AI infrastructure spending to hit $1.5 trillion a year by 2031

Justifying investment requires $6 trillion in annual revenue

Existing AI market seen generating only $1.8 trillion

Robots, new drugs among sectors needed to close $4.2 trillion gap

Microsoft. [Reuters]
Microsoft. [Reuters]

The world's biggest technology companies are pouring vast sums into AI data centers, but a new analysis warns that the industry will need to generate $6 trillion in annual revenue by 2031 to justify the scale of investment — and that roughly $4.2 trillion of that will have to come from markets that do not yet exist.

According to Bain & Company's annual global technology report released Tuesday, major tech firms including Microsoft, Google, Amazon, Meta and Oracle are rapidly expanding data center investment to meet the enormous computing demands of AI.

If that trend continues, Bain projects cumulative data center-related spending will reach between $5 trillion and $6.5 trillion by 2030.

The central question, the report argues, is whether AI can actually generate enough revenue to match the investment being made.

Bain estimates that by 2031, annual spending on AI infrastructure — including data center construction, expanded computing capacity, AI accelerator upgrades and memory chip procurement — could reach as much as $1.5 trillion.

To sustain that level of spending while delivering adequate returns to investors, the end market for AI services and applications would need to produce roughly four times that amount, or $6 trillion in annual revenue, the consultancy said.

Existing AI markets are unlikely to close that gap on their own. Bain estimates that current consumer and enterprise AI services could generate at most $1.8 trillion in annual revenue by 2031 — just 30 percent of the $6 trillion target.

The remaining $4.2 trillion would have to come from entirely new sectors — a market more than twice the size of the AI industry as it stands today. Without that growth, the economics of the data center investment race simply do not add up, the report said.

Bain identified several potential new revenue sources, including autonomous machines and robotics, drug development, mental health services and energy generation — all areas where AI remains in early stages.

Today's AI industry is largely focused on boosting employee productivity through chatbots, document drafting and coding assistance. But sustaining the current pace of infrastructure investment will require AI to move beyond cutting corporate costs and into generating trillions of dollars in entirely new revenue, the report said.

"To make the current pace of development sustainable, the AI industry needs a wave of innovation that dwarfs what mobile and cloud computing once delivered," said David Crawford, Bain's global head of technology, media and telecommunications.

Crawford added that AI infrastructure is being built far ahead of the demand curve, and that sustaining the current investment pace in a viable way would require an economic impact large enough to add roughly 1 percentage point to annual global GDP growth.

Data centers themselves are also scaling up rapidly. Bain found that data center size and costs are roughly doubling every 12 to 16 months, driven by rising prices for Nvidia AI accelerators, SK hynix memory chips, networking equipment and other components.

Power supply presents another challenge. The report projects that at least 150 GW of additional electricity capacity will need to be secured by 2031 to support current AI investment trends.


sjy@heraldcorp.com