Goldman Sachs predicts global AI spending will hit $1tn by 2026, shaping the next phase of economic expansion

Goldman Sachs forecasts that worldwide spending on AI will reach approximately $1 trillion in 2026, highlighting the sector’s rapid growth and its significant impact on global GDP over the coming years.

Goldman Sachs expects global spending on artificial intelligence to reach about $1tn in 2026, underscoring how quickly the build-out of computing capacity, chips and data centres has become a major driver of investment. The bank said that AI-related outlays could equal 0.9% of global GDP next year, rising to 1.3% in 2027 and 1.4% in 2028, suggesting the cycle may stay expansionary for at least the next two years.

The estimate is wider than the commonly cited figure for US hyperscalers alone. Goldman Sachs said its calculation includes spending by private companies, non-hyperscaler firms and businesses outside the US, lifting the projected total to roughly $1tn in 2026, including about $581bn in the US. Axios reported that the bank’s economists also put US AI investment at around $600bn next year, equal to 2% of US GDP, 10% of business fixed investment and 15% of equipment investment, while arguing that the crowding-out effect on the wider economy is less severe than many critics suggest.

Goldman Sachs said the cumulative total invested in AI since 2022 could reach $1.8tn by the end of 2026, although it warned that the pace of future spending remains one of the main uncertainties for markets. The bank said its projections are supported by indicators such as semiconductor equipment imports in Taiwan and South Korea, purchasing managers’ surveys, memory prices and GPU rental rates, all of which are close to the upper end of their range since 2022. At the same time, it cautioned that the estimate is difficult to verify precisely and may involve some double-counting, particularly where companies do not separately disclose leased hardware and capital expenditure.