Google’s greenhouse gas emissions are increasing thanks to AI

new York
CNN
—

As Google has rushed to integrate artificial intelligence into its core products—with sometimes lackluster results—there is a problem brewing behind the scenes: The systems needed to power its AI tools have significantly increased the company’s greenhouse gas emissions.

AI systems need a lot of computers to run. The data centers needed to run them, essentially warehouses filled with powerful computing equipment, consume tons of energy to operate. process data and manage the heat produced by all those computers.

According to the tech giant’s annual environmental report, Google’s greenhouse gas emissions have increased by 48% since 2019. The tech giant attributes this growth primarily to “increasing data center energy consumption and supply chain emissions.”

Google now calls its goal of reaching net-zero emissions by 2030 “extremely ambitious,” and adds that this commitment will likely be affected by “uncertainty surrounding the future environmental impact of AI, which is complex and difficult to predict.” In other words: a sustainability strategy driven by the company — which once included the The slogan “Don’t be evil” in its code of conduct — has become more complicated thanks to AI.

Google, like other tech rivals, has invested heavily in artificial intelligence, widely seen as the next major technological revolution that is set to change the way we live, work and consume information. The company has integrated its Gemini generative AI technology into some of its major products, including Search and Google Assistant, and CEO Sundar Pichai has called Google an “AI company.”

But AI has a major drawback: the energy-hungry data centers that Google and other Big Tech rivals are currently spending tens of billions of dollars each quarter to expand to fuel their AI ambitions.

The International Energy Agency estimates that a Google search query requires an average of 0.3 watt-hours of electricity, while a ChatGPT query typically consumes about 2.9 watt-hours. An October study by Dutch researcher Alex de Vries estimates that the “worst-case scenario” suggests that Google’s AI systems could potentially consume as much electricity as Ireland each year, assuming widespread adoption of AI in their current hardware and software.

“As we further integrate AI into our products, reducing emissions may be challenging due to the increasing demand for energy from the increased computational intensity of AI and the emissions associated with expected increases in our investments in technical infrastructure,” Google said in its report, released Monday. It added that data center electricity consumption is currently growing faster than it can bring carbon-free electricity sources online.

Google said it expects its total greenhouse gas emissions to continue to increase before decreasing as the company seeks to invest in clean energy sources, such as wind and geothermal, to power its data centers.

The large amounts of water used as a coolant to keep data centers from overheating also pose a sustainability challenge. Google says it aims to replenish 120% of the fresh water used in its offices and data centers by 2030. Last year, it replenished just 18% of that water, though that was up sharply from 6% the year before.

Google is also among the companies experimenting with ways to use AI to combat climate change. A 2019 Google DeepMind project, for example, trained an AI model on weather forecasts and historical wind turbine data to predict wind power availability, helping increase the value of this renewable energy source for wind producers. The company has also used AI to suggest more fuel-efficient routes for drivers using Google Maps.

“We know that scaling AI and using it to accelerate climate action is just as critical as addressing the environmental impacts associated with it,” Google said in the report.