New Delhi, June 10: The rapid expansion of artificial intelligence (AI) is expected to drive a significant increase in electricity consumption by data centres worldwide, with global demand projected to rise by more than 26 per cent in 2026, according to a new report.
Research and advisory firm Gartner said global electricity consumption by data centres is expected to increase from 447 terawatt hours (TWh) in 2025 to 565 TWh in 2026, representing a year-on-year growth of 26.4 per cent.
The surge is being attributed primarily to the rapid adoption of compute-intensive AI applications, particularly generative AI technologies, which require significantly greater processing power than conventional computing workloads.
According to the report, data centre power demand is expected to reach 132 gigawatts (GW) in 2026, up from 104 GW in 2025. The figure is projected to rise further to 290 GW by 2030 as demand for AI infrastructure continues to expand globally.
Linglan Wang, Director Analyst at Gartner, said power availability is increasingly becoming a key factor influencing AI development and deployment.
“AI capacity is now constrained by power availability, making data centre power security the new battleground for scaling and protecting margins in the global AI race,” Wang said.
The report highlights the growing role of AI-optimised servers, which are expected to account for nearly 31 per cent of total data centre electricity consumption in 2026. By 2027, power consumption from AI-focused servers is projected to exceed that of conventional servers for the first time.
Industry experts believe the trend reflects the increasing reliance on advanced AI models that require enormous computational resources for training and deployment.
The report also warns that electricity demand from data centres could exceed 1,200 TWh annually by 2030, placing substantial pressure on power grids worldwide. Such demand levels may outpace the ability of existing electricity infrastructure to support future data centre construction and expansion projects.
To address these challenges, Gartner urged infrastructure and operations leaders to focus on improving efficiency and securing long-term power access.
The report recommends investments in high-efficiency cooling technologies, advanced energy management systems and edge computing infrastructure to reduce pressure on centralised data centres and improve energy utilisation.
Cooling requirements are also expected to increase sharply as computing power expands. Gartner estimates that cooling and supporting infrastructure demand will rise by 22.6 per cent in 2026 and by a further 24.6 per cent in 2027.
The findings come amid growing concerns about the environmental impact of AI-driven computing growth. A recent industry assessment found that global data centres consumed as much electricity as Saudi Arabia during the previous year.
The same study suggested that if electricity consumption by data centres doubles by 2030, offsetting the associated carbon emissions would require approximately 6.7 billion trees grown over a ten-year period.
As governments and technology companies continue investing heavily in AI capabilities, the challenge of balancing technological advancement with energy security and environmental sustainability is expected to become increasingly important.
With AI applications expanding across sectors ranging from healthcare and education to finance and manufacturing, experts say the race to build sufficient computing infrastructure is now closely tied to the availability of reliable and sustainable power supplies.
