Goldman Sachs analysis indicates that while US AI investment is projected to reach nearly $600 billion by 2026, its crowding-out effect on other investments will be limited, with an estimated incremental crowding-out effect of approximately $50 billion. The report found that large tech companies primarily finance AI projects by reducing stock buybacks and borrowing, rather than by significantly cutting other investments. Furthermore, the expansion of data center construction has coincided with a decline in subsidized manufacturing facility construction, freeing up resources. Goldman Sachs believes that although AI-related financing accounts for nearly a quarter of investment-grade bond issuance, its impact on other companies' borrowing costs is also limited.