AI demand is surging, but Gartner says data centre owners may need $8.5tn in revenue to justify infrastructure spending through 2030, leaving more than $3.5tn potentially exposed to impairment.
Gartner says the scale of the artificial intelligence infrastructure boom now carries a material risk of asset write-downs, with more than $3.5 trillion of spending potentially exposed to impairment by the end of 2030. The research firm said data centre owners would need to generate $8.5 trillion in cumulative revenue over the period to cover capital outlays, financing costs and a 12% return on invested capital. Yet it projects only $3.6 trillion in attributable AI revenue from enterprises and consumers, leaving a large gap between spending and the cash flow needed to justify it. According to Gartner’s latest forecast, the imbalance raises questions about how much of the current build-out can ultimately earn an economic return.
That risk is most acute for specialist “neocloud” operators, which depend heavily on demand for computing capacity rather than the broader software and services businesses enjoyed by large hyperscalers such as Microsoft, Alphabet and Amazon. Gartner says more than 80% of neocloud providers are likely to stop operating independently by 2030 as cost pressure and limited product differentiation squeeze margins. Will Sommer, a senior director analyst at Gartner, said many of these companies are trying to move further up the stack into higher-margin cloud services, but that few are likely to succeed. “All of them want to be the full-stack. They can’t all be,” he said. “For every dollar that’s spent on compute, only a fraction of that is going to the owners of compute.”
The warning comes despite Gartner’s view that AI spending will remain exceptionally strong in the near term. In September, the firm forecast worldwide AI expenditure at $2.7 trillion in 2026, up 49.5% from the previous year, with AI infrastructure cited as the largest and most capital-intensive part of the market. It has also projected that spending on AI-optimised infrastructure as a service will almost double in 2026 to $42 billion, rising further to $66 billion in 2027. In earlier forecasts, Gartner said worldwide AI spending would total nearly $1.5 trillion in 2025 and later revised its 2026 estimate upwards to $2.59 trillion, underscoring how quickly demand for servers, network fabric and semiconductors has accelerated.
The contradiction at the heart of the market is that demand for AI capacity remains real, but the economics of building that capacity are becoming harder to defend. Gartner’s separate guidance on AI spending warns that costs are rising quickly while many organisations still struggle to connect investment to measurable outcomes. That tension is now being felt most sharply in the infrastructure layer, where hardware depreciates fast, competition is intense and the revenue model may not be broad enough to support the scale of construction under way. If Gartner’s revenue assumptions prove accurate, the current AI data centre expansion could leave a sizeable portion of the sector facing impairments before the decade ends.
Disclaimer: This content is intended for informational purposes only. Readers are advised to exercise their own judgement, conduct due diligence, or consult a qualified expert before acting on any information provided.
