The world’s electricity consumption is set to increase by 4% annually through 2027, marking the fastest growth rate in recent years, according to a new report by the International Energy Agency (IEA). This surge is fueled by the rapid expansion of artificial intelligence (AI), increased industrial activity, rising air conditioning usage, and the growing adoption of electric vehicles (EVs).
Soaring Electricity Demand: A Growing Concern
In 2023, global electricity consumption jumped by 4.3%, and the IEA projects an additional 3,500 terawatt-hours will be added to the grid over the next three years. To put it in perspective, that’s equivalent to adding the annual electricity demand of a country the size of Japan to the global power network each year.
A key area of concern is data centers, which are becoming increasingly energy-intensive due to their role in powering AI models and cloud computing. In China alone, the share of electricity consumed by data centers is expected to double by 2027, reaching 6%. While much of the demand growth will come from emerging economies—China alone accounted for over half of last year’s increase—developed nations are also experiencing rising consumption after a period of relative stability.
DeepSeek’s AI Efficiency: A Game Changer?
As AI-driven technologies demand increasing amounts of power, a new player has entered the scene—DeepSeek, a Chinese AI startup that claims to have developed more energy-efficient AI models than its competitors. The company’s reported breakthroughs have sent shockwaves through Silicon Valley and disrupted energy market expectations.
DeepSeek’s AI models, including its R1 system, which reached the top of Apple’s App Store shortly after launch, reportedly deliver high performance at a fraction of the energy cost incurred by major U.S. tech firms. This raises critical questions about whether AI’s expected electricity consumption might be overestimated. If DeepSeek’s technology proves to be significantly more efficient, it could potentially curb the energy surge expected from AI-driven data centers. However, experts warn that greater efficiency could also lead to wider AI adoption, ultimately negating any energy savings.
Implications for Energy Providers and Policymakers
The uncertainty surrounding AI’s electricity demand is already impacting financial markets, with energy company stocks fluctuating as analysts question whether current projections for power demand remain valid. Jefferies analysts have noted that DeepSeek’s energy-efficient AI model challenges expectations of massive electricity consumption growth in the U.S., prompting calls for policymakers to reassess long-term infrastructure investments.
Meanwhile, the U.S. Energy Department has estimated that data centers could account for anywhere between 6.7% and 12% of the country’s electricity use by 2028, up from 4.4% in 2023. The wide variation in these estimates underscores the challenge of predicting AI’s true impact on power grids.
While DeepSeek’s efficiency improvements add a new layer of complexity to the discussion, one thing remains certain: AI and data centers will continue to be major drivers of electricity demand in the coming years. Whether the world can keep pace with this rising consumption—or whether new technological breakthroughs will shift the balance—remains to be seen.
