Bain says AI will need to earn $6 trillion a year to justify the data center buildout

The global artificial intelligence industry must generate $6 trillion in annual revenue by 2031 in order to justify the massive investments being made in data centers worldwide, according to a new report from Bain. Existing AI services for consumers and businesses could contribute up to $1.8 trillion, leaving a $4.2 trillion gap that needs to be filled with new revenue sources.

These sources are expected to come from emerging areas such as autonomous machines, robotics, drug discovery, mental health care, and energy generation. To sustainably fund this growth, the infrastructure supporting AI development must expand at a rate comparable to the global GDP growth rate, said David Crawford, the lead author of the report and head of Bain's Global Technology, Media, and Telecommunications practice.

The report underscores the challenges ahead in making the pace of AI development sustainable. Major tech companies like Microsoft, Alphabet's Google, Amazon, Meta, and Oracle are pouring trillions of dollars into building data centers to meet AI's growing demand for computational resources. Data center sizes and costs are doubling approximately every 12 to 16 months, driven by rising chip prices from Nvidia, SK Hynix, and other suppliers, as well as the escalating costs of networking equipment and other components.

The report coincides with growing scrutiny over the elusive returns for AI service providers. Critics warn of a growing web of interdependencies between technology manufacturers and AI developers, which fuels high expectations and necessitates even larger financial commitments. While current discussions focus on boosting employee productivity, Bain highlights that the economics of AI infrastructure will ultimately require trillions in new revenue beyond productivity gains.

The consultancy predicts data-center spending will range between $5 trillion and $6.5 trillion by 2030, adding at least 150 gigawatts of capacity that will put additional strain on countries' energy resources. Annual spending on AI infrastructure — encompassing data centers, computing capacity, and upgrades in accelerators and memory chips — could reach up to $1.5 trillion by 2031, the report says.

Data-center developers already face shortages of transformers, water, and power supplies, as well as local opposition that has blocked or delayed $68 billion worth of projects in the U.S. in the June quarter.