Google has released its 2024 Environmental Report, an 80-plus-page document detailing all the efforts the company is making to apply technology to environmental issues and mitigate its own impacts. But it completely dodges the question of how much energy AI uses — perhaps because the answer is “way more than we’d like to say.”
You can read the full report here (PDF)and honestly, there’s a lot of interesting stuff in there. It’s easy to forget how many plates a company as big as Google is spinning, and there’s some truly remarkable work here.
For example, he is working on a water replenishment programthrough which it hopes to offset the water used in its facilities and operations, ultimately creating a net positive outcome. It does this by identifying and funding watershed restoration, irrigation management, and other work in this area, with dozens of such projects around the world funded at least in part by Google. It has managed to replenish 18% of its water consumption (however that word is defined here) in this way, and is improving every year.
The company is also careful to highlight the potential benefits of AI for climate change, such as optimizing watering systems, creating more fuel-efficient routes for cars and boats, and predicting floods. We’ve already highlighted some of these benefits in our AI article , and they could actually be very useful in many areas. Google doesn’t need to do this kind of thing, and many large companies don’t. So credit where credit is due.
But then we come to the section “Responsibly managing AI resource consumption”. Here Google, so sure so far about all the statistics and estimates, suddenly spreads its hands and shrugs its shoulders. How much energy does AI use? Can anyone Really be certain?
But that must be bad, because the first thing the company does is downplay the entire data center energy market, claiming that it accounts for only 1.3% of global energy consumption, and that the amount of energy used by Google is only 10% of that figure at most. So, according to the report, only 0.1% of all global energy powers its servers. A trifle!
In 2021, the company notably decided to reach the goal of net zero emissions by 2030, although it admits that there is a lot of “uncertainty,” as it likes to say, about how this will actually happen. Especially since its emissions have increased every year since 2020.
In 2023, our total GHG (greenhouse gas) emissions were 14.3 million tCO2e, representative an increase of 13% year-on-year and 48% compared to our target baseline year of 2019. This is primarily due to increased data center energy consumption and supply chain emissions. As we further integrate AI into our products, reducing emissions may be challenging due to the increasing energy demand from increased AI computational intensity and emissions associated with expected increases in our technical infrastructure investments.
(Emphasis added here and in the quote below.)
But AI’s growth is lost in the aforementioned uncertainties. Google offers the following excuse for why it doesn’t specify the contribution of AI workloads to its overall data center energy bill:
Predicting the future environmental impact of AI is complex and evolving, and our historical trends likely do not fully reflect the future trajectory of AI. As we deeply integrate AI into our product portfolio, the distinction between AI and other workloads will be meaningless, so we focus on data center-wide metrics because they include the overall resource consumption (and therefore environmental impact) of AI.
“Complex and evolving”; “trends probably don’t fully reflect”; “the distinction…won’t make sense”: this is the kind of language used when someone knows something but would really, really rather not tell you.
Does anyone really believe that Google doesn’t know, down to the penny, how much AI training and inference has added to its energy costs? Isn’t being able to break down those numbers so precisely part of the company’s core competency in cloud computing and data center management? The company has all these other claims about how efficient its custom AI server units are, how it does all this work to reduce the energy required to train an AI model by 100x, etc.
I am sure that Google is doing a lot to protect the environment, and you can read more about it in the report. But it is important to highlight what the company apparently refuses to highlight: the enormous and growing energy cost of AI systems. The company may not be the main culprit for global warming, but despite its potential, Google does not seem to have achieved a net positive result yet.
Google has every incentive to minimize and blur these numbers, which, even in their reduced and highly efficient state, can hardly be good. We will be sure to ask Google to be more specific before we find out if the numbers get even worse in the 2025 report.
