is a near-perfect illustration of the conversion gap. India has built clean capacity faster than its grid can turn it into round-the-clock clean power.
The regional picture is even more fragmented. Nepal has hydropower that Bangladesh wants to buy. Their tripartite arrangement with India enabled the first 40-megawatt transfer in November 2024 and regular seasonal exports from June 2025. When the parties sought to add another 20 megawatts in 2026, India’s Central Electricity Authority withheld approval because of transmission constraints.
Cross-border trade is possible, but each expansion still depends on infrastructure and bilateral permissions controlled by the transit country. South Asia can produce renewable energy and digital demand at exceptional scale. Its next test is institutional: turning several national energy relationships into a more predictable regional market.
ASEAN’s Corridor Opportunity
ASEAN enters from a more uneven energy base than Latin America and without India’s single-market scale. Its strength lies in the close proximity of data center clusters, submarine cables, manufacturing bases, growing digital demand, and renewable resources distributed across nearby countries.
Singapore, Johor, and Indonesia’s Riau Islands already resemble an emerging compute corridor. Singapore provides capital, connectivity, and enterprise demand. Johor offers land and expanding capacity. Indonesia can contribute a larger market and renewable potential, although its electricity system remains dominated by coal. Elsewhere, Laos and Vietnam bring hydropower, the Philippines has geothermal resources, and Thailand is attracting new data center investment.
The missing layer is conversion across borders. The ASEAN Power Grid has spent decades moving from aspiration towards limited interconnection. The World Bank estimates that realizing its 2045 vision will require around US$800 billion in generation and transmission investment. Without faster regional power trade, clean energy surplus in one part of Southeast Asia cannot easily support compute demand in another.
Johor shows why speed now matters. New projects must explain how they will source electricity, with renewable energy increasingly treated as a condition of approval. The state has stopped approving two categories of highly water-intensive facilities and asked operators to defer water-cooled expansion until new supply infrastructure is expected in 2027. Selangor is adding another test by seeking 30 percent local content in areas such as chip design and cooling systems.
These policies move the discussion beyond attracting hyperscalers. They ask how a host economy can protect water and power while gaining skills, suppliers, and technology. Johor’s protests make the deadline social as well as technical.
ASEAN could build the Global South’s most diversified clean-compute corridor. It could also reproduce today’s fragmentation at a larger scale, with data centers concentrated near Singapore while renewable resources, grid risk, and community costs sit elsewhere.
Africa’s Potential and the Gulf’s Speed
Africa contains some of the world’s most attractive solar, wind, hydro, and geothermal resources. It also faces the widest conversion gap. Many of its best renewable locations remain distant from high-capacity grids, fiber routes, and established data center markets. Governments must weigh new industrial demand against the unfinished task of providing reliable electricity to citizens and firms.
Kenya turns that tension into a concrete case. Microsoft and G42 announced a US$1 billion geothermal-powered campus at Olkaria in 2024, beginning with 100 megawatts and potentially expanding to one gigawatt. By 2026, the project had stalled. The full plan would have required roughly one-third of Kenya’s existing electricity supply, while the Treasury had not approved the proposed capacity arrangement.
The geothermal resource was present. The grid, financing model, and national demand architecture were not yet aligned. Africa’s opportunity remains enormous, but a green server enclave beside communities without dependable electricity would be politically fragile and developmentally difficult to defend.
The Gulf states provide the opposite model. The UAE can combine sovereign capital, centralized coordination, and privileged technology partnerships to build quickly. Its planned five-gigawatt AI campus in Abu Dhabi is expected to draw on nuclear, solar, and natural gas, with an initial 200-megawatt phase due in 2026.
Yet speed carries another dependency. Access to advanced chips and US technology remains tied to export licenses and geopolitical trust, while extreme heat raises cooling requirements. The Gulf can assemble compute faster than most Global South regions, but its model is capital-intensive and difficult to replicate.
The Competition Between Corridors
Latin America has the strongest renewable starting point, although finance and transmission keep its lead incomplete. South Asia’s scale and the Gulf’s speed rest on different foundations: one still relies on coal at peak hours; the other depends on concentrated capital and geopolitical access. Across Africa, exceptional resources remain separated from much of the grid and digital infrastructure needed to use them.
ASEAN sits in the most dynamic middle position. Its electricity is not yet the cleanest, but its proximity between digital hubs, industrial markets, and varied renewable resources creates a plausible corridor strategy. Johor also reveals the cost of getting that strategy wrong before regional governance catches up.
The meaningful unit of competition is becoming the renewable compute corridor: a connected geography where low-carbon power, transmission, fiber, technology access, investment rules, and public legitimacy reinforce one another. Some will cross national borders. Their governance will decide who finances capacity, who receives its benefits, and which communities carry the residual cost.
This changes how Global South governments might assess AI investment. Capital expenditure and installed megawatts capture only part of a data center’s value. Domestic capabilities, added grid resilience, and the scarcity left behind belong in the same calculation.
Renewable energy gives countries more room to manage geopolitical shocks, meet the expectations of global buyers, and negotiate a social license at home. Whether that room produces an inclusive AI economy will depend on institutions treating energy, compute, and public legitimacy as one strategic system.
The next map of AI will be drawn where institutions can convert low-carbon energy into reliable compute without transferring scarcity to the people living beside it.
