AI Compute Bottleneck Driven by Power and Labor Constraints

⚡ AI Executive Summary

Morgan Stanley's research indicates that despite accelerating corporate adoption of artificial intelligence, the sector faces significant supply limitations stemming from insufficient computing capacity. While 25% of S&P 500 companies now report measurable AI returns—up from 14% a year prior—the ability to scale remains constrained by structural challenges. For the power sector, this represents a critical emerging demand driver that will sustain elevated electricity consumption for years. The 10–20% power deficit Weaver identifies suggests utilities must plan aggressively for data-center load growth, particularly in regions competing to host new facilities. Grid operators should expect sustained pressure on peak capacity and transmission infrastructure, while regulators face mounting political pressure to balance development with environmental and rate concerns. This dynamic will likely reshape investment priorities in generation, storage, and transmission—making data-center-ready infrastructure a competitive advantage for forward-thinking regional grids.