The Gulf Is Betting On Everyone’s AI And Building Its Own – Can That Strategy Actually Work?

With unprecedented, methodical investment, the Gulf states are pacing the global AI race.

Saudi Arabia’s Public Investment Fund and Abu Dhabi’s MGX have taken positions in OpenAI, Anthropic and Mistral simultaneously, while Saudi Aramco has launched Metabrain AI, a 250-billion-parameter model, and G42’s Jais series has produced bilingual Arabic language models now serving regional AI deployment through Core42. According to reporting from Mobile World Live and Innovation Library, PIF has also pursued stakes in Databricks and is in active discussions around further OpenAI funding through its AI arm, Humain, which has announced $77 billion in planned data centre investment and a national Arabic large language model.

The sheer volume of simultaneous activity is remarkable. On one side, sovereign wealth funds with some of the largest pools of capital in the world are buying into the companies defining the global landscape. On the other side, state-backed entities are investing in domestic compute, Arabic-language models and sector-specific applications for energy, finance and public services.

How this dual approach translates into long-term competitive advantage is one of the most interesting strategic questions in global tech.

 

The Dual Strategy, Explained

 

The investment positions on the global side are active – MGX co-led Anthropic’s most recent funding round, participated in OpenAI share sales and has been in negotiation over Mistral. These are active positions with strategic dimensions attached: the G42 partnership with Microsoft, for instance, required divesting from Chinese hardware but secured priority access to Nvidia Blackwell chips and US export licences for up to 35,000 GB300 systems. The Gulf isn’t simply writing cheques into Western AI; it’s acquiring infrastructure access and supply chain positioning at the same time.

The domestic aspect operates on a different rationale. The goal is unlikely to be direct competition with frontier model providers on raw capability. Saudi Aramco’s Metabrain and G42’s Jais are targeting specific use cases: industrial AI applications for the energy sector, Arabic-language deployment across the wider Arab world, and sovereign AI infrastructure that doesn’t depend on US or Chinese computing power. Humain’s planned 5GW data centre campus, if completed, would represent the largest AI infrastructure site outside the United States.

 

 

What Success Looks Like

 

Measuring the Gulf against OpenAI is probably the wrong benchmark.

A better question to ask is whether the Gulf can establish itself as the dominant AI deployment market and infrastructure provider for the Global South – the parts of the world that are too large to ignore and too specific to be served well by generic Western models. Arabic-language capability, compliance with regional data sovereignty requirements and sector-specific applications in energy and finance are all areas where the Gulf’s structural position generates authentic competition.

The risk that analysts flag isn’t the ambition but the execution pattern. The metrics that would progress – gigawatts of operational compute, share of non-US, non-Chinese inference workloads, adoption rates of Arabic models across the Arab world, Africa and South Asia – are harder to track than headline investment figures, and the region is still in the early stages of demonstrating them at scale.

We asked analysts and founders with direct exposure to Gulf AI strategy to weigh in on whether the dual approach holds up under scrutiny.

 

Our Experts

 

 

  • Nic Puckrin, UAE-based macro analyst and Co-Founder, Coin Bureau
  • Oscar Asly, Group CEO, M4Markets
  • Dr V. Charles, Reader in AI for Business and Management Science, Queen’s University Belfast
  • Leo Derikiants, CEO and Co-Founder, Mind Simulation Lab

 

 

Nic Puckrin, UAE-based macro analyst and Co-Founder, Coin Bureau

 

Nic Puckrin, UAE-based macro analyst and Co-Founder, Coin Bureau

 

“It’s neither – it’s more of a hedged bet, one based on a very sober assessment of where the UAE stands on global AI competitiveness. It simply can’t out-research the US AI labs over the next five years, so the strategy is both funding American frontier development through MGX’s Stargate stake and the Microsoft-G42 partnership and building sovereign infrastructure at home. It’s essentially buying optionality. It has a stake in the success of US AI labs, but also regional infrastructure if a more decentralised, open-source model emerges. It’s the best of both worlds.

“It’s definitely not building a regional ChatGPT competitor. It’s more about dominating the AI infrastructure space for the Global South – basically anywhere outside the US and China. The metrics that matter are GW of operational compute, share of non-US, non-Chinese inference workloads, and adoption of UAE-developed Arabic and multilingual models across the wider Arab world, Africa, and South Asia. The UAE is already delivering on the first two – Stargate UAE’s planned 5GW campus would be the largest AI site outside the US, and OpenAI’s recent pause on its UK project over energy costs and regulatory friction makes that execution speed look even sharper. The open question, though, is adoption across the region.

