• Global data centre capital expenditure could rise from $800 billion in 2026 to $1.8 trillion annually by 2050.
• The US is projected to attract 48% of cumulative investment, while Asia Pacific could secure $8.2 trillion.
• Power availability, chip trade restrictions and digital sovereignty policies will determine which markets capture the investment.

Power and chips redraw the investment map

From the United States to Asia Pacific, the race to secure computing power is developing into one of the largest infrastructure investment cycles in history.

Global investment in artificial intelligence infrastructure is projected to reach $31.6 trillion through 2050, according to PwC’s Global Data Centre Outlook. Annual capital expenditure could more than double, climbing from about $800 billion in 2026 to $1.8 trillion by mid-century.

The scale of that spending will connect technology policy with energy security, real estate, supply chains and sustainable finance. For corporate leaders and investors, data centres can no longer be assessed as conventional property assets. Their performance will depend on access to advanced chips, reliable electricity, regulatory certainty and local support.

The US is expected to attract $15.1 trillion, equal to 48% of projected global investment. Its position at the centre of the advanced-chip ecosystem gives it a powerful advantage.

Asia Pacific could account for $8.2 trillion, led by China and India. Europe and the Middle East are also increasing investment as governments pursue sovereign AI capacity and seek greater control over critical digital infrastructure.

Equipment drives the capital cycle

Unlike infrastructure expansions that slow after construction, AI investment is expected to continue rising. Chips and other information and communications technology equipment must be replaced or upgraded every few years.

ICT equipment currently represents about 70% of AI infrastructure investment. PwC expects that share to reach 93% by 2050. The shift places recurring technology expenditure, rather than building construction, at the centre of long-term capital planning.

That structure creates a complex risk profile. Hardware can become obsolete quickly, while demand forecasts remain sensitive to advances in AI models and computing efficiency. Investors must also account for supply concentration and changing export controls.

“AI infrastructure is becoming one of the defining capital allocation challenges of the next generation. It cuts across technology, energy, real estate, supply chains, regulation and financing. This changes how infrastructure investors need to think about capital requirements, risk and returns.” Clara Cutajar, Global Infrastructure Leader, PwC Australia

RELATED ARTICLE: PwC Joins PCAF Global Accredited Partner Program

Electricity becomes the decisive constraint

PwC identifies power as the leading factor influencing where AI infrastructure capital will flow. Data centre operators need electricity that is affordable, reliable and increasingly low carbon. Many markets will struggle to provide all three at sufficient scale.

Connectivity, security, policy certainty, community consent and access to graphics processing units will also shape investment decisions. These factors place utilities, grid operators and governments directly inside the AI investment equation.

The climate implications are substantial. Rapid electricity demand could delay emissions targets where grids remain dependent on fossil fuels. Markets with abundant renewable generation, firm low-carbon power and faster grid connections may gain a competitive advantage.

For policymakers, attracting data centres will require clearer planning rules and credible energy strategies. Community concerns over land, water and power allocation will also require early engagement.

Trade policy could remove US$6 trillion

PwC tested how trade restrictions and digital sovereignty could alter the outlook. Under a scenario involving tighter chip export controls, annual investment falls to about half the baseline forecast by 2030. Spending later recovers as supply chains adapt.

Even after that adjustment, cumulative investment reaches only $25.5 trillion through 2050. That is about US$6 trillion below the central projection, a reduction of nearly 20%.

A stronger push for digital sovereignty has a different effect. Global spending declines only slightly, but capital moves toward countries with strong domestic demand and limited existing capacity. Governments and regulated industries become important buyers as they prioritise local data storage and trusted infrastructure.

“The AI buildout is not a rising tide that will naturally lift all boats. Capturing this investment requires active positioning. Investors should recognize data centres as hybrid assets with a complicated risk profile.” Clara Cutajar, Global Infrastructure Leader, PwC Australia

Covering 46 countries and territories, the outlook extends capital forecasts to 2050. Its central message for executives is clear: the AI economy will be built where technology access, energy policy and public consent align. Those choices will influence global competitiveness and determine whether the infrastructure boom supports, or complicates, climate commitments.