The artificial intelligence boom is generating enormous capital commitments from the world’s largest technology companies, but the gap between spending and revenue is drawing serious scrutiny from investors.
Data from Capital on Tap shows the share of UK small and medium-sized enterprises paying for AI services quadrupled from 3.2% in Q2 2024 to 12.8% in Q2 2026, a significant jump in adoption by any measure.
Despite that rapid growth in business uptake, typical spending among those SMEs remains relatively small, meaning the customer base is expanding far faster than the actual revenue flowing to AI providers.
On the infrastructure side, the numbers are of an entirely different magnitude, with America’s biggest technology companies projected to spend roughly $900 billion on AI infrastructure in 2026 alone, according to The Economist.
That figure is expected to climb even further in 2027, when The Economist estimates the same group of major technology firms will collectively deploy approximately $1.4 trillion in AI-related capital expenditure.
The contrast between those infrastructure commitments and the modest commercial revenues being generated by AI services is rapidly becoming one of the defining questions of the current technology investment cycle.
Investors are increasingly focused on when, and whether, the returns on that extraordinary level of capital deployment will materialize at a scale sufficient to justify the spending.
The SME adoption figures do point to genuine and accelerating demand for AI tools in the broader business community, offering some evidence that commercial uptake is moving in the right direction.
However, the pace of that commercial adoption would need to accelerate dramatically and consistently over the coming years to close the gap with infrastructure expenditure of this size.
The trillion-dollar question hanging over the AI industry is not whether the technology works, but whether the business models surrounding it can generate returns that match the ambitions embedded in today’s capital commitments.
The widening divergence between what is being spent to build AI infrastructure and what is currently being earned from it represents a central risk that analysts and institutional investors will be watching closely throughout 2026 and into 2027.
