Goldman Sachs: AI Added 'Basically Zero' to US Economic Growth in 2025

The Sobering Assessment

Goldman Sachs has published an analysis concluding that AI technology contributed "basically zero" to US economic growth in 2025, despite the massive wave of AI investment and widespread hype surrounding the technology. The report challenges the prevailing narrative that AI is already driving significant productivity gains across the economy.

Investment vs. Output

The disconnect is striking: major AI companies including OpenAI, Anthropic, Google, and Meta have attracted hundreds of billions in investment. GPU shortages, datacenter buildouts, and AI integration projects have dominated corporate agendas. Yet the macroeconomic data shows little evidence of a productivity boost that would register in GDP statistics.

Goldman's analysis points out that translating technological investment into measurable economic output takes time — and AI may still be in the early adoption phase where businesses are acquiring tools but haven't yet restructured workflows to realize gains.

Historical Context

Economists often point to the "productivity paradox" from the PC era, where computers were everywhere but productivity statistics showed no improvement for years. The same pattern may be playing out with AI: the transformation may be underway, but not yet visible in the aggregate data.

Counterarguments

Some experts argue Goldman's analysis is too narrowly focused on short-term GDP metrics. AI's impact may be more diffuse — showing up in healthcare outcomes, scientific research, drug discovery — before registering in traditional economic measures. The debate about when (and whether) AI's economic impact will become undeniable continues to intensify.