Gartner expects AI infrastructure to account for 59% of global AI spending next year, driven by expanding computing workloads and rising server prices.
- The firm expects AI infrastructure spending to jump 53% to $2.4 trillion in 2027.
- It cited growing inference and training workloads, hyperscaler and neocloud build-outs, and “memflation.”
- Spending on AI agents and assistants is set to more than double to $67 billion from $30 billion, the fastest growth of any category.
Gartner on Thursday forecast global spending on artificial intelligence (AI)-optimized servers to reach $957.2 billion in 2027, an increase of around 81% from $527.6 billion in 2026.
The research firm expects total worldwide AI spending to rise 47.5% to $4.05 trillion next year, with infrastructure accounting for approximately 59%, or $2.4 trillion.
The forecast comes amid renewed turbulence in AI-linked stocks, as uncertainty over Nvidia (NVDA) CEO Jensen Huang’s future at the firm rattled investors following reports of two extramarital affairs, including one involving a woman suspected of ties to Chinese intelligence.
At the time of this writing, Nasdaq-100 futures were down 0.7%, with a selloff across semiconductor and AI infrastructure stocks.
Why AI Server Spending Is Rising
Gartner attributed the projected spending growth to faster-than-anticipated expansion of AI inference workloads, which involve running trained AI models, alongside increasingly compute-intensive training requirements.
John-David Lovelock, Distinguished VP Analyst and Chief of Research at Gartner, said both trends continue to increase spending on AI infrastructure globally.
The firm also cited continued infrastructure expansion by hyperscalers and neocloud providers, alongside “memflation,” its term for memory-driven increases in AI server prices.
Gartner said in April that meaningful relief from elevated memory prices was unlikely before late 2027. Lovelock separately told CIO Dive in October that memflation could persist into 2030.
The spending outlook is relevant to companies supplying AI servers and computing hardware, including Dell Technologies (DELL), Super Micro Computer (SMCI) and Nvidia (NVDA).
AI Agents, Cybersecurity And Software Spending To Rise
Beyond infrastructure, Gartner expects spending on AI agents and assistants to more than double from $30 billion in 2026 to $67 billion in 2027, making it the fastest-growing category in its forecast.
AI cybersecurity spending is projected to increase from $51 billion in 2026 to $86 billion in 2027, representing approximately 69% growth, according to Gartner. The forecast puts cybersecurity companies such as CrowdStrike (CRWD), Palo Alto Networks (PANW), Zscaler (ZS) and Okta (OKTA) in focus.
Meanwhile, spending on generative AI models is expected to rise 82% to $51 billion. Gartner forecasts spending on domain-specific models to grow 99%, although foundation models are expected to retain 81% of the generative AI model market.
The firm also expects AI software spending to increase 42% to $649 billion and AI services spending to rise 29.5% to $746 billion.
AI Infrastructure Investment Risks Remain
The latest forecast follows Gartner’s October warning that more than $3.5 trillion in AI infrastructure spending could face impairment by 2030 if investment continues at its current pace.
The firm said enterprise and consumer AI revenue would be insufficient to generate profitable returns on all planned capital expenditures. It also predicted that more than 80% of neocloud providers would cease independent operations by 2030, citing cost pressures and a lack of differentiated business models.
At the time of this writing, the Invesco QQQ Trust Series 1 (QQQ), which tracks the Nasdaq-100 index, was down over 0.5% in Thursday’s premarket. Retail Investors’ sentiment around the ETF remained ‘bullish’ on Stocktwits.
AI-linked ETFs were under pressure, with the iShares Semiconductor ETF (SOXX) falling over 1%. The Defiance AI & Power Infrastructure ETF (AIPO), which tracks companies tied to AI and power infrastructure, declined marginally. Meanwhile, the First Trust Nasdaq Cybersecurity ETF (CIBR) edged up 0.2%.
See Also: Anthropic IPO Could Be Bitcoin’s Next Catalyst: Arthur Hayes Turns ‘Mini-Bulled Up’ Ahead Of Nasdaq Listing
For updates and corrections, email newsroom[at]stocktwits[dot]com.<