Gartner Warns Over $3.5 Trillion in AI Infrastructure Spending Could Be Written Down by 2030
Research firm Gartner warned in October 2026 that more than $3.5 trillion of capital invested in AI infrastructure is at risk of impairment by 2030. The firm projects that AI revenue from data centers will reach only $3.6 trillion by 2030, far short of the $8.5 trillion needed to justify current spending levels.
Research firm Gartner warned in October 2026 that more than $3.5 trillion of capital invested in artificial intelligence infrastructure is at risk of impairment by the end of 2030.
The firm projects that AI providers will spend nearly $1 trillion on infrastructure in 2026 alone, with an additional $6 trillion projected through 2030. To achieve a 12 percent return on that total investment, the industry would need to generate approximately $8.5 trillion in cumulative revenue by the end of 2030.
However, Gartner forecasts that revenue from enterprise and consumer AI services directly attributable to these data centers will reach no more than $3.6 trillion during that period. The gap between projected spending and expected revenue creates a high risk of "impairment," where companies are forced to write down the recorded value of their infrastructure assets.
Gartner further warned that the risk could escalate, with potential asset write-downs exceeding $4 trillion by 2032.
The report specifically identified "neocloud" providers as particularly vulnerable. These are specialized operators that provide AI-optimized infrastructure as a service. Unlike large hyperscalers with diversified revenue streams, neocloud providers rely heavily on returns from their specific infrastructure investments. Gartner predicts that over 80 percent of independent neocloud providers may exit the market through mergers, acquisitions, or closures by 2030.
Will Sommer, a senior director analyst at Gartner, attributed part of the problem to "asset-light" business models favored by many AI-native companies. These firms often retain the value of AI applications while the heavy capital burden remains with infrastructure owners.
The warning comes as companies including TSMC, Microsoft, Google, and Amazon continue to announce massive data center and chip manufacturing investments. Gartner said the industry faces a challenge of capital efficiency rather than a lack of demand.
