The Race for Sovereign AI: How a UAE-India Deal Is Reshaping the Global Compute Map
For years, the global AI arms race has been cast as a contest between two superpowers. But on February 20, 2026, an announcement at the AI Impact Summit in New Delhi offered a striking alternative vision, one in which middle powers, acting in concert, could carve out their own space in the world's most consequential technological competition.
Abu Dhabi-based technology group G42 revealed plans to deploy an 8 exaflop AI supercomputer in India — a system of a scale that, until recently, only a handful of nations could dream of hosting. Built in partnership with US chip company Cerebras Systems, the Mohamed Bin Zayed University of Artificial Intelligence (MBZUAI), and India's Centre for Development of Advanced Computing (C-DAC), the project positions both the UAE and India as serious actors in the global AI infrastructure landscape.
A New Axis of AI Power
The announcement reflects a broader geopolitical trend: nations that are neither the US nor China are investing heavily to ensure they are not left behind in the AI era. The Gulf states, led by Abu Dhabi, have emerged as some of the most aggressive investors in AI infrastructure globally, deploying sovereign wealth into compute capacity, frontier research, and strategic technology partnerships.
India, for its part, brings scale. With 1.4 billion citizens, one of the world's fastest-growing digital economies, and a vast pool of technical talent, it represents both an enormous market for AI applications and a formidable potential source of AI innovation. What it has lacked — until now — is the raw computational infrastructure to train and deploy frontier AI models domestically.
"Sovereign AI infrastructure is becoming essential for national competitiveness," said Manu Jain, CEO of G42 India. The 8 exaflop system, hosted entirely within Indian borders and operating under Indian governance frameworks, is designed to change that calculus — giving India the ability to build, train, and scale AI without routing sensitive data or workloads through foreign infrastructure.
Sovereignty as Strategy
The emphasis on data sovereignty is not incidental. It is the central strategic logic of the project. As governments worldwide grapple with the risks of dependence on foreign cloud and compute providers, the ability to host and govern AI infrastructure domestically is increasingly viewed as a matter of national security, not just economic policy.
Andy Hock, Chief Strategy Officer at Cerebras, pointed to the company's prior success delivering Condor Galaxy supercomputers in the United States as proof that the technology is ready for deployment at this scale. The Cerebras Wafer Scale Engine 3 — the hardware underpinning the system — is the world's largest AI processor, and delivers training and inference speeds that its makers claim surpass conventional GPU-based systems by a wide margin.
Diplomacy Meets Deep Tech
The project is also a product of deepening diplomatic ties. It follows the 5th India-UAE Strategic Dialogue in December 2025 and the state visit of UAE President Sheikh Mohamed bin Zayed Al Nahyan to India in January 2026, which established a sweeping bilateral framework spanning defence, space, energy, and technology. That the most visible early deliverable of this framework is an AI supercomputer speaks volumes about where both nations see the future of geopolitical leverage.
G42 and MBZUAI have already laid groundwork at the model layer. In December 2025, they released NANDA 87B, an open-source large language model with 87 billion parameters trained across Hindi and English — one of the most capable multilingual models built specifically for the subcontinent. The new supercomputer would provide the computational foundation to go significantly further.
Richard Morton, Executive Director of the Institute of Foundation Models at MBZUAI, described the collaboration as part of a wider mission to ensure that the benefits of advanced AI are not concentrated in a handful of wealthy Western nations: "This collaboration with India represents a shared commitment to expanding access to advanced AI compute for researchers and students, enabling breakthroughs in critical areas like healthcare, agriculture and education."
If successful, the India deployment could serve as a template, demonstrating that exaflop-scale sovereign AI infrastructure is achievable not just for superpowers, but for any nation willing to build the right partnerships. In a world where AI capability is rapidly becoming synonymous with national power, that may be the most significant development of all.