The AI industry must generate $6 trillion in annual revenue by 2031 to keep funding the infrastructure required to meet anticipated demand, according to Bain & Company's 2026 Global Technology Report. That figure assumes capital expenditure will amount to about 25 percent of industry revenue, a ratio Bain says is ambitious but reasonable given cloud provider trends. The forecast is a sharp escalation from Bain's own estimate a year earlier, when it projected $2 trillion in revenue by 2030.
Bain expects existing AI applications—consumer subscriptions and ads, plus enterprise software, sales, and IT operations—to bring in only $1.2 trillion to $1.8 trillion. To close the gap, it points to AI replacing search engines with ads, autonomous vehicles and industrial automation, and physical AI such as simulations and robotics. Those categories add up to roughly $1.5 trillion, leaving $2.7 trillion that Bain says must come from "new products and uses that don't exist today," including drug discovery, mental health support, and materials science.
The report also highlights the scale of the infrastructure buildout: Bain estimates hyperscaler capital expenditure by Microsoft, Google, Amazon, Meta, and Oracle could reach $780 billion in 2026, nearly five times the level three years earlier, and $1.5 trillion by 2031—close to Omdia's separate $1.6 trillion forecast. Yet The Register notes reasons for skepticism, citing reports that AI rollouts are not paying off and Jefferies data showing only half of US datacenter capacity scheduled for 2026 is under construction, with chip manufacturing constraints likely to limit how many planned server farms come online.