The race for sovereign AI is, at its core, a race for compute. Across the Middle East and North Africa, governments are spending at a scale not seen since the oil infrastructure buildouts of the 1970s. But where exactly is that compute landing? This audit traces the physical and legal footprint of MENA's declared sovereign AI infrastructure, site by site.
The Gulf Cluster
## By The Numbers - **40 per cent** - **$1.5 billion** - **$2 trillion - Combined Gulf sovereign wealth deployed toward AI and technology diversification** - **40% - Projected increase in MENA AI market size year-on-year through 2028**Saudi Arabia's HUMAIN project, announced at the May 2025 US-Saudi summit, has committed to 300,000 Nvidia H100-equivalent GPUs by end of 2026. The primary data centre sites are Neom's Sindalah island facility and a second campus in Riyadh's King Abdullah Economic City. Cooling is the constraint: both sites use hybrid air-liquid systems, and HUMAIN's engineers have acknowledged that Neom's ambient temperature requires 40 per cent more cooling energy than a comparable Nordic facility., as highlighted by Saudi Data and AI Authority (SDAIA)
For related analysis, see: [Saudi Arabia's AI Development: A Future Blueprint?](/voices/opinion-saudi-arabia-ai-development-future-blueprint).
The UAE's G42, through its Khazna Data Centres subsidiary, operates seven facilities across Abu Dhabi and Dubai. The Mussafah campus alone carries an estimated 120,000 GPU capacity, with expansion underway. G42's partnership with Microsoft, formalised in April 2024 with a $1.5 billion investment, brings Azure sovereign cloud to the same physical infrastructure, creating a hybrid model that complicates the "sovereign" classification.
For related analysis, see: [Stability AI in Turmoil](/news/stability-ai-in-turmoil).
Qatar's AI infrastructure is smaller but strategically placed. The Qatar Computing Research Institute at Hamad Bin Khalifa University houses the HBKU HPC cluster, a 5,000-node system that underpins the country's Arabic NLP research programme. Qatar is explicitly not attempting GPU-scale compute parity, positioning instead as a governance and research hub., as highlighted by UAE Artificial Intelligence Office
The North Africa Tier
Egypt's Karnak national LLM project relies on infrastructure spread across three Cairo data centres, none of which was purpose-built for AI workloads. The Ministry of Communications has allocated LE 2.3 billion (approximately $46 million) through 2027 for GPU procurement, but independent estimates suggest current Karnak training runs are constrained to batches of under 10,000 A100-equivalents, sufficient for fine-tuning and RAG but not for training a frontier model from scratch.
For related analysis, see: [Beyond Search: Google Unveils App-Based Gemini Chatbot](/news/beyond-search-google-unveils-app-based-gemini-chatbot).
Morocco's AI ambitions are anchored at the Mohammed VI Polytechnic University UM6P campus in Benguerir, which hosts the continent's largest academic GPU cluster. The 2,048-GPU system is modest by Gulf standards but has produced Arabic-French multilingual models with commercial traction in Francophone Africa., as highlighted by Nvidia AI
For related analysis, see: [Opinion: Saudi Arabia's AI Dominance](/voices/opinion-saudi-arabia-ai-dominance-strategic-approach).
The Sovereignty Problem Nobody Is Talking About
Every sovereign AI project in this survey is grappling with the same structural problem: the GPUs are either Nvidia (US export-controlled) or Huawei Ascend 910B (subject to its own political constraints). True sovereignty, meaning no foreign vendor kill-switch, is not achievable on current hardware. The more honest framing, used quietly by SDAIA officials, is "strategic autonomy within managed dependency."
The next episode in this series will map the model layer: which foundation models are being trained, fine-tuned or adapted at each site, and which teams are doing the work.
THE AI IN ARABIA VIEW
The intersection of AI and energy in the Middle East is not merely an efficiency play; it is existential. These economies must use AI to optimise their hydrocarbon present whilst accelerating their renewable future. The organisations that master this dual mandate will shape the region's economic trajectory for decades.
- HUMAIN official announcements, May 2025
- G42 investor presentation, Q4 2025
- Egypt MCIT budget allocation, FY2026
- UM6P AI Lab publications, 2025-2026
Several MENA nations, led by Saudi Arabia and the UAE, have committed billions in sovereign AI infrastructure, talent development, and regulatory frameworks. These investments aim to diversify economies away from hydrocarbon dependence whilst establishing the region as a global AI hub.
### Q: What role does government policy play in MENA's AI development?Government policy is the primary driver. National AI strategies, dedicated authorities like Saudi Arabia's SDAIA, and initiatives such as the UAE's AI Minister role have created top-down frameworks that coordinate investment, regulation, and adoption across sectors.
### Q: How is AI transforming the energy sector in the Middle East?AI is being deployed across the energy value chain, from predictive maintenance in oil and gas operations to optimising solar farm output and managing smart grid distribution. The technology is central to the region's energy transition strategies.
### Q: What are the key smart city AI projects in the Arab world?- Major projects include Saudi Arabia's NEOM
- Dubai's Smart City initiative
- Abu Dhabi's Masdar City
- all showcasing AI-driven traffic management
- waste optimisation
- citizen services integrated from the ground up