Artificial intelligence has been called the engine of the new global economy. Yet while AI feels virtual, its foundations are intensely physical. Every algorithm, every chatbot, every digital assistant depends on a vast ecosystem of chips, data centres and power grids. And as the world races to deploy more AI, these physical foundations are fast becoming the new frontiers of global competition.

Put simply, chips are the new oil – concentrated in just a few countries. Data centres are the new refineries – processing the digital crude. And power grids are the new pipelines, carrying the energy that fuels intelligence. Every AI model depends on a global chain of energy, minerals and capital. That means the future of AI will be shaped as much by energy policy and infrastructure as by algorithms and code.

Globally, AI’s growth is accelerating at a breath-taking pace. Goldman Sachs estimates that it could add $7 trillion to global gross domestic product over the next decade. McKinsey suggests it may drive up to 40 per cent of future economic growth in advanced economies. But this transformation comes with a surge in energy demand: the International Energy Agency forecasts that AI and data centres together could consume more than 8 per cent of global electricity by 2030, up from about 2 per cent today. Each hyperscale data centre already uses as much electricity as a small city.

AI’s hunger for power is already reshaping energy flows – from natural gas and