AI's economic boost isn't showing up in GDP, and Goldman says that's a $115 billion blind spot

Goldman Sachs analysts report that AI's economic contribution to the US economy is significantly understated in official GDP figures, with an estimated $160 billion in AI-driven economic activity since 2022 only partially reflected as $45 billion. This methodological gap primarily arises because the Bureau of Economic Analysis treats high-performance semiconductors as intermediate inputs rather than investment in intangible assets like AI models, obscuring substantial spending on AI infrastructure and development and suggesting a larger, currently unmeasured, impact on national economic growth.

Analysis

A Goldman Sachs analysis reveals a significant disparity between the real economic impact of Artificial Intelligence and its measurement in official US GDP statistics. Since 2022, AI has contributed an estimated $160 billion to US economic activity, equivalent to 0.7% of GDP, yet only $45 billion, or 0.2% of GDP, is reflected in government data, leaving approximately $115 billion uncounted. This discrepancy stems from the Bureau of Economic Analysis's methodology, which classifies high-performance semiconductors as intermediate inputs, meaning their value is deducted from GDP when imported and their use in building AI systems is not captured as investment. The report posits that this spending is creating intangible assets, with an estimated $75 billion in uncounted investment for AI model development and enterprise cloud solutions. Complicating the data, a surge in server imports in the first half of 2025, likely driven by frontloading ahead of tariffs, has exaggerated normal investment demand. This macroeconomic measurement challenge is mirrored at the corporate level, where a record share of S&P 500 firms mention AI but few are able to quantify its impact on earnings, suggesting the true scale of the AI-driven investment cycle is substantially larger than headline data indicates.

Key Decisions for Investors

• Investors should recognize that headline GDP figures likely understate the true economic impact of the AI boom, suggesting the underlying economic strength and investment cycle may be more robust than currently reported.
• The significant, uncounted investment in AI infrastructure reinforces the fundamental growth narrative for companies in the AI value chain, particularly in semiconductors and cloud services, as their revenues represent real economic activity not fully captured in macro data.
• Exercise caution and prioritize bottom-up analysis, as the current inability of most companies to quantify AI's direct impact on earnings, coupled with data distortions from trade policies, introduces uncertainty and makes company-specific validation of the AI theme critical.