The most frequently cited figure for AI investment is $800 billion in capital expenditures by hyperscalers in 2026. In a global economic analysis report released on August 2, Goldman Sachs noted that this figure underestimates global AI investment by approximately $200 billion and overestimates U.S.-only investment by about $200 billion. After adjusting for private companies, non-U.S. enterprises, and non-AI expenditures, Goldman Sachs estimates total global AI investment in 2026 to be approximately $1.019 trillion, with the U.S. accounting for about $581 billion. Three independent estimation methods yield highly consistent results, indicating that global AI investment is nearing $1 trillion. AI investment as a share of U.S. GDP is projected to rise from 1.8% in 2026 to 2.5% in 2027, and further to 2.8% in 2028.
The commonly cited market figure of $800 billion underestimates the global total of $200 billion and overestimates the U.S. total of $200 billion.
Goldman Sachs points out that the commonly used capex figure for hyperscalers—approximately $800 billion—has four issues. First, it overlooks substantial AI investments by U.S. private companies that play a critical role in the AI ecosystem. Second, it ignores AI investments by non-U.S. companies, particularly those in China and broader Asia. Third, hyperscalers had already invested over $150 billion in AI-related infrastructure prior to the current AI boom, and some of today’s capex is unrelated to AI. Fourth, U.S. hyperscalers operate globally, with a portion of their investments occurring outside the United States.
Goldman Sachs addressed each of the above issues by incorporating AI capital expenditures from other U.S. public and private companies, as well as spending by non-U.S. AI-related companies, on top of the hyperscale providers. It excluded existing investment levels from 2022, retaining only the incremental portion. Additionally, based on the geographic locations of hyperscale providers’ announced investment projects, the global total was allocated across countries and regions.
The adjusted figures show that global AI investment totals approximately $1.019 trillion, with the United States accounting for about $581 billion. The commonly cited figure of $800 billion underestimates global AI investment by approximately $200 billion and overestimates U.S.-only AI investment by approximately $200 billion. Goldman Sachs estimates that approximately 70% of U.S. hyperscaler capex is directed toward domestic projects, 15% toward Asia, and 9% toward Europe.
Corporate profit revisions and official trade data have confirmed the $1 trillion assessment.
Goldman Sachs cross-validated the above results using two additional methods. The first method is based on revisions to gross profit forecasts for publicly traded companies related to AI. AI investments will ultimately manifest as increased revenue and profits for upstream suppliers; by tracking the incremental growth in gross profits of AI-related companies relative to 2022 levels, the scale of AI investment can be inferred in reverse. This method estimates global AI investment in 2026 at approximately $1.06 trillion.
The second method is based on official national accounts and trade data. For the United States, Goldman Sachs employs the commodity flow approach, summing domestic production, net imports, and changes in inventories to estimate annualized AI hardware investment at approximately $500 billion through May 2026, plus approximately $100 billion in AI-related R&D and intellectual property investment, totaling around $600 billion. On a global scale, Goldman Sachs extrapolates global AI investment by correlating each country’s net AI-related imports with known data from countries such as the United States, arriving at an estimated global AI investment of approximately $1.002 trillion.
The figures from the three methods are highly consistent: approximately $1 trillion globally and nearly $600 billion in the United States. Goldman Sachs believes this data provides a reliable benchmark for the scale of AI investment. Cumulative global AI investment since 2022 is projected to reach approximately $1.8 trillion by the end of 2026.
AI investment as a share of GDP will rise to 2.8%, with continued resilience in the short term.
Goldman Sachs extrapolated AI investment pathways for 2027 and 2028 using consensus forecasts for public company capex. AI investment as a share of U.S. GDP is projected to rise from 1.8% in 2026 to 2.5% in 2027, and further to 2.8% in 2028. Globally, AI investment as a share of global GDP is expected to increase from 0.9% in 2026 to 1.3% in 2027, and to 1.4% in 2028. These levels are consistent with historical peaks in investment shares during infrastructure booms for general-purpose technologies, which ranged from 2% to 5%.
When AI investment growth will slow is the key uncertainty in today’s macro market. Goldman Sachs has compiled a set of leading indicators, including semiconductor manufacturing equipment imports, PMI-related components, import prices, memory purchases, and GPU rental prices. All leading indicators are currently at the upper end of their ranges since 2022, pointing to continued strong short-term growth.
However, based on trade data released earlier by economies such as Taiwan and South Korea, Goldman Sachs’ real-time forecasts indicate a modest slowdown in AI-related investment in June and July, following extremely high levels. Official U.S. data show that AI investment is increasingly driven by cost inflation rather than growth in actual investment volume. Approximately 8% of nominal AI investment growth in the first half of 2026 is attributable to price factors rather than increases in real investment volume. If this trend continues, AI investment’s contribution to real GDP growth in 2026 may be smaller than in 2025. Since semiconductor purchases are not classified as investment in U.S. national accounts and the high import content of AI hardware is excluded from GDP calculations, Goldman Sachs believes the overall impact of AI investment on U.S. GDP levels remains limited.
AI investment is approaching $1 trillion. The market has been discussing AI investment at $800 billion, but Goldman Sachs has revised this figure. Global AI investment is larger than the market realized, while U.S.-only AI investment is smaller than previously thought. The share of AI investment relative to GDP is entering the range typical of historical infrastructure cycles for general-purpose technologies. The very framework through which the market understands AI investment needs to be recalibrated.
Disclaimer: This article is a compilation and interpretation by Chaoxiang Research of a third-party brokerage research report (Goldman Sachs, August 2, 2026), combined with publicly available market information. The ratings, price targets, earnings forecasts, and related judgments cited herein reflect the views of the brokerage’s analysts and represent the position of their respective institution, not the views of Chaoxiang Research, nor do they constitute any investment advice. The market carries risks; decisions must be made independently. This article should not be used as a basis for buying or selling any securities.
Written by: Rita