AI data centers to consume 945 TWh of electricity year by 2030, UN study warns

08 / 06 / 2026

By Marwa Nassar - -

Artificial intelligence data centers could consume 945 terawatt-hours (TWh) of electricity annually by 2030, according to a new study from the United Nations University (UNU), highlighting the technology’s rapidly growing environmental footprint.

The projected electricity demand is nearly three times the combined annual power consumption of Pakistan, Bangladesh, and Nigeria, countries with a combined population of more than 650 million people.

Environmental costs extend beyond carbon emissions:

While discussions around AI’s environmental impact often focus on greenhouse gas emissions, the UNU study found that the technology’s footprint also includes significant water and land demands associated with data center cooling, electricity generation, and supply chains.

Researchers estimate that AI-related water consumption could reach the equivalent of the basic annual domestic needs of 1.3 billion people by the end of the decade. The sector’s land footprint could also exceed 14,500 square kilometers, roughly twice the size of the Jakarta metropolitan area.

The report notes that measures aimed at reducing carbon emissions may sometimes increase pressures on water resources and land use, particularly in regions already facing resource constraints.

Daily AI use drives most energy demand:

The study found that day-to-day AI usage accounts for approximately 80% to 90% of the technology’s total energy demand, exceeding the energy required to train advanced AI models.

One widely used AI service is estimated to process around 2.5 billion prompts per day, consuming hundreds of gigawatt-hours of electricity annually.

Energy requirements vary significantly depending on the application. Generating a single AI image can require more than 1,000 times the energy needed for simple text classification tasks, while AI-generated video demands even greater resources.

According to the report, efficiency gains alone are unlikely to curb overall resource consumption, as lower costs and improved performance often drive increased usage.

Growing pressure on resources:

The environmental impacts of AI infrastructure are not evenly distributed, the study said. In some countries, data centers already account for a significant share of national electricity consumption, while in others they are placing increasing pressure on water supplies, including in drought-prone regions.

The report also warned that AI infrastructure could generate up to 2.5 million tonnes of electronic waste annually by 2030, with much of the disposal burden expected to fall on lower-income countries with limited waste-management capacity.

Demand for critical minerals used in AI hardware is also raising concerns about environmental degradation and social inequities in extraction regions.

Calls for responsible AI development:

Despite the findings, UNU researchers stressed that the report is not an argument against AI. Instead, it calls for the development of a responsible AI ecosystem based on transparency, efficiency, equity, lifecycle responsibility, global cooperation, and sustainable use.

The study urges governments to incorporate AI infrastructure into energy, water, and land-use planning, while encouraging companies to design systems that minimize resource consumption.

Ultimately, the report argues that the environmental impact of AI will depend not only on technological advances but also on the policy and governance decisions made today.