The artificial intelligence sector needs to generate $6 trillion in yearly revenue by 2031 to justify the massive infrastructure investments required to meet anticipated demand, according to Bain & Company's 2026 Global Technology Report. That ambitious target will require the industry to develop creative new applications far beyond simply boosting worker productivity. The consultancy warns that current and projected uses of AI won't come close to hitting that revenue threshold.
Bain estimates that combined capital spending by Microsoft, Google, Amazon, Meta, and Oracle could reach $780 billion throughout 2026, nearly five times what these companies spent just three years ago. By 2031, annual expenditure on AI infrastructure could climb to $1.5 trillion. The firm calculates the $6 trillion revenue figure based on the assumption that capital spending will represent roughly 25 percent of industry revenue, a ratio it describes as ambitious but reasonable given trends among cloud providers. Established AI applications—including consumer AI through subscriptions and advertising, plus enterprise AI for software development, sales, marketing, customer service, and IT operations—are expected to produce between $1.2 trillion and $1.8 trillion in revenue, leaving a $4.2 trillion gap the industry must fill from other sources.
The report identifies four possible categories to bridge this shortfall. AI models replacing search engines and incorporating advertisements could unlock $100 billion to $200 billion in additional revenue. Autonomous vehicles, drones, and other industrial automation could create products and services worth approximately $400 billion. Physical AI—encompassing simulations, digital twins, and robotics—might generate as much as $900 billion. Those three categories total $1.5 trillion, still leaving a $2.7 trillion gap that Bain attributes to "new products and uses that don't exist today," including AI-driven drug discovery, mental health support, materials science breakthroughs in battery technology and semiconductors, and accelerated scientific research in fields from neuroscience to fusion energy.
"The debate today is fixated on employee productivity. The economics of AI infrastructure demand trillions in new revenue beyond productivity gains," said David Crawford, chairman of Bain's global Technology practice. The firm notes this projection has ballooned dramatically—just a year ago, a previous Bain report forecast the industry would need to reach $2 trillion in revenue by 2030, meaning the new target has tripled in 12 months. The report points to several headwinds: multiple analyses indicate AI deployments aren't delivering expected returns, investment bank Jefferies found only half the US datacenter capacity scheduled for 2026 is actually under construction, and chip manufacturing constraints may limit how many planned AI server farms can come online over the next several years. The industry faces a fundamental challenge—it must invent entirely new revenue streams worth trillions, not just scale existing applications, to sustain the infrastructure arms race already underway. Questions about whether such transformative uses will materialize on this timeline remain unanswered, and the widening gap between infrastructure spending and proven returns suggests the sector may be building on speculation rather than demonstrated value. The stakes couldn't be clearer: without breakthrough applications no one has yet imagined, the current AI boom risks becoming the most expensive technology bet in history.