The Gulf’s bid to become a global AI powerhouse is, in 2026, primarily a race for compute, chips and Arabic-language capability. Saudi Arabia’s Public Investment Fund launched HUMAIN in May 2025 to anchor the kingdom’s AI strategy, while Abu Dhabi’s G42 and the $100 billion investment vehicle MGX carry the UAE’s push. The two states are buying advanced chips and building gigawatts of data-centre capacity — but talent and dependence on US hardware remain the binding constraints.
What is the GCC actually building in 2026?
The Gulf’s AI ambition rests on three pillars: sovereign compute, home-grown models and the capital to fund both. HUMAIN, established under the PIF and led by chief executive Tareq Amin, secured 2025 partnerships with Nvidia, AMD and Qualcomm and has stated plans to build several gigawatts of data-centre capacity. Its flagship model, ALLaM, is an Arabic-first large language model, with a 34-billion-parameter version reportedly trained on more than 500 billion Arabic tokens and now fronting a consumer chat app.
In the UAE, Abu Dhabi has positioned itself as the region’s compute hub. G42 builds and operates AI infrastructure and cloud capacity, while MGX — backed by Mubadala and others — deploys capital into global AI infrastructure at scale. As we set out in our look at how Abu Dhabi is betting $100 billion on AI, the emirate’s strategy is to own the physical layer of the AI economy rather than merely consume it.
Why is compute the real battleground?
Models are commoditising; the scarce resource is the compute to train and run them. That is why access to advanced Nvidia and AMD accelerators — and the electricity to power them — has become the currency of AI sovereignty. Nvidia was reported to be routing tens of thousands of top-tier chips to Saudi Arabia in 2025, and both Riyadh and Abu Dhabi are racing to secure US export approvals for successive chip generations.
The Gulf’s advantage is unusually cheap land, energy and sovereign capital; its exposure is that the most valuable chips are made in Taiwan and licensed under US rules. Control of the hardware pipeline sits in Washington, not the Gulf — a dependence no amount of capital fully removes.
The Gulf AI landscape at a glance
| Initiative | Country | Role | Anchor / backer |
|---|---|---|---|
| HUMAIN | Saudi Arabia | National AI champion; ALLaM model; data centres | Public Investment Fund |
| G42 | UAE (Abu Dhabi) | AI infrastructure, cloud, applied AI | Abu Dhabi |
| MGX | UAE (Abu Dhabi) | $100bn AI-infrastructure investment vehicle | Mubadala and partners |
| Falcon / Jais | UAE | Open Arabic-capable language models | TII / national institutions |
Where does the Gulf still fall short?
Capital is not the constraint; people are. The GCC must import most of its senior AI research talent, and retaining it against Silicon Valley pay and academic ecosystems is hard. Building a domestic pipeline of PhD-level researchers is a decade-long project, which is why the UAE has tied AI literacy into its schools — a theme in our coverage of the UAE AI Strategy 2031. Energy is the second ceiling: gigawatt-scale data centres compete with desalination and cooling for power in some of the hottest markets on earth.
The strategic risk is spending heavily on infrastructure that depreciates fast while the highest-value layer — frontier models and the researchers who build them — stays offshore. As our editorial on GCC AI acceleration argued, buying compute is necessary but not sufficient.
What would make the Gulf a genuine AI leader?
The measure of success will not be the number of chips imported but whether the region can turn compute into applied advantage. Three markers matter. First, a domestic research base: universities and labs producing Arabic-native models and the engineers who maintain them, rather than licensing talent on short contracts. Second, adoption at home — banks, hospitals, ports and government services running Gulf-built AI at scale, which turns infrastructure into productivity. Third, exports: selling compute capacity and Arabic-language models to the wider Middle East, Africa and South Asia, where demand for sovereign, non-Western AI is real. If the GCC hits those three, the “powerhouse” label is earned; if it stops at data centres, it will have bought influence rather than built capability.
Frequently asked questions
Is Saudi Arabia or the UAE ahead in AI?
They lead in different layers. The UAE, through G42 and MGX, is further ahead on infrastructure and global investment reach; Saudi Arabia, through HUMAIN and ALLaM, is moving fast on national compute and Arabic models backed by the PIF’s balance sheet.
What is HUMAIN?
HUMAIN is a Saudi AI company launched in May 2025 under the Public Investment Fund to drive the kingdom’s AI strategy, spanning data centres, the ALLaM model and consumer applications.
Why does Arabic-language AI matter?
Global frontier models underperform in Arabic. Models such as ALLaM, Falcon and Jais give the Gulf digital sovereignty over how AI handles its language, culture and public services.
What is the biggest risk to Gulf AI ambitions?
Dependence on US-controlled chips and a shortage of senior research talent, combined with the risk of over-building infrastructure that ages quickly.
Bottom line: The GCC has the capital, cheap energy and political will to be a serious AI power, and HUMAIN, G42 and MGX are turning that into real compute. Whether it becomes a genuine powerhouse — rather than a well-funded customer of Western hardware — depends on closing the talent gap and reducing chip dependence over the next five years.
