By Stuart Kerr, Technology Correspondent, LiveAIWire

On May 13, 2025, Donald Trump landed in Saudi Arabia for the first stop of his Middle East tour and departed with technology commitments that reshaped the global AI infrastructure map. Saudi Arabia pledged 600 billion dollars in orders from American companies for advanced chips and services. The United Arab Emirates signed a 200 billion dollar agreement with the United States to jointly build what was announced as the world’s largest AI data centre complex in Abu Dhabi. Trump simultaneously ended the chip export restrictions that the Biden administration had placed on Gulf state AI purchases, restrictions designed to prevent advanced American chips from reaching China through intermediary countries, as confirmed by the Middle East Institute’s policy analysis of the reversal.

The UAE-OpenAI partnership that followed weeks later did not, as widely reported at the time, give every resident free personal access to ChatGPT Plus. That claim went viral and was subsequently debunked by regional outlets including Khaleej Times and Gulf News. What OpenAI and the UAE actually announced was that ChatGPT would become integrated nationwide across government, education, healthcare, and transportation services, the first such national deployment anywhere, alongside the Stargate UAE data centre project in Abu Dhabi. In January 2026, Saudi Arabia inaugurated Hexagon, the world’s largest government data centre by megawatt capacity, at 480 megawatts. Saudi Arabia declared 2026 the Year of AI.

The speed and scale of Gulf state AI investment has moved these countries from the periphery to the centre of the global AI infrastructure story within 18 months. The GCC data centre market reached 3.48 billion dollars in 2024 and will grow to 9.49 billion dollars by 2030 at an 18.2 percent compound annual growth rate. Technology spending across the MENA region will reach 169 billion dollars in 2026, according to Gartner forecasts. PwC projects AI will contribute 320 billion dollars to Middle East GDP by 2030, with Saudi Arabia contributing 135.2 billion dollars. Understanding what the Gulf states actually want from AI, as distinct from what they are buying, requires examining the strategic logic behind investments that have surprised even seasoned geopolitical analysts with their ambition.

The Diversification Imperative

The central strategic driver behind Gulf states AI strategy is economic diversification away from hydrocarbon dependence, a transition that Saudi Arabia’s Vision 2030 and the UAE’s parallel agenda have framed as existential. Non-oil GDP share has reached 55 percent in Saudi Arabia in 2025. AI is being deployed as a cross-cutting enabler for the sectors Vision 2030 has identified as growth drivers: tourism, logistics, entertainment, advanced manufacturing, and financial services. AI improves public services, supports economic diversification, accelerates sustainability goals, and helps organisations operate at new levels of efficiency and scale, as Microsoft’s December 2025 analysis of Saudi Arabia’s AI foundations put it directly.

The diversification logic extends to the labour market. Saudi unemployment stood at 7.2 percent in Q4 2025, concentrated among young Saudis who have historically struggled to compete with foreign workers in certain sectors. SDAIA, the Saudi data and AI authority that leads the national AI strategy, trained more than one million Saudi citizens in AI technologies through its SAMAI initiative within a single year. The strategic calculus is that AI skills, developed at scale in the Saudi population, will enable young Saudis to participate in the AI-driven economy that is replacing the oil-revenue-subsidised public sector employment of the previous generation.

The Infrastructure Play

The physical infrastructure investments Saudi Arabia and the UAE are making are not merely expressions of AI enthusiasm. They are deliberate bets on AI infrastructure becoming a form of strategic endowment comparable to oil reserves. The Gulf states possess structural advantages for AI data centre development that explain why American technology companies are investing there at this scale. Energy is abundant and cheap, capital is abundant through the sovereign wealth funds, and land and governance are predictable in ways that create a lower-risk environment for multi-decade capital commitments than many alternative locations.

Humain, Saudi Arabia’s state-backed AI entity and a subsidiary of the Public Investment Fund, is pursuing a 77 billion dollar infrastructure strategy targeting 1.9 gigawatts of data centre capacity by 2030. Google Cloud and PIF announced a 10 billion dollar partnership to build a global AI hub in Saudi Arabia through Humain. AWS committed 5.3 billion dollars to develop new Saudi data centres. Groq committed 1.5 billion dollars in collaboration with Aramco Digital to establish what could become the region’s largest AI inference data centre. The UAE’s Stargate project, involving OpenAI, NVIDIA, and G42, is building a 5-gigawatt AI campus that will be the largest such facility outside the United States. The Middle East’s data centre capacity is projected to triple from 1 gigawatt in 2025 to 3.3 gigawatts by 2030.

It is worth noting that this buildout is proceeding against a genuinely destabilised regional backdrop. Since late February 2026, Saudi Arabia has faced direct missile and drone strikes tied to escalation with Iran, with reports of damage to some regional cloud infrastructure and the postponement of Saudi Arabia’s flagship LEAP technology conference. How that conflict ultimately affects the infrastructure timelines described here is a live and unresolved question that this piece does not attempt to answer in full.

The Geopolitical Positioning

Gulf states AI strategy cannot be understood without its geopolitical context. Both Saudi Arabia and the UAE spent the preceding decade developing deep relationships with Chinese technology companies in telecommunications, data centres, cloud computing, and AI, relationships that created significant US security concern about Chinese hardware in critical Gulf state infrastructure. The Trump administration’s decision to lift chip export restrictions and enable massive US technology investment in the Gulf was partly a deliberate counter-move, drawing Gulf state AI infrastructure investment toward American companies and away from Chinese alternatives, a dynamic Axios documented in detail at the time of the announcement.

