The AI math isn't mathing as infrastructure spending forecast to outstrip expected revenue

AI services are unlikely to generate sufficient revenue to cover the anticipated capital expenditure for IT infrastructure infrastructure between now and 2030, according to Gartner. The research firm found that neocloud operators, which specialize in AI compute, may either exit the market or be acquired due to the financial strain.

Gartner's report, "Prepare for AI Market Consolidation," estimates that AI providers will spend nearly $1 trillion on AI infrastructure in 2026 alone, with an additional $6 trillion invested between 2027 and 2030. However, the firm estimates that enterprise and consumer AI revenue will only amount to $3.6 trillion during this period, failing to meet the required $8.5 trillion to achieve a 12 percent return on invested capital.

This shortfall could lead to write-downs of assets, resulting in a shake-out of AI infrastructure service providers. Smaller neocloud operators are expected to suffer the most, with Gartner predicting that 80 percent of them will not survive as independent companies beyond 2030. Larger hyperscalers, with businesses beyond AI services, are expected to fare better in the market shake-out.

Financial market concerns over the excessive AI capital spending have led some lenders to become more cautious in their approach. Several reports have warned about the potential profitability threats faced by AI developers, including those from open-source and Chinese models.