Artificial intelligence (AI) data centers face operational constraints due to a shortage of electricity, warns market research group Gartner. According to their predictions, by 2027, approximately 40% of today’s AI data centers could be affected by power availability issues. The rapid growth of tech giants’ AI-focused data centers has outpaced the expansion of energy capacity by utility providers, says Gartner analyst Bob Johnson. As a result, the power consumed by data centers could surge by up to 160% over the next two years.
The immediate consequence of these power shortages is the potential jeopardy of the AI tools market, as constraints drive up electricity prices. However, the long-term implications are even more concerning. Gartner highlights that data center expansion may lead to increased carbon emissions, as grid operators are compelled to keep fossil fuel plants operational for longer periods than scheduled.
In response to this looming crisis, industry leaders such as Amazon, Google, and Microsoft are exploring alternative energy sources, including nuclear power. However, the implementation of nuclear plants is a time-consuming process, requiring significant lead time.