SKN CBBA -
SKN CBBA
Cross Border Banking Advisors
SKN | Goldman Sachs Reassesses the AI Investment Cycle as Spending Enters a New Phase

Finance

SKN | Goldman Sachs Reassesses the AI Investment Cycle as Spending Enters a New Phase

By Or Sushan

September 10, 2026

Key Takeaways:

  • Goldman Sachs expects the AI investment boom to eventually slow, even under a constructive long-term outlook for the technology.
  • Chief economist Jan Hatzius argues that the transition from infrastructure build-out to technology exploitation will naturally reduce investment volumes.
  • Goldman’s baseline view remains that AI investment is productive and sustainable, with the potential to support stronger productivity growth.
  • The bank is highlighting a critical distinction between a healthy investment cycle and assumptions that exceptionally high spending can continue indefinitely.

Goldman Sachs is taking a more measured view of the artificial intelligence investment boom, warning that the extraordinary pace of AI-related capital spending will eventually moderate. Chief economist Jan Hatzius does not characterize the current build-out as fundamentally unsustainable; rather, he argues that even a successful technology cycle inevitably moves from an intensive investment phase toward a period in which existing infrastructure generates economic returns.

Goldman Sachs Sees a Normalization, Not an End to AI Investment

Hatzius’s baseline assessment remains constructive. Goldman Sachs assumes that AI spending is largely sustainable and productive, with investment contributing to stronger productivity growth over time. The important qualification is that productive investment does not require investment volumes to keep accelerating indefinitely.

In Goldman’s framework, new technologies typically generate a substantial capital-allocation phase as companies build infrastructure, acquire equipment and establish capacity. Once that infrastructure is sufficiently developed, the economic emphasis shifts toward exploiting the technology rather than continuously expanding the physical investment base.

For Goldman Sachs, that transition represents a normal feature of technological development rather than evidence that the AI thesis has failed.

Why Goldman Is Watching the Investment Cycle Closely

The scale of the current infrastructure cycle makes the timing of that transition particularly important. PwC estimates global AI infrastructure investment could reach $31.6 trillion through 2050, while annual data-center capital expenditures are projected to increase from approximately $800 billion in 2026 to $1.8 trillion by 2050.

Goldman’s concern is therefore less about whether companies should invest in AI and more about what happens when the exceptionally high rate of investment begins to normalize. Businesses and markets that have been valued on the assumption of permanently accelerating capital expenditure could face adjustment when spending growth eventually slows.

The Goldman Sachs Signal for Capital Allocation

For sophisticated investors, the distinction between AI adoption and AI infrastructure spending is increasingly important. A moderation in capital expenditure would not necessarily mean declining AI usage or weaker productivity. It could instead indicate that companies are moving from building the infrastructure toward generating returns from infrastructure already deployed.

That distinction is particularly relevant for HNWI portfolios exposed to technology, private markets and businesses dependent on AI-related capital expenditure. Goldman Sachs is effectively signaling that the quality of future returns may matter more than the absolute size of future spending.

Hatzius also leaves open a more adverse scenario in which some AI investments prove unproductive. Goldman’s baseline remains optimistic, but acknowledging that risk reinforces the importance of distinguishing genuine productivity gains from capital spending that has yet to demonstrate durable economic returns.

Going forward, Goldman Sachs will be watching the transition from investment intensity to economic exploitation. For wealth holders, the key question is not whether the AI cycle ends, but which businesses and financial structures remain resilient when its exceptional investment phase eventually matures.

For a confidential discussion regarding your global technology exposure, cross-border banking structure or international wealth strategy, contact our senior advisory team.

Leave a Reply

Your email address will not be published. Required fields are marked *

More like this