The AI industry’s energy consumption could match that of the Netherlands by 2027, a study warns.

A new study has cautioned that the artificial intelligence (AI) industry may consume as much energy as a country the size of the Netherlands by 2027. The rapid expansion of AI, which became increasingly prevalent with the emergence of technologies like ChatGPT, has raised concerns about its energy-intensive requirements. AI-powered services demand more power than traditional applications, leading to heightened energy consumption.

The study, conducted by Alex De Vries, a PhD candidate at the VU Amsterdam School of Business and Economics, is based on certain assumptions, such as AI’s continued growth rate, the availability of AI chips, and servers running at full capacity. It estimates an annual energy consumption for AI in the range of 85–134 terrawatt-hours (TWh). At the upper end, this would be equivalent to the annual electricity consumption of a small country like the Netherlands, amounting to about 0.5% of global electricity consumption.

De Vries recommends that AI be employed only where necessary. His study, published in the journal Joule, underscores the need for more transparency in the sector, particularly concerning the energy required for cooling the data centres used for AI. Many tech companies do not provide specific data on their energy consumption or water usage for cooling.

The AI industry, which relies on data centres filled with specialised computers, has become significantly energy-hungry. Demand for AI-powered equipment is surging, necessitating substantial energy consumption to support these systems. AI also has the potential to address environmental challenges; for instance, it has been used to optimise flight paths and reduce contrails known to contribute to global warming. Moreover, AI could accelerate research in fields like nuclear fusion, which has the potential to provide a limitless, green power supply.

As the AI industry expands and innovates, it is essential to consider the environmental implications and develop more energy-efficient AI technologies. This study serves as a warning that, without careful management, the growth of the AI industry could lead to substantial increases in energy consumption.