Bruce Flatt says AI developers cannot build enough data centers and compute to meet demand, making a slower rollout inevitable.
The artificial intelligence race is already slowing because developers cannot build infrastructure fast enough to meet the industry’s expectations, according to Brookfield CEO Bruce Flatt.
Flatt said AI companies are demanding more data centers and compute than the market can currently deliver. “We as an industry can’t build enough,” he said at Brookfield’s annual investor day. “We can’t even build a fraction of what everyone thinks they need.”
His comments come as executives including Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman call for a slower pace of AI development. Their arguments focus on unpredictable risks and the possibility that advanced models could move beyond effective human control.
Flatt said the infrastructure constraint existed even before those safety concerns became central to the debate. In his view, the technology is slowing anyway because the physical capacity required to run it cannot be built quickly enough.
Brookfield has estimated that AI infrastructure will require more than $7 trillion in capital investment over the next decade. Flatt said a slower rollout could benefit the sector by bringing more discipline to spending, construction, and deployment.
The comments underscore that AI’s pace will be shaped not only by model capabilities and safety decisions, but also by the practical limits of power, data centers, and computing capacity.
