America Spent $300 Billion on AI Last Year. Goldman Sachs Says It Added Nothing to the Economy.

Source: Dev.to

Introduction

The most expensive technology bet in corporate history has a GDP contribution of approximately zero. Goldman Sachs Chief Economist Jan Hatzius told clients this week that massive AI investment spending contributed “basically zero” to U.S. economic growth in 2025. Morgan Stanley reached the same conclusion independently. Two of Wall Street’s most influential economics desks agree: the AI boom is not showing up in the numbers that matter.

AI Investment and GDP Impact

• Spending vs. Output: When an American company spends $50 billion on Nvidia GPUs, the majority of that money flows to foreign manufacturers. In GDP accounting, domestic spending on imported goods is subtracted from output. The investment appears on corporate balance sheets but does not show up in GDP.
• Scale of Spending: Meta committed $65 billion to AI infrastructure in 2026, Microsoft $80 billion, and Amazon, Google, and Oracle each committed tens of billions more.
• Capital Expenditure vs. GDP: Goldman estimates AI investment will contribute roughly 1.5 percentage points to U.S. capital‑expenditure growth in 2026, but the net GDP impact will be only 0.1–0.2 percentage points. Almost all of the impact leaks offshore.

Hardware Supply Chain

• Import Dependence: About three‑quarters of AI data‑center costs go toward chips and equipment manufactured in Asia—primarily TSMC in Taiwan, SK Hynix and Samsung in South Korea, and assembly operations across Southeast Asia.
• Reshoring Efforts: The CHIPS Act allocated $52 billion to reshoring semiconductor manufacturing, yet total AI‑chip demand in 2026 alone will exceed $200 billion.
• Supply Constraints: SK Hynix announced its entire 2026 production of HBM4 memory chips is already sold out and plans to increase advanced DRAM capacity eightfold, from 20,000 to 190,000 wafers per month. Prices will keep rising, and the revenue flows to South Korea, not the U.S.

Corporate Spending vs. Domestic Production

The companies spending the money are American, but the factories collecting the money are not. The political narrative that AI drives American economic growth rests on a category error: growth is driven by domestic production, while AI currently drives demand for Taiwanese semiconductors.

Productivity Gains Timeline

Goldman expects meaningful productivity gains from AI to begin appearing in 2027 and to compound through the late 2030s. The actual economic case for AI is not the spending itself but the eventual output gains from workers using AI tools. At current adoption rates, those gains remain theoretical.

Industry Adoption and ROI

• Adoption Rates: McKinsey’s most recent survey found 72 % of companies have adopted AI in some form, up from 55 % a year earlier.
• ROI Reality: MIT’s Project NANDA found 95 % of custom enterprise AI deployments deliver zero measurable ROI. Anthropic’s research shows developers using AI code assistants scored 17 % lower on code‑comprehension tests. The tools are everywhere; the output gains are not.

Historical Parallel

Hatzius draws an explicit parallel to the late‑1990s internet investment cycle. Companies spent aggressively on fiber‑optic cable, web servers, and e‑commerce platforms for years before productivity statistics moved. The dot‑com crash intervened. Productivity gains eventually materialized between 2003 and 2007, but they arrived for different companies than those that spent the money.

Economic Implications

• Trade Balance Issue: The situation is not a technology problem; it is a trade‑balance problem wearing a technology costume.
• Short‑Term Outlook: America is financing the most expensive infrastructure build‑out in history, but the infrastructure is being built with foreign components. GDP measures what gets produced domestically, and what gets produced domestically is largely the electricity bill.

Conclusion

Goldman Sachs is not saying AI is a bad investment. It is saying the investment and the economic growth are happening in different countries. The companies spending the money are American; the factories collecting the money are not. The economic returns, if they arrive, will do so years from now, while current GDP figures remain largely unaffected.