Goldman Sachs: US Cloud Giants Need $300B Annual AI Revenue to Break Even
According to a Goldman Sachs report, the top five US tech giants—Amazon, Alphabet, Microsoft, Oracle, and Meta—are projected to spend $1.2 trillion on AI infrastructure next year. To break even on this investment, these companies will collectively need to generate around $300 billion in annual AI revenue over the coming years. The report highlights growing financial scrutiny surrounding the massive capital expenditure on artificial intelligence and its long-term return on investment. As infrastructure spending currently exceeds operational cash flows, tech giants are increasingly forced to rely on debt and equity financing while navigating energy and supply constraints. Goldman Sachs estimates capital expenditure growth across the five companies will slow from nearly 100% this year to 54% next year, eventually tapering to 12% by 2028 when cumulative spending reaches $1.4 trillion. Furthermore, analysts calculate that end-users will need to spend nearly $1 trillion annually on AI applications for software providers to achieve solid profit margins after paying cloud computing costs.
## BACKGROUND
Hyperscalers are massive cloud infrastructure providers capable of operating scale-out data centers worldwide. Developing and deploying modern artificial intelligence models requires immense computing power, driving record investments in hardware like graphics processing units (GPUs), data center infrastructure, and energy supply.