“It’s a hedge, and a clear-eyed one. By investing in OpenAI, Microsoft, and the broader Stargate ecosystem, the UAE is signalling a realistic assessment of the landscape. Namely, that no country can fund the frontier AI model race alone, so partnering with a leader simply makes strategic sense. What matters is whether the investment it’s making comes with strategic access, which it does. The G42 partnership required divesting from Chinese hardware, but in exchange the UAE secured priority access to Nvidia Blackwell chips, US export licences for up to 35,000 GB300 systems, and integration into the most advanced AI supply chain in the world. That’s something the UAE is willing to pay the price for.”

 

Oscar Asly, Group CEO, M4Markets

 

 
Oscar Asly, Group CEO, M4Markets

 

“The Gulf’s AI strategy makes sense because it reflects how frontier technology is now built. No single market owns the full stack. The model layer, compute, energy, data, regulation and commercial deployment are all being fought over at once. Saudi Arabia and Abu Dhabi are taking strategic positions in the companies setting the global pace, while also building the domestic conditions needed to deploy AI at scale. That means data centres, sovereign cloud capacity, Arabic language models, sector-specific applications and regulation that allows deployment across finance, energy, logistics and public services.

“The region does not need to produce the next OpenAI for this strategy to work. A more realistic measure of success is whether MENA can become one of the world’s most effective AI deployment markets. If AI changes how banks manage risk, how energy companies optimise production, how governments deliver services and how trading platforms process data, then the region has captured real economic value. Backing Western AI companies should be seen as access to talent, research and commercial relationships. The frontier model race is expensive and concentrated. Sovereign investors are right to take exposure to the leaders while local capability develops.

“The real risk is duplication. If every institution wants its own flagship model, the region will burn capital and scatter talent. The Gulf’s advantage is unusually strong: capital, energy, infrastructure, government coordination and long investment horizons. The next phase has to convert those advantages into practical capability and measurable economic output, not another round of announcement-led spending.”

 

Dr V. Charles, Reader in AI for Business and Management Science, Queen’s University Belfast

 

 
Dr V. Charles, Reader in AI for Business and Management Science, Queen’s University Belfast

 

“Investing in OpenAI, Anthropic, and Mistral AI is not a sign of lacking confidence in homegrown capability. It is how a sovereign actor buys frontier proximity while domestic capability matures. These are not merely passive financial positions. In many cases, they include compute agreements, ecosystem partnerships, talent pipelines, and institutional visibility into how the world’s leading labs actually operate. What looks like contradiction is actually coopetition – cooperative investment and competitive building operating as a single integrated strategy, not two separate ones. Saudi Arabia’s Public Investment Fund and Abu Dhabi’s MGX are not simply writing cheques; they are acquiring strategic optionality at the most consequential moment in the technology’s development.

“Success for the Gulf’s sovereign AI ambitions should not be measured against Western frontier benchmarks. The goal is likely not to train a model that directly competes with GPT-5. It is to own the inference layer serving Arabic-speaking populations, to build vertical AI applications in energy, logistics, finance, and public services where general-purpose Western models may structurally underperform, and to establish the data sovereignty infrastructure that underpins genuine long-term autonomy. By that definition, the region appears meaningfully on track.

“The most underappreciated dimension is geopolitical. Gulf states are actively positioning themselves as a credible AI bridge between Western frontier development and the Global South. That is a role few Western nations can occupy as credibly and no purely domestic AI programme could claim. It requires exactly the dual presence the Gulf has built: inside the tent of global AI development, while simultaneously constructing sovereign capability at home. That is not a contradiction. That is a sovereign strategy.”

 

Leo Derikiants, CEO and Co-Founder, Mind Simulation Lab

 

Leo Derikiants, CEO and Co-Founder, Mind Simulation Lab
“The Gulf’s strategy makes sense as long as sovereign AI ambitions are matched with resilient infrastructure. Right now, the real bottleneck is access to hardware, power stability and supply chains. We’ve seen GPU and memory costs increase dramatically in recent months, while some components have become almost impossible to source through traditional UAE routes. Gulf investment into frontier Western AI companies should be viewed less as a lack of confidence and more as a way to secure immediate access to ecosystems and compute capacity while domestic capabilities mature.”

 

For any questions, comments or features, please contact us directly.
techround-logo