According to Morgan Stanley's latest analysis, the burgeoning AI infrastructure market is on track to experience a dire electricity shortfall by 2028. The research firm estimated that US data-centre developers could face a 32-gigawatt power shortfall through 2028, even after accounting for additional power-generation solutions. This represents around 34% of projected power requirements and threatens data-center development and chip deployment across the industry. As reported by Morgan Stanley, "Nvidia confirmed our thesis that there is not enough power to meet demand, and there are more semiconductors being built than there are places to put them." The assessment comes as rapid investment in AI infrastructure threatens to outpace the availability of electricity, raising concerns that smaller companies could struggle to compete with financially stronger technology giants.
The power shortage is creating significant market concentration, with Morgan Stanley estimating that major technology companies and specialist cloud providers could account for approximately 94% of global data-centre capacity additions in 2026. This concentration is expected to continue, with 64% in 2027 and 57% in 2028. As reported by Morgan Stanley, "We believe we're in an environment where data center energization is being concentrated across a handful of large, credit-worthy players," potentially squeezing out smaller developers over time. The research noted that "companies with strong credit support and large power requirements were securing electricity more quickly, potentially leaving smaller developers at a disadvantage." Morgan Stanley believes this could potentially "squeeze out" or put pressure on other players over time.
The power constraints could significantly impact semiconductor manufacturers if data centres are unable to install the chips they have ordered. According to Morgan Stanley's report, "If there's chip capacity expected to come online that cannot be deployed at data centers, we could see customers push out delivery or start canceling orders." However, the research expects Nvidia and Broadcom to meet their 2027 forecasts, citing their visibility into customer demand and infrastructure availability. The report warns that "greater risks for suppliers of memory chips, optical equipment and other supporting components if deployment slows." The findings suggest that expanding chip production alone may not be enough to sustain the AI boom unless companies can also secure electricity, suitable land and operational data centres.
The infrastructure challenge could encourage technology companies to expand computing infrastructure outside the United States, including in Asian markets such as India. Morgan Stanley's Asia data-centre team estimates that 14 GW of demand from US hyperscalers could shift to Asia through 2030, although the report does not specify India's share. However, the research warns that "while we expect international expansion to continue, we don't see it as a solution to the US power problem." The shortage could encourage technology companies to expand computing infrastructure outside the United States, including in Asian markets such as India, as the infrastructure challenge could persist beyond 2028.
The findings underline a growing challenge for the AI industry: technological advances and spending on chips may not translate into additional computing capacity unless electricity supply and supporting infrastructure expand at a similar pace. Morgan Stanley's report indicates that the infrastructure shortage could persist beyond 2028, with the projected cumulative US data-centre power shortfall potentially widening to 68 GW by 2029. The global race to expand AI capabilities is increasingly determined by which companies can secure enough electricity, financing and data-centre infrastructure to operate advanced chips, rather than simply by their ability to acquire the latest technology. Access to electricity was increasingly favouring large companies with strong financial backing and predictable revenues, with the shortage potentially affecting smaller players' ability to compete in the AI market.
