Policy

Google's AI Expansion Drives Record Emissions, Water Use in 2025

The tech giant's electricity demand jumped 37% and greenhouse gas emissions rose 18% despite aggressive clean energy investments.

Omega Editorial· June 30, 2026· 3 min read

Google's artificial intelligence infrastructure buildout drove the company's electricity consumption, water use, and greenhouse gas emissions to unprecedented levels in 2025, according to the company's environmental report released Tuesday.

The figures underscore a growing tension in the technology sector: even companies with substantial clean energy commitments are struggling to contain their environmental footprint as they race to expand AI capabilities. Google has historically been among the most aggressive tech companies in pursuing renewable energy, yet the scale of its AI expansion is outpacing those efforts.

The numbers tell the story

Google's electricity demand surged 37% in 2025, accelerating from a 27% increase the previous year. The company now consumes roughly 3.5 times the electricity it used in 2019.

Greenhouse gas emissions climbed 18%, marking the largest single-year increase Google has reported. The rise stems primarily from manufacturing AI hardware components, including specialized chips and servers required for machine learning workloads.

Water consumption reached 10.9 billion gallons, a 34% increase from 2024 and more than double the 2021 level. Data centers accounted for the majority of this growth, as cooling systems for high-density AI computing require substantial water resources.

Why it matters

The report signals a fundamental shift in how technology companies measure environmental progress. Rather than reducing total emissions, the new benchmark has become limiting how fast they grow. This recalibration affects not just corporate sustainability goals but also the broader conversation about AI's role in climate strategy. As other major tech companies prepare to release similar reports in coming weeks, investors and policymakers will gain a clearer picture of whether the industry can reconcile AI ambitions with climate commitments.

Efficiency gains can't keep pace

Google's data centers are becoming more efficient on a per-unit basis, but the sheer volume of new AI infrastructure is overwhelming those improvements. The company acknowledged this dynamic directly in its report: "This rapid expansion in energy demand is a reality we must manage actively, and we're committed to ensuring that the growth of AI doesn't become a rationale for lowering our environmental standards."

The company did secure 12 gigawatts of clean energy agreements in 2025, a record amount. It also maintained a roughly flat share of carbon-free electricity despite soaring overall demand. Electricity-related emissions fell 2% compared to 2024, though this represents a slower decline than the 12% drop achieved the year before.

The broader context

Technology companies have published annual environmental reports for years—Google since 2016. These documents previously served as platforms for highlighting clean energy achievements and climate leadership. The AI boom has transformed them into accountability measures that reveal the gap between sustainability pledges and operational reality.

Google expanded its discussion of AI's potential environmental benefits this year, growing from five to nine initiatives where the technology might reduce emissions in other sectors. The company continues to argue that AI applications can deliver net climate benefits across the broader economy.

Microsoft and Amazon are expected to release their annual environmental reports in the coming weeks, according to Axios, which first reported the details of Google's filing. Those disclosures will provide additional data points on whether the pattern Google is experiencing extends across the industry's leading AI developers.

#google#artificial intelligence#data centers#emissions#sustainability#clean energy

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy

Policy· 3 min read

Cleveland Fed's Hammack: AI Infrastructure Demand Driving Inflation

Central bank official warns that unchecked spending on data centers and AI systems may require interest rate increases to control price pressures.

Via AI Watch · Jun 30, 2026
Policy· 2 min read

Trump Administration Controls Access to Advanced AI Models

A new regulatory regime gives the U.S. government approval authority over cutting-edge AI releases, raising questions about transparency and long-term viability.

Via AI Watch · Jun 30, 2026
Policy· 3 min read

AWS Launches $2B in Cloud Programs for Defense and Intelligence

New classified infrastructure for contractors, migration credits for IC agencies, and embedded AI engineering teams mark major public sector push.

Via AI Watch · Jun 30, 2026