Goldman Sachs Projects AI Infrastructure Spending to Top $1 Trillion

Research from The Goldman Sachs Group, Inc. indicates that capital expenditures for AI infrastructure among hyperscalers are projected to exceed $1 trillion by 2027. The firm notes that these expenditures may surpass cash flow from operations between 2026 and 2028. Despite recent turbulence in AI stocks, the firm expects the broader market to stabilize as strong corporate earnings support the bull market.

Individual companies have reported significant spending increases. Alphabet Inc. raised its full-year guidance to between $195 billion and $205 billion and signaled further increases for 2027. Meta Platforms Incorporated tightened its 2026 spending range to $135 billion–$145 billion. These surges in spending have negatively impacted the stock prices of both Alphabet and Meta, whereas Microsoft Corporation maintains a stronger near-term free cash flow position.

AI infrastructure contributed approximately one-third of S&P 500 profit growth in the second quarter. Goldman Sachs projects this contribution could rise to over half of all growth through 2027, driven largely by Alphabet, Amazon, Microsoft Corporation, Nvidia, and Broadcom.