Bain Warns AI Industry Faces $6 Trillion Revenue Test

Bain & Co. said on 29 September that for the capital pouring into data centres to be worthwhile, the artificial intelligence industry will need to bring in $6 trillion a year by 2031, cautioning that $4.2 trillion of that figure has no identified source so far.²³¹

The revenue gap

According to Bain's 7th Global Technology Report, annual spending on AI infrastructure could climb to $1.5 trillion by 2031 as data centres, chips, networks and power systems expand. To make that spending economically sustainable, the consulting firm estimates the AI industry needs to earn roughly $6 trillion a year by the same date.²⁴¹⁶

Bain estimates that current AI offerings for consumers and businesses will bring in somewhere from $1.2 trillion to $1.8 trillion of that overall figure. That leaves about $4.2 trillion in new revenue still needed from segments of the AI market that barely exist today.²³⁵

FourWeekMBA, commenting on the report, said the $6 trillion should not be read as a spending or cost figure. "The $6 trillion is not money being spent on AI — it is the revenue the industry would have to earn for the spending to make sense," the outlet wrote, adding that roughly 70 per cent of that required market has not yet been earned.⁶

Where the new revenue could come from

Bain identified several nascent categories that could help close the gap. Businesses using AI across functions like coding, sales, marketing, customer support and IT could generate $1 trillion to $1.4 trillion for providers by 2031, and AI products aimed at consumers could add another $200 billion to $400 billion.¹⁴

Search and advertising built around chatbots and AI-based search could unlock $100 billion to $200 billion, Bain said, while what it calls "autonomous everything" — including autonomous vehicles, robotaxis and logistics automation — represents a $400 billion opportunity. Physical AI and robotics across automotive, electronics, semiconductor and aerospace sectors could add $900 billion, assuming a 10 per cent reduction in research, development and manufacturing costs.⁴

Bain also pointed to entirely new products that do not yet exist. "New products and uses that don't exist today will enable new markets and opportunities from abundant intelligence – these may include drug discovery, mental health and energy generation," the firm said.¹

David Crawford, who authored the report and leads Bain's technology, media and telecommunications practice, argued that the amount of investment needed cannot be justified through corporate efficiency gains alone. "Current conversations are overly focused on worker output. Beyond productivity improvements, the financial reality of AI infrastructure calls for trillions in fresh income. What's required is a surge of innovation far larger than what mobile technology and cloud computing produced," Crawford said.¹⁵

Scaling infrastructure

According to the report, both the scale and expense of data centres are roughly doubling every 12 to 16 months, a trend partly fueled by rising chip costs. Today's cutting-edge AI data centres are nearing 1 gigawatt of power capacity, and Bain projects that many will near 2 gigawatts by 2027, with facilities of 9 gigawatts appearing before 2030 ends.²⁴⁶¹

Bain further noted that spending by major cloud providers may hit $780 billion in 2026 - almost five times what it was three years prior - and that top AI labs are pouring more than $9.75 billion into forward-deployed engineering models. Sustaining this level of investment, the firm estimates, would require adding roughly 1 percentage point to the annual global GDP growth rate.⁴⁵²

Bain framed the challenge as one of economic justification rather than physical capacity. "The unprecedented speed and scale of the AI buildout, with billions flowing into chips, data centres, networks and power systems, have focused attention on the challenge of building capacity," the firm said, adding: "But the more important question may be whether enough economic value can be created to justify it."¹