As the United Arab Emirates accelerates its bid to position itself at the center of the global artificial intelligence economy, officials are framing AI tokens, hyperscale compute, and data-driven productivity as the foundations of a new industrial era one in which intelligence, rather than oil, becomes the country’s defining export.
A National Bet on Intelligence as the Next Strategic Resource
At the Milken Middle East and Africa Summit in Abu Dhabi, the UAE laid out one of its most ambitious technological visions to date: the creation of what officials describe as the “world’s factory of intelligence.” Central to this plan is the production of up to 60 trillion AI tokens at the Stargate campus, a scale that would represent roughly 60 percent of global token production.
Omar Al Olama, the UAE’s Minister of State for Artificial Intelligence, Digital Economy, and Remote Work Applications, framed this shift as part of a broader economic transformation. AI tokens, he said, will serve as the currency of tomorrow units of computation that can be converted into insights, predictions, and decisions. Just as hydrocarbons once defined geopolitical influence, the UAE believes control over AI compute and token throughput may define the next era of global competitiveness.
The strategy marks a reframing of national resources. Instead of rare earth minerals or oil, Al Olama argued, token capacity will shape which countries export intelligence and which depend on others to power their digital industries.
Economic Returns and Energy Efficiencies Already Taking Shape
The government is already citing tangible financial gains. AI tools deployed across the UAE’s energy sector have saved approximately $136 million according to Al Olama. Companies such as Adnoc Distribution are using machine-learning systems to reduce production costs, assess output with real-time precision, and streamline logistics.
“These savings are only the beginning,” Al Olama said, noting that the metric for evaluating energy production is shifting from barrels and refineries to models and compute efficiency. He pointed to accumulating productivity gains as evidence that the UAE’s early AI investments are generating economic returns.
The emphasis on efficiency reflects a broader national calculus: the belief that AI can multiply the value of existing industries while creating entirely new categories of economic activity. But it also underscores the scale of infrastructure needed, from hyperscale data centers to stable, high-capacity power generation.
Building Stargate: A Hyperscale Campus for an Emerging AI Economy
The Stargate AI campus is the centerpiece of this emerging infrastructure. Designed as a hyperscale facility, it is projected to ultimately form part of a 5-gigawatt AI hub in Abu Dhabi. By 2026, Stargate is expected to operate at 200 megawatts, with phased expansion planned as global demand for AI compute continues to rise.
“We are not looking at this blindly,” Al Olama said. “We’re going to increase capacity as demand increases.”
He acknowledged that a 5GW facility could fluctuate in scale over time, but maintained that the UAE’s trajectory is clear: the country intends to become an indispensable node in the global AI supply chain.
Unlike conventional data centers, the facilities envisioned under Stargate are described as “AI factories” optimized not just to store and process data, but to manufacture intelligence. These centers specialize in handling massive volumes of AI tokens, converting raw data into real-time outputs such as predictions, reasoning, analytics, and content.
Navigating Uncertainty as the UAE Seeks a Global Role
Even as officials outline an expansive vision, they acknowledge that uncertainties remain. Scaling to 60 trillion tokens has no established precedent, and timelines remain fluid. “No one is certain about anything,” Al Olama said.
“But the investment is there. As long as our partners work with us and we’re able to get what we need at the right time, then we are certain.”
He also addressed concerns about overvaluation in the global AI sector, drawing parallels to the early internet era. The dot-com surge, he argued, was considered a bubble only by those evaluating it over a short time horizon.
“If your time frame was five years, then the word bubble was right. If your time frame was 20, then the word bubble was wrong,” he said, noting how internet companies ultimately reshaped the world economy.
