Global AI infrastructure investments face mounting financial risks, with over $3.5 trillion potentially exposed to asset impairment by 2030, according to Gartner, as capital expenditure outpaces revenue generation.
AI providers are projected to spend nearly $1 trillion in 2026, followed by another $6 trillion between 2027 and 2030, raising concerns about investment sustainability.
Gartner estimates AI data centre operators must generate $8.5 trillion in cumulative revenue to recover investments, financing costs and achieve a 12% return on invested capital.
However, attributable enterprise and consumer AI revenue is projected to reach only $3.6 trillion, creating a substantial profitability gap.
The biggest casualties could be neocloud providers, whose business models depend heavily on AI infrastructure utilisation and rapidly depreciating semiconductor assets.
Despite AI-optimised infrastructure spending potentially reaching $174 billion by 2030, Gartner predicts over 80% of neocloud providers could cease independent operations through consolidation or financial pressures.
The larger concern is whether AI monetisation can justify unprecedented infrastructure investments.
Without sustainable revenue, operational efficiency and differentiated services, today's AI infrastructure boom risks becoming tomorrow's massive asset write-down crisis.
