Google’s AI expansion drives surge in emissions and power use

Google wants to organize the world’s information. Turns out, that goal now comes with a carbon footprint roughly the size of a small nation’s.

The company’s 2025 Environmental Report reveals that total greenhouse gas emissions reached approximately 11.5 million tons of CO2-equivalent in fiscal year 2024. That’s a 51% increase from Google’s 2019 baseline, driven overwhelmingly by the infrastructure required to power AI models like Gemini.

Data center electricity consumption alone climbed 27% year-over-year, a spike the company attributes primarily to AI workloads and growth across its supply chain operations. For context, the International Energy Agency projects that global data center electricity usage could double to around 945 TWh by 2030.

The efficiency paradox

The company actually achieved a 12% reduction in data center energy emissions in 2024, thanks to clean energy investments and more efficient hardware. But total emissions still ballooned because demand for AI compute is growing faster than efficiency gains can offset.

Google’s numbers on individual AI tasks are genuinely impressive when viewed in isolation. The median energy consumption for a single Gemini text prompt clocks in at just 0.24 watt-hours, producing 0.03 grams of CO2-equivalent and using 0.26 milliliters of water. The company says that represents up to a 44x improvement in carbon footprint per inference task over the past year.

Google’s data centers are now delivering over six times more compute per unit of electricity than they did five years ago. On the hardware front, Google points to its 7th-generation Ironwood TPU as nearly 30 times more power-efficient than its first Cloud TPU from 2018.

The clean energy balancing act

Google reports an average carbon-free energy matching rate of 66%, supported by contracts for over 8 gigawatts of clean power capacity.

The IEA projects that US data centers are on track to hit record power demand peaks between 2026 and 2027.

What this means for investors

For Alphabet investors specifically, the rising cost of energy procurement is worth watching closely. Securing 8 GW of clean power contracts is not cheap, and the competition for renewable energy capacity is only intensifying as every hyperscaler races to build out AI infrastructure simultaneously. These costs will eventually show up in capital expenditure figures and could pressure margins if AI monetization doesn’t scale proportionally.

European regulators have already shown willingness to impose environmental standards on tech companies, and US policymakers are increasingly focused on the energy demands of data centers. Companies that can demonstrate genuine progress on sustainability, not just efficiency-per-task metrics but actual reductions in total emissions, will be better positioned if that regulatory environment materializes.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.