Two-thirds of the revenue needed to justify the AI buildout are still unaccounted for, says major consulting firm

A major consulting firm estimates that roughly two-thirds of the revenue needed to justify the artificial-intelligence buildout is not yet accounted for. According to the firm, at least about $4.2 trillion of the estimated yearly costs required to fund the buildout is currently not being covered by identified revenue sources.

The figure matters because AI infrastructure—data centers, chips, and power capacity—is capital-intensive, and those commitments are typically made years before the revenue they depend on materializes. If the gap persists, the risk shifts to how projects are financed, how quickly capacity is used, and whether pricing for AI services can rise enough to close the difference.

For markets, the key takeaway is that the debate is moving from technological capability to funding durability. What to watch: updates to these estimates, disclosed capex plans and depreciation schedules, data-center utilization rates, and any signs that AI revenue is scaling faster or slower than the cost base assumes. Those signals will shape how investors assess both AI-linked equities and the credit behind them.