The world is about to embark on an infrastructure buildout that could dwarf anything seen before. According to a new analysis from PwC, global spending on data centers—the physical backbone of artificial intelligence—could reach $31.6 trillion by 2050. That's more than the entire annual output of the U.S. economy, and it could climb even higher if AI adoption accelerates.
PwC modeled spending across 46 countries, factoring in the needs of consumers, companies, and governments all racing to deploy AI. The baseline estimate of $31.6 trillion is already staggering, but the report also outlines a hotter scenario where the bill could reach $50 trillion. To put that in perspective, the buildout would rival—and likely surpass—the investments that defined previous technological eras, including railways, electrification, and the internet.
Why data centers are the new engine room
Data centers are the physical facilities that house the servers and computing power required to run AI models. As AI tools become more integrated into everyday life—from chatbots to autonomous vehicles to industrial automation—the demand for computing capacity is exploding. That means more data centers, more power, and more cooling systems, all of which require massive capital investment.
The scale of the projected spending is hard to grasp. $31.6 trillion is roughly equivalent to the combined GDP of the world's largest economies. It's a number that underscores how central AI infrastructure has become to the global economy. But it also raises questions about where the money will come from and whether the returns will justify the outlay.
Investors have already seen signs of this spending spree. Tech giants and cloud providers are pouring billions into new data centers, and the demand for power gear, not just chips, has become a key bottleneck. The race to build out AI infrastructure is also affecting borrowing costs, as companies and governments issue debt to fund these projects.
What it means for investors
For everyday investors, the PwC forecast is a reminder that the AI boom is not just about software and algorithms—it's about physical infrastructure. Companies that build, equip, and power data centers could see sustained demand for years. That includes firms involved in construction, electrical equipment, cooling systems, and renewable energy, as well as the semiconductor makers that supply the chips.
But the scale of the investment also carries risks. A $31.6 trillion buildout implies a massive allocation of capital that could crowd out other investments or strain supply chains. It also raises the stakes for AI adoption: if the technology fails to deliver the expected productivity gains, some of that spending could be wasted. Historically, infrastructure booms have often led to overcapacity and eventual busts, as seen in the dot-com era's fiber-optic glut.
Investors should also consider the broader economic backdrop. The projected spending comes at a time when borrowing costs are already rising in many parts of the world. Higher interest rates make it more expensive to finance large projects, which could slow the pace of data center construction or push companies to seek alternative funding sources.
For those looking to position their portfolios, the key is to focus on companies with strong balance sheets and clear exposure to the AI infrastructure trend. But it's also important to remember that forecasts like PwC's are just that—projections. They assume a certain trajectory of AI adoption, and that trajectory could change. The actual spending could be higher or lower, depending on technological breakthroughs, regulatory decisions, and economic conditions.
In the meantime, the data center buildout is already underway. Global markets are watching closely, and the ripple effects are being felt across industries. As PwC's report makes clear, the AI era is not just about software—it's about the physical infrastructure that will power it for decades to come.
