You Won’t Believe How Much Revenue AI Titans Need To Haul In For Data Center Binge

Artificial intelligence companies will need to generate at least $4.2 trillion in new annual revenue by 2031 to financially support the industry’s massive infrastructure buildout, according to an analysis from Bain & Company.

The AI industry would need roughly $6 trillion in annual revenue by 2031 to support projected spending on data centers, chips and computing infrastructure, according to Bain. Uses that can be identified today are projected to generate only about $1.2 trillion to $1.8 trillion annually, leaving at least $4.2 trillion that would have to come from new products, robotics, autonomous systems and other applications.

The $6 trillion figure is not a forecast of how much revenue AI companies will actually generate. Rather, it represents Bain’s estimate of the revenue required to fund the amount of computing investment currently projected for the industry.

The enormous gap highlights the scale of the bet technology companies are making as they pour hundreds of billions of dollars into infrastructure before many of the products expected to eventually pay for it have been developed.

Bain estimates existing consumer AI applications could generate roughly $200 billion to $400 billion annually, while enterprise uses could contribute approximately $1 trillion to $1.4 trillion. Even at the high end of those estimates, existing applications would account for less than one-third of the $6 trillion needed.

The remainder would have to come from areas including search and advertising, autonomous systems, physical AI and robotics, along with trillions of dollars in revenue from products that have not yet been developed, according to Bain’s analysis.

The revenue challenge comes as Big Tech is rapidly increasing both its spending and borrowing to finance the AI race.

Combined capital expenditures by six major hyperscalers are expected to exceed $1.3 trillion by 2027, according to S&P