Brookfield CEO Bruce Flatt warns AI race is slowing due to infrastructure gaps

The AI boom has a speed limit, and it’s not set by software. It’s set by concrete, copper wire, and electrical grids that take the better part of a decade to upgrade.

That was the core message from Brookfield Corp. CEO Bruce Flatt during the firm’s annual Investor Day on September 17, where he told investors that the AI industry is decelerating because developers simply cannot build infrastructure fast enough to keep pace with demand. The bottleneck isn’t a lack of ambition or capital. It’s a lack of power.

The $7 trillion problem

Flatt’s framing was blunt: even if AI companies wanted to slow down voluntarily, physical constraints would force their hand regardless. He pointed to power shortages as the single most acute constraint facing the industry today.

Brookfield estimates that realizing AI’s full potential will require more than $7 trillion in capital investment over the next decade. That figure spans data centers, power generation, networking, and compute infrastructure.

US AI data center power requirements are expected to exceed 100 gigawatts within the next decade. Meanwhile, grid interconnection queues, the line companies must join to connect new facilities to the electrical grid, average between six and ten years.

Hyperscaler spending is surging, but not fast enough

Total capital expenditure guidance from major hyperscalers is projected to reach approximately $760 billion in 2026. That’s up from around $200 billion in 2022, nearly a fourfold increase in just four years.

Those numbers reflect the urgency that companies like Microsoft, Google, Amazon, and Meta feel about scaling their AI capabilities. But Flatt’s argument is that even this torrent of spending isn’t sufficient when the physical infrastructure simply doesn’t exist yet and takes years to build.

Brookfield itself is positioning aggressively in this space. The firm recently secured a $2 billion commitment from Nvidia as an anchor investor in its AI infrastructure fund.

Why this matters beyond the data center

Flatt’s comments carry weight because Brookfield isn’t a tech company making speculative projections. It’s one of the largest alternative asset managers in the world, with deep exposure to infrastructure, renewable energy, and real estate.

For policymakers, the six-to-ten-year interconnection queue is a glaring problem. Permitting reform and grid modernization have become de facto AI policy issues, whether regulators frame them that way or not.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.