Gulf Data Center Boom Set to Attract US Hyperscaler Billions
S&P sees GCC capacity tripling as AI capex reshapes regional power demand
S&P Global Ratings expects the six largest US hyperscalers, among them Amazon, Microsoft, Alphabet, Meta, Oracle and SpaceX, to spend more than $7 trillion on data centers and AI-related capital expenditure between 2025 and 2030. That is the scale of the capital cycle now under way in artificial intelligence, and the ratings agency argues in a new report that Gulf Cooperation Council countries are well positioned to capture the next wave of it, citing government funding, advanced infrastructure and relatively low electricity costs.
The regional opportunity is sizeable, even if the global share stays modest. Data center capacity across the GCC is expected to nearly triple to more than 3 gigawatts by 2030, with the UAE and Saudi Arabia accounting for more than 80% of new capacity. Together, the two countries are projected to add between 2 and 3 GW of IT power capacity by the end of the decade, an expansion driven by national AI strategies, growing demand for cloud computing and sustained investment in digital infrastructure. For market-watchers, the tripling of capacity signals where power, policy support and capital are converging in the region.
According to the report, continued spending on power networks and telecommunications infrastructure, combined with supportive government policies, has been steadily strengthening the GCC’s position as a hub for digital and AI infrastructure.
The flagship projects show where the money is going. In Abu Dhabi, the 5 GW UAE-US AI Campus, which includes the Stargate UAE project, ranks among the largest developments in the region, and its initial 1 GW phase is expected to be completed over the next three years. In Saudi Arabia, HUMAIN, an AI company backed by the Public Investment Fund, is targeting 1.9 GW of data center capacity by 2030, with plans to expand to between 6 GW and 6.6 GW by 2034.
That buildout carries a substantial power bill. Surging AI investment is the force pushing electricity demand upward. HUMAIN’s planned 1.9 GW of capacity alone could consume approximately 17 terawatt-hours of electricity annually if fully utilized, equivalent to roughly 6% of Saudi Arabia’s current electricity consumption. In the UAE, S&P expects electricity consumption by data centers to more than double by 2030, driven by the scale of the country’s projects.
Operators and government bodies are directing capital toward efficiency to manage that load. The UAE’s Ministry of Energy and Infrastructure, Khazna Data Centers and Agility have launched a pilot using technology from Phaidra, a Seattle-based AI company, to improve energy efficiency in data centers and district cooling systems. Phaidra claims its technology can reduce cooling-related energy consumption by up to 40%. The company is also developing a 100-megawatt AI data center in Ajman that uses advanced liquid-cooling technology to manage the heat generated by high-performance AI chips while improving energy efficiency.
Saudi Arabia is pursuing comparable savings. A 2026 report by the Kingdom’s Communications, Space and Technology Commission points to locally developed geothermal cooling technology from Strataphy, which can reduce cooling-related electricity consumption by 40% to 50% without using water during its cooling cycle. Saudi Energy, meanwhile, has signed agreements to expand electricity infrastructure for AI data centers, including HUMAIN’s planned facility near Riyadh.
For investors and operators, the central question is whether the power base can keep pace with demand. S&P expects the region’s combination of gas-fired generation, nuclear power and expanding renewable energy capacity, including gigawatt-scale solar, to provide a diversified mix capable of supporting the continuous demand of AI infrastructure. The agency cautioned, however, that while the GCC has the energy resources and infrastructure advantages needed to attract AI investment, its ability to sustain the expansion will increasingly depend on keeping power generation, cooling and connectivity aligned with growing data center demand.
Even so, the Gulf’s footprint in the global market remains small. The UAE and Saudi Arabia’s projected additions represent less than 2% of worldwide IT power capacity growth between 2026 and 2030, compared with approximately 70% expected to come from the US and Canada. The region’s wager is that low-cost electricity, state funding and new cooling technology can convert a modest slice of a vast capital cycle into a lasting position in AI infrastructure. Whether the power grid holds that pace will decide the bet.
Q&A
How much are the largest US hyperscalers expected to spend on data centers and AI capex?
S&P Global Ratings expects the six largest US hyperscalers, including Amazon, Microsoft, Alphabet, Meta, Oracle and SpaceX, to spend more than $7 trillion between 2025 and 2030.
How much data center capacity will the GCC add, and where?
GCC capacity is expected to nearly triple to more than 3 GW by 2030, with the UAE and Saudi Arabia accounting for more than 80% of new capacity, adding 2 to 3 GW of IT power capacity combined.
What are the flagship projects in the region?
In Abu Dhabi, the 5 GW UAE-US AI Campus, including the Stargate UAE project, with an initial 1 GW phase due within three years; in Saudi Arabia, PIF-backed HUMAIN targets 1.9 GW by 2030, expanding to 6-6.6 GW by 2034.
What are the main risks flagged for the Gulf buildout?
Power demand is the central risk: HUMAIN's 1.9 GW alone could use about 17 TWh annually, roughly 6% of Saudi consumption, and S&P cautions that sustaining expansion depends on keeping generation, cooling and connectivity aligned with demand.