AI race may hinge on power access and infrastructure, not just chips: Morgan Stanley

According to Morgan Stanley's latest analysis, the rapidly expanding AI infrastructure market is expected to face a severe electricity shortage by 2028. The predicted 32-gigawatt power deficit poses a significant challenge for data-center development and chip deployment, particularly affecting smaller enterprises attempting to compete with larger rivals.

The study suggests that while some global strategies may offer relief, the United States will continue to grapple with internal power constraints. Access to reliable electricity and robust infrastructure may ultimately prove more critical than chip supply in determining competitive advantage in the AI race.

The warning underscores growing concerns within the technology sector about whether existing power grids can support the exponential growth in computing demands driven by artificial intelligence development and deployment.