The UAE’s ambassador explicitly framed the chip export approval as validation of the UAE’s strategic pivot away from Chinese infrastructure. The strategic partnership model the Gulf states are pursuing, with US companies for AI infrastructure, with NVIDIA for chip supply, and with domestic entities like G42 and Humain for sovereign AI development, is a deliberate positioning that avoids full dependence on any single external party. LiveAIWire’s coverage of the parallel debate over Nvidia’s chip exports to China covers the other side of the same export-control story now shaping Gulf access to American hardware.

The Sovereign AI Dimension

Beyond infrastructure investment, both Saudi Arabia and the UAE are pursuing what analysts describe as sovereign AI, systems that reflect local values, operate in Arabic, and are governed by local frameworks rather than being imported wholesale from American or Chinese providers. The UAE released Jais 2, its Arabic large language model, and created an index to benchmark AI models against local values. Saudi Arabia’s SDAIA has signed seven major partnerships with US technology firms to accelerate AI capability building, while simultaneously investing in domestic talent development to reduce dependency on imported expertise.

Bahrain, Qatar, and Kuwait are pursuing parallel strategies at different scales, with Bahrain launching a comprehensive National Policy for AI governance compliant with the GCC’s ethical framework. This sovereignty question, over who controls access to the most capable AI systems and on what terms, is not unique to the Gulf. LiveAIWire’s reporting on Austria’s request for the EU to host Anthropic after US export controls cut off European access shows the same dynamic playing out among Washington’s own allies.

The Risks and the Questions That Remain

The Gulf states’ AI strategy faces genuine questions the current investment momentum does not fully resolve. The talent pipeline challenge is structural: there are not enough trained AI professionals in the world to meet demand, and the ambition to build self-sustaining AI ecosystems depends on producing domestic talent faster than current education systems can deliver. The 7.2 percent Saudi unemployment figure coexists with a chronic shortage of nationals with the technical AI skills the AI economy requires, a paradox that training programmes alone cannot quickly resolve.

The data question is similarly unresolved. AI systems require high-quality, domain-specific training data, and the aspiration to build systems that serve Arabic-speaking populations with local cultural context requires investments in data collection and labelling that are only beginning. The security dimension has also sharpened: LiveAIWire’s coverage of state actors using AI for espionage examines the same access-control questions now shaping who gets to build and use the most capable systems, a question the Gulf states are answering for themselves through sovereign AI investment rather than waiting for it to be answered for them.

The Chip Export Politics

The Trump administration’s decision to lift export restrictions on advanced AI chips to Saudi Arabia and the UAE, overriding the Biden-era framework that had classified these countries as requiring controlled chip access, was one of the most consequential technology policy decisions of 2025. The rationale was explicit: draw Gulf state AI investment toward American companies and away from Chinese alternatives, using chip access as a diplomatic tool. The decision was also commercially driven. Nvidia, Microsoft, Google, Amazon, and other US companies stood to gain hundreds of billions of dollars in Gulf state contracts that the export restrictions had been blocking.

The security concerns that motivated the original restrictions, that advanced chips imported to Gulf states might be re-exported to China or provide a vector for technology transfer, were addressed through licensing conditions and safeguard requirements that Microsoft, the first company to receive export approval under the new framework in September 2025, was required to satisfy. Whether those safeguards are sufficient is a judgment on which qualified analysts genuinely disagree, and the evidence needed to assess their effectiveness will take years to accumulate. The Gulf states’ technology relationship with China, including existing Huawei telecommunications infrastructure and Chinese cloud investments that predate the current US alignment, does not disappear simply because new US chip investments flow into the region.

The Arabic AI Gap and Why It Matters

The Gulf states’ investment in Arabic large language models, the UAE’s Jais 2, Saudi Arabia’s investments in Arabic AI capability through SDAIA partnerships, and similar efforts across the GCC, addresses a genuine gap in the global AI landscape. Arabic is spoken by more than 400 million people as a first language and by several hundred million more as a second language across the MENA region. It is among the most morphologically complex languages in the world, with a root-and-pattern system that creates challenges for tokenisation-based language models that do not arise in European languages. Frontier AI models, trained overwhelmingly on English-language data, perform significantly worse on Arabic-language tasks, and especially poorly on Arabic dialects, which differ substantially from Modern Standard Arabic and from each other.

An Arabic-language AI that performs well on Gulf, Egyptian, Levantine, and Maghrebi Arabic, reaching the populations that speak those dialects for daily communication rather than formal written language, would represent a genuine contribution to AI accessibility for hundreds of millions of people currently underserved by global AI infrastructure. Whether the Gulf states’ investments will produce models that genuinely outperform global frontier models on Arabic-language tasks, or locally-branded products that trail the technical frontier while serving sovereignty goals, is the practical question that will determine whether the investment generates the population-level benefit its proponents describe.

About the Author

Stuart Kerr is Technology Correspondent at LiveAIWire, covering artificial intelligence, emerging technology, and their impact on business, society, and everyday life. LiveAIWire publishes original AI journalism every weekday at liveaiwire.com.