According to an analysis by Bain & Company, artificial intelligence firms will need to bring in at least $4.2 trillion in new annual revenue by 2031 to fund the extensive infrastructure development required in the industry.
Bain’s report suggests that for the AI sector to finance anticipated spending on data centers, chips, and computing infrastructure, it would require roughly $6 trillion in annual revenue by that year. However, current applications are expected to only generate between $1.2 trillion and $1.8 trillion annually, leaving a significant gap of at least $4.2 trillion that needs to be filled by new products, robotics, autonomous systems, and other innovations.
Importantly, the $6 trillion figure isn’t a prediction of actual revenue AI companies will make. Instead, it’s Bain’s estimate of the revenue necessary to support the levels of investment in computing that are forecasted for the industry.
This considerable gap underscores the enormous risk that technology companies are taking as they invest billions into infrastructure, even before many of the products that are expected to eventually generate returns have been created.
Bain also estimates that current consumer AI applications might bring in around $200 billion to $400 billion annually, while enterprise applications may add about $1 trillion to $1.4 trillion. Still, even at the higher end, existing applications would account for less than a third of the required $6 trillion.
The rest would need to come from sectors like search and advertising, autonomous systems, physical AI, and robotics, in addition to trillions from new products that haven’t yet reached the market, according to Bain’s findings.
This revenue challenge comes as major tech companies ramp up spending and borrowing to compete in the AI field. Combined capital expenditures by six leading hyperscalers are projected to surpass $1.3 trillion by 2027, as per S&P Global Ratings. S&P anticipates that these six firms will experience negative free operating cash flow in 2026 and 2027, largely because expenses will exceed the cash generated internally.
Increasingly, these companies are turning to debt, leases, and special-purpose vehicles to fund their investments. Reports indicate that tech firms have backed as much as $300 billion in AI infrastructure through various guarantees and arrangements, allowing some liabilities to exist outside traditional corporate debt totals.
These arrangements often involve residual value guarantees, where tech companies commit to ensuring that assets like chips and data centers retain certain values over time. While this can reduce the financing burden directly on their balance sheets, it also places them at risk if these assets depreciate far more than expected.
Moreover, the debt load is anticipated to grow steadily. Hyperscaler borrowing is projected to reach an all-time high of $420 billion by 2027, marking an increase of about 60% compared to 2026, according to data from Goldman Sachs.
