Goldman Sachs says AI delivered 'basically zero' GDP growth despite massive tech spending

Goldman Sachs Challenges AI Investment Narrative

The AI economic impact that tech companies promised has failed to materialize in measurable terms, according to Goldman Sachs chief economist Jan Hatzius. Speaking with the Atlantic Council, Hatzius stated that AI investment contributed "basically zero" to US GDP growth in 2025, directly challenging the widespread belief that massive tech spending is propping up the American economy 1

1

. "We don't actually view AI investment as strongly growth positive," Hatzius explained. "We think there's been a lot of misreporting of the impact that AI investment had on GDP growth in 2025, and it's much smaller than it's often perceived" 1

1

.

Economic analyst Joseph Politano calculated that of the US economy's 2.2 percent growth in 2025, only 0.2% likely came from AI investment 1

1

. This minimal contribution stands in stark contrast to the AI hype dominating boardrooms and earnings calls, where 70% of S&P 500 management teams discussed AI in their quarterly reports 5

5

.

Overseas Manufacturing Drains US Economic Benefits

The core issue behind basically zero GDP growth from AI lies in where the money actually flows. Roughly three-quarters of AI infrastructure spending goes toward computing components manufactured overseas, primarily by TSMC in Taiwan 1

1

4

4

. When US tech companies invest in data centers, they're effectively purchasing imported components that boost Taiwanese and Korean GDP rather than American economic output.

"A lot of the AI investment that we're seeing in the U.S. adds to Taiwanese GDP, and it adds to Korean GDP but not really that much to U.S. GDP," Hatzius noted 3

3

. While Nvidia remains headquartered in the US, its manufacturing happens elsewhere, and semiconductors represent the major component in capital expenditure for AI data centers 1

1

.

The five top US tech companies are collectively expected to spend as much as $700 billion on AI infrastructure in 2026 1

1

. IDC research predicts AI infrastructure spending could reach $758 billion by 2029, up from $82 billion in the last full calendar quarter 4

4

. Yet this massive capital expenditure for AI primarily benefits Asian manufacturing economies rather than generating domestic economic growth.

AI and Productivity Show Disconnected Results

Goldman Sachs found "no meaningful relationship between AI and productivity at the economywide level," according to senior US economist Ronnie Walker's analysis of fourth-quarter earnings 5

5

. While 54% of S&P 500 management teams framed AI around productivity and efficiency during earnings calls, only 10% quantified its impact on specific use cases, and a mere 1% quantified its impact on earnings 5

5

.

However, companies that successfully measured AI implementation reported median productivity gains of around 30% for two specific, localized use cases 5

5

. This suggests AI adoption delivers tangible benefits at the task level but hasn't yet translated into economy-wide impact. US Census survey data indicates fewer than 20% of establishments currently utilize AI for any business functions 5

5

.

A 2023 Goldman Sachs Research report forecasted AI beginning to have measurable impact on labor productivity in 2027, potentially increasing US productivity growth by 1.5 percentage points annually with widespread AI adoption over a 10-year period 3

3

. Long-term productivity gains remain possible, but the gap between current reality and future promises continues to widen.

Growing Concerns About AI Market Bubble

The disconnect between massive spending and minimal economic returns fuels growing warnings about an AI market bubble. OpenAI remains the biggest capital-burning company in history, with revised estimates showing capital expenditure on AI infrastructure reaching $600 billion by 2030 and potentially $1.4 trillion by 2033 1

1

. Yet the company's entire revenue for 2025 was less than $20 billion 1

1

.

J.P. Morgan claimed AI needed to generate over $600 billion in annual revenue just to achieve a 10% return on infrastructure expenditures 1

1

. Analysts at Morgan Stanley, JPMorgan Chase, and other major financial institutions have expressed similar concerns that technology sector growth may be indirectly benefiting Asian manufacturing economies more than the US 2

2

.

Market fervor around AI continues despite these economic realities. Tax advisor Joe Brusuelas acknowledged that AI's economic effects remain difficult to estimate, with everyone "trying to peer through the fog to understand what is driving growth" 2

2

. The challenge for investors and policymakers lies in distinguishing between AI's genuine long-term potential and the current mismatch between investment scale and measurable returns. Job displacement concerns also loom, with Goldman Sachs estimating AI could displace 6-7% of the US workforce if widely adopted, though new job opportunities may eventually emerge 3

3

.