How “AI” greenwashing became the climate movement’s new fault line

Hey there, Earth Month has again felt quieter on the greenwashing front, with more brands apparently giving up even the pretence that they care.

But greenwashing is alive and well in some corners, and it’s particularly prevalent right now around the use of so-called “artificial intelligence.” “AI” is perhaps the buzziest and most divisive term right now in the fashion industry (for instance, the State of Fashion 2026 report from the Business of Fashion and McKinsey named “AI” one of its defining themes for the year).

And the greenwashing around the tech rarely faces the kind of scrutiny that fashion brands do. In November, we reported on the greenwashing around “fashion slop,” where replacing creative workers with generative “AI” is sold as a way to reduce carbon emissions.

Zooming out beyond the fashion context, the fossil fuel-powered boom in data centres is often defended with vague claims that “AI” will somehow “solve” climate change. That smelt off to climate analyst Ketan Joshi, who has checked the sources on more than 150 claims about AI and the climate.

JD spoke to Ketan for this week’s examination of the greenwashing playbook.

Thanks for reading!

JD and Amy

Want to sour the mood at your next function? It’s easy: simply bring up “AI,” and watch the conversation quickly go off the rails. At least in climate and sustainability circles, nothing divides a room quite like artificial intelligence does.

You’ve no doubt heard some variation of the claim that “AI” will somehow solve the climate crisis. You’ve also probably heard the warnings: that the rapid expansion of all the data centres needed to deploy more of this technology comes with considerable environmental costs. There’s a widening rift, with competing voices making wildly different arguments about the tech’s climate impacts.

Is it all just hot air? Is anyone backing up their claims with data? To try and get some answers, independent climate analyst Ketan Joshi recently published the first report of its kind. In it, he interrogates the argument that “AI” will have net-positive benefits for the climate. Supported by a consortium of environmental organisations — including Stand.earth, Beyond Fossil Fuels, and Climate Action Against Disinformation, among others — he looked into the green claims companies make about “AI.” He checked the footnotes and scoured the available data to try and come to some conclusions.

In the end, Joshi found zero verifiable evidence that the new wave of consumer generative “AI” was reducing emissions in any way, despite what the Big Tech companies would have you believe.

Yes, some forms of “AI” do have positive environmental benefits, as we get into in the conversation that follows. But Joshi found that those benefits were being misattributed to generative “AI” systems like ChatGPT, which consume up to 13 times more energy than the lower-energy technologies actually delivering the climate gains.

The vast majority of “AI” climate claims are unproven, and companies are using a greenwashing playbook borrowed from the fossil fuel industry to avoid accountability.

It’s tempting to tune out of “AI” discourse entirely given all the other bad things happening in the world. But if you care about environmental justice, this is hard to ignore, as it springs from the same systems of exploitation as the climate crisis itself. “AI” is the reason Big Tech companies have backed away from their greenhouse gas emissions targets, and running them requires burning more fossil fuels at the exact moment climate scientists say we must be phasing them out.

With all the competing claims, I wanted to hear Joshi’s take on this rift. His conclusion: The evidence that generative “AI” will have any positive environmental impact is missing, while the evidence for substantial environmental harm is strong. Worse, all this attention paid to “AI” is distracting the climate movement from the actual solutions. Here’s what Joshi had to say:

JS: I’d like to begin with what a lot of people find frustrating about the moment we're in: we have this rift in the climate movement between folks concerned about the fact that the world’s richest companies are significantly increasing their environmental impacts to build massive data centres for “AI” versus those who seem to think “AI” is somehow going to save us from the climate catastrophe, which it’s worsening. How did we get here?

KJ: Back in 2020 and 2021, when a lot of Big Tech companies began announcing their net zero targets, it felt pretty good. I had a sense of trust in what they were saying. And as a technology fanboy myself, I always kind of thought, well, yeah, these companies, they’ve got it sorted, right? They generally tend to consume electricity. They have enough money to fund renewable energy. They have good knowledge of technical solutions around integrating renewables, and they’re probably going to do a good job getting in line with their climate targets.

But now we see most of these companies veering away from the targets they set for themselves.

When you look at some of the charts that I provide in the report, it is such a betrayal of that enthusiasm and trust. Their emissions are significantly rising because of the over-deployment of generative AI. I cannot believe how there are still people who listen to those companies and go, “OK, well, yeah, they’re going to get right back on track soon.”

Part of it is simply the fact that a lot of those tech companies have ingratiated themselves with sections of the climate movement. They’re considered to be trusted partners, even though I would argue that they’ve broken that trust many times over.

Another part of it is that people who work in the climate space often come from business backgrounds where they’re primarily focused on growth and profits. And that generally tends to bring a mindset where the physical nature of climate — the impacts, the fossil fuel emissions, biodiversity loss, water usage, etc — takes a back seat. And if there is one issue where you will find the narratives of growth and profits being put ahead of the current climate harms, it’s going to be AI.

And then this all sits in the broader context in which climate action itself has been deprioritised in both the media and among elite institutions.

JS: “AI” is one of those buzzwords that means everything and nothing simultaneously. It speaks to a range of technologies that have vastly different impacts. Some technologies you refer to as “traditional AI” in your report genuinely do help with climate modelling or grid optimisation. But that’s not what’s driving the data centre boom, and it’s not what most people mean when they talk about ChatGPT, Claude and other “generative AI” products. So before we get too far into it, what do we mean when we say “AI”?

