Google’s carbon footprint balloons as it goes all in on AI

Despite its ambitious climate goals, from net-zero emissions to carbon-free energy, Google’s greenhouse gases are on the rise. Like its competitors, the company points to artificial intelligence as the culprit.

Google wants to cut its pollution in half by 2030 compared to its 2019 base year. But the company’s total greenhouse gas emissions have actually ballooned. According to its most recent environmental report, carbon emissions rose to 13.3 million metric tonnes last year, an increase of 48 per cent since 2019. Last year alone, its emissions hit 14.3 million metric tonnes, a 30 per cent increase on the previous year.

Google attributes the jump largely to increases in the electricity consumption of its data centres, which are needed for the broad rollout of artificial intelligence, and supply chain emissions.

Ambitions under threat

“As we further integrate AI into our products, reducing emissions may be challenging due to increasing energy demands from the greater intensity of AI compute, and the emissions associated with the expected increases in our technical infrastructure investment,” the company said in the report. 

Data centres play a crucial role in training and operating the models that underpin AI models like Google’s Gemini and OpenAI’s GPT-4. But they are huge energy consumers. So much that they might endanger the tech companies’ zero or negative carbon targets, something that was signalled earlier this year by Google’s competitor Microsoft. 

Promises to reduce CO2 emissions thus clash with promises to invest heavily in AI products across the globe. At the same time, there are growing questions about increasing water use as a result of AI hardware production.

^{The Google logo at the company's Mons data centre site © BELGA PHOTO NICOLAS MAETERLINCK}