From ordinary users to politicians, these AI-generated “80s” avatars have swept across social media.

While the trend may appear harmless, creating an instant AI-generated image comes with a hidden environmental cost. Behind every such image are data centres, specialised chips, electricity, cooling systems and water-intensive infrastructure working to process the request.

Although it remains unclear how many AI-generated 1980s-style images have been produced since the trend began, the potential environmental impact is difficult to ignore.

According to a new study by the United Nations University (UNU), AI-related water consumption could equal the basic annual domestic water needs of 1.3 billion people by the end of the decade. Its land footprint could also exceed 14,500 square kilometres — roughly twice the size of the Jakarta metropolitan area.

The report highlights a critical gap in the way AI’s environmental impact is measured. Greenhouse gas emissions, particularly those associated with training large AI models, tend to receive the most attention, while other environmental costs are often overlooked.

Solutions considered “green” in one respect can also intensify pressures elsewhere, particularly in regions already facing resource scarcity. For instance, switching to certain renewable energy sources may reduce carbon emissions but significantly increase demand for water and land.

Daily AI Use Is the Main Culprit

Public debate has largely focused on the energy required to train advanced AI models. However, the study finds that day-to-day AI use accounts for roughly 80% to 90% of total energy demand.

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

Energy requirements also vary significantly depending on the task. Generating a single AI image can require more than 1,000 times the energy needed for simple text classification, while video generation can demand even greater resources.

The report warns that efficiency improvements alone may not be enough to offset rising demand. It points to the so-called “rebound effect”, whereby lower costs and improved performance encourage greater usage, ultimately increasing overall resource consumption.

Local Burdens, Global Benefits

The environmental impacts of AI infrastructure are not evenly distributed. While the benefits of AI are global, its environmental costs are often concentrated in particular regions.

In some countries, data centres already account for a significant share of national electricity consumption, putting additional pressure on energy systems. In others, expanding data-centre infrastructure is placing heavy demands on water supplies, sometimes in areas already experiencing drought conditions.

The report also warns of a growing electronic waste problem, with AI infrastructure projected to generate up to 2.5 million tonnes of e-waste annually by 2030. Much of this burden could fall on lower-income countries with limited capacity for safe disposal.

The extraction of critical minerals required to manufacture AI hardware also raises concerns about environmental degradation and social inequalities in resource-rich regions.

A Widening Digital and Environmental Divide

The expansion of AI infrastructure is creating new disparities in access and influence. According to the report, more than 90% of AI-specialised computing capacity is concentrated in just two countries — the United States and China.

At the same time, more than 150 countries lack significant domestic AI infrastructure.

This imbalance not only limits economic opportunities but also raises questions about environmental justice, as some countries may bear the environmental costs of AI-driven growth without sharing proportionately in its benefits.

Towards Responsible AI

Despite its stark findings, the UNU study is not an argument against AI itself. Instead, researchers are calling for urgent action to ensure that the technology develops within planetary limits.

The study outlines a framework for a “responsible AI ecosystem” based on principles including transparency, efficiency by design, equity, lifecycle responsibility, global cooperation and sustainable use.

Governments are urged to incorporate AI infrastructure into energy, water and land-use planning, while companies are encouraged to design systems that minimise resource consumption.

Users, too, can play a role by choosing lower-impact AI applications where possible and being mindful of unnecessary usage.

Ultimately, the report argues that the future of AI will depend not only on technological innovation but also on the governance decisions made today.