Guterres launches a transparency initiative at London Climate Action Week, citing rising data center water, electricity, and land demands.

United Nations Secretary-General António Guterres is urging major AI firms to be much more transparent about the environmental toll of their operations, including the electricity, water, and land needed to support data centers.

He says the industry should stop treating those costs as hidden side effects, and instead publish clear public disclosures while also moving toward renewable energy use across data centers by 2030.

The appeal came during London Climate Action Week (20–28 June 2026), where Guterres introduced a new transparency initiative focused on the AI sector’s climate footprint. His message: the explosive growth of AI cannot be separated from the real-world resources it consumes, and those impacts need to be measured openly rather than assumed away. The effort is framed as a way to make the sector accountable at a time when its energy appetite is rising quickly.

When voluntary pledges cannot cut it

That concern is backed by increasingly stark numbers:

  • A UN University study cited in the coverage predicts that AI-related water use could reach a level comparable to the basic annual domestic needs of 1.3bn people by 2030.
  • The same reporting says data centers could draw about 945 terawatt-hours of electricity a year, nearly three times the combined annual use of Pakistan, Bangladesh, and Nigeria.
  • Land use is also part of the problem. The study estimates the infrastructure footprint linked to AI could exceed 14,500 square kilometers, which is roughly twice the size of the Jakarta metropolitan area.

In other words, the environmental cost of AI is not limited to electricity bills or emissions reports; it also reaches into physical space, water systems, and long-term planning for cities and regions. The pressure is especially sharp because AI firms have mostly leaned on voluntary climate pledges instead of firm, mandatory transparency rules. That has left outside observers with only partial visibility into how fast emissions are growing as data centers expand and models become more resource-intensive.

The result is a sector that often markets itself as innovative and efficient while keeping its environmental burden largely out of public view. The issue goes beyond individual firms because governments and international organizations are starting to treat data-center growth as a policy problem, not just a business trend.

Guterres’s remarks suggest that disclosure is becoming the first step towards wider accountability, since companies cannot be managed effectively if their resource-use remains opaque. The central argument is simple: if AI is shaping the future, its climate footprint should be visible now, not after the damage is already locked in.