AI Needs $6 Trillion in Annual Revenue by 2031 to Support Its Data Center Boom

Bain & Company estimates that the AI industry would need about $6 trillion in annual revenue by 2031 to support the scale of infrastructure investment it projects. Existing consumer and enterprise AI services could account for up to $1.8 trillion, leaving roughly $4.2 trillion to come from new sources of revenue. These are forecasts, not guaranteed outcomes. 35

How Bain arrives at the $6 trillion estimate

Bain projects annual AI infrastructure spending could reach $1.5 trillion by 2031, covering new data centers and computing capacity as well as upgrades to GPUs, memory, and networking equipment. Its estimate assumes capital expenditure would equal about 25% of industry revenue; on that basis, supporting the projected investment would require an AI market approaching $6 trillion annually. 3

That is an assumption-based revenue threshold—not a prediction that the industry will necessarily reach it. The estimate depends on both the scale of investment and the share of revenue directed to capital spending. 3

Existing AI services leave a large gap

Bain’s report identifies consumer AI—such as subscription and advertising products—and enterprise AI applications, including software development, sales, marketing, customer service, and IT, as sources of future revenue. Together, existing applications could provide up to $1.8 trillion, according to Bain, but would not close the full gap. 35

The remaining opportunity therefore depends on new AI-enabled products and markets, rather than productivity improvements alone. Bain-related reporting names possibilities including autonomous machines, robotics, drug discovery, mental health, and energy generation; these are potential areas, not assured revenue sources. 517

The buildout is large—and faces delivery constraints

Bain separately estimates $5 trillion to $6.5 trillion in data-center buildout by 2030, adding about 150 gigawatts or more of capacity. It says that would nearly triple global capacity over five years. The buildout figure is a cumulative investment estimate, while the $1.5 trillion figure is projected annual AI-infrastructure spending by 2031; they describe different measures and time periods. 23

Bain also warns that power, chips, skilled labor, and permitting are constrained at the same time. Those bottlenecks could delay construction and the arrival of capacity. Meanwhile, the revenue case rests on new applications becoming substantial businesses quickly enough to support the projected investment. The available forecasts do not establish whether that timing will work out. 23

The central question: can new AI markets scale in time?

Bain’s figures frame a two-sided challenge: the industry must build infrastructure while developing new products that can generate revenue at a much larger scale than today’s consumer and enterprise services. The $6 trillion estimate is best read as a test of the economics behind the buildout—not as a certainty that either the infrastructure or the new markets will materialize on schedule. 235