AI Servers Could Trigger an Energy Crisis by 2026

IT Sector Faces Energy Crisis Forecast

According to Главком: According to a June 2026 report from analytics firm Gartner, the IT industry is heading toward an energy crisis driven by surging electricity consumption in data centers. Global data center power usage is expected to jump 26.4% in 2026, reaching 565 TWh compared to 447 TWh in 2025. The primary culprit is the skyrocketing demand from artificial intelligence servers, which are projected to account for 31% of all data center energy consumption by 2026.

AI server energy use is growing at an alarming rate. In 2025, these systems consumed 95 TWh-an 83.6% increase from the prior year. By 2026, that figure is forecast to climb another 84.2% to 175 TWh, with projections for 2027 hitting 258 TWh. This makes AI servers one of the biggest drivers of rising power demands in data centers worldwide.

Infrastructure and Business Models Under Pressure

At the same time, cooling systems and other data center infrastructure are consuming more resources due to the intense heat generated by AI chips. Energy use in this segment rose from 159 TWh in 2025 to 195 TWh in 2026, a year-over-year increase of 22.6%. By 2027, it could reach 243 TWh. In contrast, traditional servers-the most stable and energy-efficient segment-show minimal change, with consumption holding at 193 TWh in 2025, 195 TWh in 2026, and a predicted 200 TWh in 2027.

Total projected data center electricity consumption for 2027 stands at 702 TWh, and by 2030 it could exceed 1,200 TWh. Linlan Wang, research director at Gartner, commented:

“Infrastructure leaders must prioritize energy efficiency and access to reliable power grids. Investment in advanced cooling systems and edge computing will be essential to mitigate the crisis.” – Linlan Wang

This escalating power consumption in the IT sector poses a serious challenge for technological advancement in the coming years.

The situation underscores the urgent need for new technologies to reduce data center energy use, as well as the necessity of adapting infrastructure to meet growing demands. Beyond that, an energy crisis could reshape business models for tech companies, pushing them toward innovative solutions that boost efficiency and cut electricity costs.

The increasing energy demands of AI servers not only pose a challenge for data centers but also prompt users to seek more cost-effective solutions. As companies grapple with rising operational costs, many are exploring affordable alternatives that can help mitigate these expenses while maintaining performance. This shift could significantly impact the future landscape of AI technology and its sustainability.