AI Investment's Economic Ripple Effects Smaller Than Expected

The artificial intelligence investment surge — projected to reach $600 billion in 2024 — represents a massive reallocation of capital that many assume must come at the expense of other economic activity. A new analysis from Goldman Sachs suggests that assumption overstates the reality.

Why it matters

Business leaders navigating capital allocation decisions need accurate data on how AI spending affects the broader investment landscape. The finding that crowding-out effects remain modest suggests companies can pursue AI initiatives without triggering severe capital constraints across other sectors — at least for now.

Where the displacement occurs

Goldman Sachs economists Jessica Rindels and David Mericle identified three channels through which AI investment displaces other economic activity, though none prove as significant as market commentary often suggests.

The most direct impact appears in corporate technology budgets. Companies facing substantial new costs for AI services and tokens are trimming spending on other software and technology. However, this represents a reallocation within the tech sector rather than a net reduction in economic activity.

Data center construction presents a clearer case of resource competition. Gross margins on data center projects exceed twice those of non-tech construction, creating powerful incentives for contractors to prioritize AI infrastructure over other building activity. Construction labor and equipment have shifted accordingly.

The third channel involves corporate debt markets. Hyperscalers' enormous capital requirements have driven a surge in AI-related bond issuance, theoretically raising borrowing costs for other companies. The Goldman analysis finds this effect has been surprisingly limited, increasing corporate borrowing costs by just 0.05 percentage points and potentially reducing non-AI investment by approximately $10 billion.

The scale in context

AI investment now accounts for roughly 2% of U.S. GDP, representing 10% of total business fixed investment and 15% of equipment investment. These figures confirm AI's significance while also establishing boundaries on its economic footprint.

The relatively contained crowding-out effects reflect several factors. The U.S. economy's size and diversity allow it to absorb substantial sector-specific investment surges. Additionally, much AI spending represents new categories of expenditure rather than simple substitution for existing investments.

The bottom line

Rindels and Mericle conclude that "both claims are exaggerated" — the assertion that AI makes very large contributions to GDP growth and the concern that it crowds out substantial other activity. The reality appears more nuanced, with AI investment creating targeted displacement in specific markets while leaving broader investment patterns largely intact.

These findings were first reported by Axios, based on the Goldman Sachs economics team's research note.