Goldman Sachs is projecting that global spending on artificial intelligence infrastructure will hit $7.5 trillion over the next five years. The forecast, released by the bank's research division, covers investments in data centers, networking, computing hardware, and other physical and digital systems needed to support AI workloads.
The scale of the forecast
The $7.5 trillion figure is a cumulative estimate for the period from 2025 through 2029. It reflects the bank's view that AI adoption will require massive upfront capital outlays, particularly from technology companies and cloud providers. The projection does not break down spending by region or sector, but it signals that the AI buildout will be one of the largest infrastructure cycles in modern history.
What's driving the spending
Goldman Sachs analysts point to the rapid expansion of generative AI and large language models as the primary catalyst. Training and running these models demand enormous computational power, which in turn requires new data centers, advanced chips, and high-speed networking. The bank expects that spending on AI infrastructure will accelerate as more industries integrate AI into their operations.
For investors, the forecast suggests sustained demand for companies that supply AI hardware, software, and services. However, the bank also warns that the spending could face headwinds, including supply chain constraints, energy costs, and regulatory scrutiny. The $7.5 trillion figure is a baseline estimate; actual spending could be higher or lower depending on technological breakthroughs and economic conditions.
What comes next
Goldman Sachs plans to update its AI infrastructure forecast quarterly as new data on capital expenditure and AI adoption becomes available. The next revision is expected in early 2026, which will give markets a clearer picture of whether the spending trajectory is on track.