KJ: AI is a really loaded term. It contains the implication of intelligence itself. But intelligence is something very specifically that applies to organic beings. These software systems are just not that. They’re doing interesting things sometimes, but they’re not intelligent.

There’s a really good book by Karen Hao, Empire of AI, in which she mentions how the term “artificial intelligence” was basically invented as a marketing term. And as you say, it doesn’t refer to one specific kind of technology.

That can be deceptive because some kinds of AI have far worse impacts than others. Older machine learning systems are often lower-energy compared to the very energy-intensive generative AI systems we see in ChatGPT today, which is what’s driving the data centre boom.

Generative AI is where the greenwashing is really happening around the tech.

JS: Karen Hao has that useful analogy where she says the term “AI” is like saying transportation without distinguishing between a bicycle and a rocket, both will get you somewhere but have very different impacts. You’ve tracked companies taking advantage of that vagueness, where they make these grand environmental claims to justify building more data centres. Is that vagueness an intentional greenwashing strategy?

KJ: Yeah, you can actually find some related examples of this in the past. Tech companies today are using an “upgraded” greenwashing playbook that fossil fuel companies have used for years.

Carbon capture and storage is one of the most fascinating greenwashing tactics of the entire fossil fuel economy because it just has been going since the 1970s and it has just lasted even though it has failed to live up to its promises for decades and decades.

I notice when I’m reading fossil fuel company reports, they’ll often really centre climate solutions – and often non-transformative climate solutions, right? So in a fossil fuel company report, you won’t see that much wind and solar, but you will see stuff like algae, carbon capture and carbon offsets.

This is really important: it’s not just that they’re providing a false promise. It has to be something that is being exaggerated or overstated. It has to be a distraction, and then the distraction can’t work. Because if it works too well, then somebody might hold your company to account.

That was why I started to pay attention to the way tech companies were talking about AI as a climate solution, because I generally assumed that it would have those two same characteristics.

So first of all, it has to distract from the main core of their climate impacts. Then secondly, it probably doesn’t work anywhere near as well as they say it does.

JS: You’ve become increasingly vocal about “AI” in the context of the climate crisis at a time when lots of your peers are generating their LinkedIn posts with ChatGPT. When did you start connecting those threads and feel the need to focus on this more?

KJ: It was one thing, actually. It was a Google statement I saw in 2023. I was looking at their sustainability report out of curiosity, and they said in their report that AI deployment could reduce global emissions by between 5% and 10%. I was like, gosh, that’s a lot. Five to 10% of global emissions is so massive. And that was by 2030.

So, I traced it through to the source of that claim, and it was based on a Boston Consulting Group report from a couple of years earlier.

What had happened was, someone from Boston Consulting Group had a conversation with a client, and that client basically said, I reckon that our emissions have gone down by 5% to 10% thanks to our use of AI. And that anecdotal conversation was extrapolated to Google saying not just in their reports but as recently as April 2025 in a set of policy recommendations to the European Union that research shows that AI could reduce emissions by 5% to 10% globally.

I was so floored by that. This is Google! Google can afford to pay a consultant to make a report that, at least somewhat, justifies this claim. I was just so shocked by it and I was just like, what is going on here? That seemed so weird and so significant.

JS: You’ve analysed more than 150 claims from companies and institutions saying “AI” would deliver climate benefits, and then you tried to track and verify the citations: 74% turned out to be unproven. Given how widespread that idea is — that generative “AI” might help solve climate change despite its many negative impacts — you think that would have stronger sources to back it up?

KJ: The majority of claims around AI having a climate benefit either cited corporate websites, like the companies making the software, for instance, or they cited no evidence at all.

One example is the largest cruise line company in the world, Carnival Corporation. There’s a line in the International Energy Agency report on AI where they say that they reduced fuel consumption by 5% with AI. That’s massive, right? This is the biggest cruise line company in the world. And I was like, damn, let’s see what the reference is here. I went to the website, and it’s a website called Sailor Speaks. Not a single reference, not a single hyperlink, no author. Every article on this website is the same. It’s just like reams and reams of text with no references, no clear human voice. I don’t know for sure, and you can ultimately never prove it, but I strongly suspect this entire website is just generated by a chatbot.

Then I went and checked the Carnival Corporation sustainability reports to see if they just said it in a report somewhere and it was just reproduced in this article. But I cannot find anywhere that Carnival Corporation makes this claim.

I bring up that example because it shows how all of this greenwashing now exists in this substrate of digital decay and slop everywhere. It was quite an experience learning how this whole system of exaggeration of a potential climate solution is now operating in an environment where it’s very easy to fabricate text.

JS: You’ve advocated for renewable energy and energy justice throughout your career. You’re clearly somebody who’s interested in tech. So it must be frustrating when people accuse critics of generative “AI” as being anti-technology or anti-progress?

KJ: I discovered this very early on in my career with wind power that you can’t just bluntly argue for more of a thing. You need to actually think about how it’s deployed and why it’s deployed. I think this is very much the same.

I feel very strongly about being an advocate for technology that is purposeful and fair. It’s driven by that love, to want to rectify the wrong that’s occurring.