UAE’s 2025 AI infrastructure expansion sets stage for low‑carbon digital revolution

The United Arab Emirates capped 2025 by consolidating its ambition to be a global hub for artificial intelligence and digital infrastructure, rolling out large-scale compute facilities, cross-border partnerships and substantial public‑sector AI adoption that together reshape the region’s technology and energy landscape.

According to a WAM report, government entities reached a 97 percent utilisation rate of AI tools during 2025, while the number of programmers in the country surpassed 450,000. The public sector has been a particular focus: national programmes introduced an AI‑driven legislative analysis system, an AI HR assistant serving more than 50,000 employees and automated workflows across 108 government services. Government figures and independent assessments indicate more than 200 public AI solutions were deployed as part of a drive to become the world’s first fully AI‑native government by 2027, supported by targeted digital infrastructure spending. IMF analysis notes AED 13 billion in direct investment into digital infrastructure as part of that effort.

Large compute and campus builds formed the most visible component of the UAE’s 2025 strategy. Abu Dhabi became home to a 5‑gigawatt UAE‑US AI Campus described as the largest supercomputing cluster outside the United States, powered by a blend of nuclear, solar and gas generation. Complementing that, the Stargate UAE initiative , a consortium including G42, OpenAI, Oracle, Cisco, SoftBank and Nvidia , was announced as a 1‑gigawatt project using advanced NVIDIA Grace Blackwell GB300 systems, with an initial phase scheduled for 2026. Industry reporting on consortium activity also showed major institutional investors coalescing around AI infrastructure: a partnership involving BlackRock, Microsoft, Nvidia, xAI and UAE‑based MGX aims to scale next‑generation data centres and energy solutions, with potential commitments in the tens of billions of dollars.

Financial vehicles anchored by UAE interests have reached into Europe and beyond. Research from Emirates NBD and public statements indicate MGX and Mubadala‑backed structures have mobilised a $100 billion vehicle targeting ICT, semiconductor and data‑centre investments, and participated in projects to build Europe’s largest AI campus in France. Independent reporting on the Artificial Intelligence Infrastructure Partnership, including coverage by Tom’s Hardware, described a consortium acquisition strategy to secure operational and planned data‑centre capacity worth roughly $40 billion, underscoring the UAE’s role as both capital provider and strategic partner in global compute supply chains.

The scale of compute deployment has required an explicit energy strategy. The large campuses announced in 2025 were planned to run on mixed energy portfolios, pairing renewables and nuclear with gas to ensure reliability for high‑performance computing (HPC) workloads. National surveys cited by WAM reported that 44 percent of UAE entities now use high‑performance computing across 91 specialised use cases spanning healthcare, finance and security. For industrial decarbonisation stakeholders, that combination highlights two concurrent pressures: rapidly rising electricity demand for AI training and inference, and a policy imperative to supply that demand with low‑carbon energy at scale.

Private‑sector partnerships have been central to capacity and capability building. Microsoft said its technology collaborations have helped accelerate AI transformation in the UAE, including cloud and platform support for government service delivery; industry sources point to Microsoft technology underpinning major public‑facing platforms. Philanthropic and development partnerships extended the UAE’s reach: Abu Dhabi Developmental Holding Company (ADQ) partnered with the Gates Foundation in December 2025 to apply responsible AI and education technology to improve foundational learning in sub‑Saharan Africa, reflecting a strategy that pairs infrastructure investment with development objectives.

On the model and software front, the UAE announced substantial domestic capabilities: Jais 2, a 70‑billion‑parameter language model claimed to be trained on the largest Arabic‑first dataset assembled , 600 billion Arabic tokens , and K2 Think, positioned as an open‑source system for advanced AI reasoning. The government also introduced the “AI in the Ring” index, intended to assess how closely models reflect national cultural values; observers noted this as an attempt to couple capability with governance and cultural alignment.

Several broader themes emerge for industrial decarbonisation professionals evaluating the UAE’s trajectory. First, the scale of announced compute and data‑centre capacity makes the country a focal point for demand‑side electrification planning: procuring low‑carbon baseloads, integrating utility‑scale storage and designing flexible grid services will be critical to avoid emissions lock‑in. Second, large investment vehicles and international consortiums mean capital flows and technology transfer are likely to accelerate regional semiconductor and advanced manufacturing programmes, with knock‑on effects for supply chains that support clean energy technologies. Third, public‑sector adoption at near‑total levels positions the UAE as an early testbed for AI governance frameworks, procurement models and workforce transformation , all of which will influence how industrial sectors adopt automation while managing environmental performance.

Not all claims are without contest. While government and industry statements highlight record investment totals , WAM cited AED 543 billion in AI‑related investment for 2024–2025 , external observers urge scrutiny of committed versus deployed capital and the timelines for operationalising large projects. Similarly, model‑scale claims such as the size of the Arabic dataset and the novelty of national indices should be assessed against independent audits and benchmarked performance in applied industrial settings.

For energy and industrial leaders, the UAE’s 2025 push offers both opportunity and warning. Massive compute campuses create demand that can stimulate large‑scale renewable and nuclear deployment, grid modernisation and new low‑carbon fuels. At the same time, rapid build‑out risks locking in fossil‑fuel‑dependent generation unless procurement and regulation prioritise decarbonisation. The coming years will test whether the UAE can translate headline infrastructure announcements and investment vehicles into enduring, low‑carbon compute ecosystems that serve both domestic transformation and global digital needs.