Analysts from PwC project that cumulative global investment in artificial intelligence will exceed $30 trillion by 2050. Researchers from Columbia University estimate that infrastructure spending in the US between 2025 and 2032 will reach $9 trillion, representing 3.22 percent of the country's annual GDP.
Market participants are investing heavily in AI infrastructure. Anthropic plans to spend $518 billion over the next few years, a figure exceeding its previous annual revenue by more than 100 times. Meanwhile, Bain and Company suggests that hyperscalers must generate $4.2 trillion in new revenue over the next five years to justify current capital expenditures.
Financial institutions remain cautious. JP Morgan notes that the economic return on AI implementation remains unclear, and current market valuations would require US labor productivity to grow by 3–5 percent annually for a decade. Official government forecasts currently anticipate a growth rate of only 1.75 percent.
Experts from Columbia University calculate that to provide investors with a 10 percent annual return, the US AI sector would need to generate an average annual revenue of $3.55 trillion through 2032. Current revenue levels in the industry remain significantly lower than these projections.
Anthropic estimates that the impact of AI on US economic growth by 2030 could range from 2.4 percent in a moderate scenario to 15.4 percent in an aggressive one. Analysts warn that historical technological booms often conclude before infrastructure investments yield financial returns, which can potentially trigger economic crises.
