A new Bain & Company report reveals that $2 trillion in annual revenue will be required to fund AI computing globally by 2030. Even with savings from AI, the world faces an $800 billion shortfall to meet rising demand.
The report states global incremental AI compute needs may reach 200 gigawatts by 2030, with the US accounting for nearly half. Shifting IT budgets to cloud and reinvesting AI savings would still not cover the full costs. AI compute demand is growing more than twice as fast as Moore’s Law.
Experts warn that governments and leading providers are entering an AI arms race. Overbuilding or underbuilding infrastructure is a major challenge. Innovation, supply constraints, and algorithmic efficiency will be critical to meeting demand over the coming years, Bain said.
While some companies have scaled AI into core workflows, achieving 10-25% EBITDA gains, most remain in experimentation mode. Global supply chains face fragmentation due to tariffs, export controls, and sovereign AI initiatives. Experts note that AI is now a strategic advantage tied to economic, political, and military power.
