Naomi Klein · Scour

This piece is part of Reframing Impact, a collaboration between AI Now Institute, Aapti Institute, and The Maybe. In this series we bring together a wide network of advocates, builders, and thinkers from around the world to draw attention to the limitations of the current discourse around AI, and to forge the conversations we want to have.

In the run-up to the 2026 India AI Impact Summit, each piece addresses a field-defining topic in AI and governance. Composed of interview excerpts, the pieces are organized around a frame (analysis and critique of dominant narratives) and a refra…

Naomi Klein is an award-winning, New York Times bestselling author of nine critically acclaimed books published in more than thirty-five languages. She is an associate professor of climate justice and the codirector of the Centre for Climate Justice at the University of British Columbia.

In this interview, Klein argues that the current “arms race” model for large-scale AI, driven by unchecked compute demands and breakneck corporate competition, is fundamentally unsustainable and incompatible with climate goals. Instead, the resource intensity of the “bigger is better” AI paradigm is proving to be a lifeline to the fossil fuel industry, particularly as renewable markets begin to emerge. She describes this dynamic as a “three-legged stool” powered by tech, fossil fuels, and the state, creating artificial markets for both the fossil fuel sector and often faulty AI technology by targeting governments as customers. Klein suggests that positive, sustainable uses of AI can exist—but they are smaller and more specialized. She calls the fight against unchecked AI development the “most unifying struggle” she has seen in her life, and cites examples of place-based, localized resistance as the beginnings of a movement.

Following is an edited transcript of the conversation.

Frame

Saying that AI can be made sustainable or can contribute to climate goals is not realistic, and obfuscates an emerging interdependence between tech, fossil fuels, and the state.

The “arms race” model of AI development is fundamentally incompatible with sustainability goals.

“AI” is such a huge term, but here we’re talking about generative AI, large language models, and specifically these “arms race” models. You have an arms race between countries, and you also have an arms race between companies. The only governing principle of it is compute—more and more and more. It’s a brute-force way of seeing the world.

To have so many titans fighting for this one prize at the same time is the most unimaginably wasteful use of resources. Because everyone is duplicating. Everybody is devouring the same data. Everybody is building identical, massive data center clusters or “AI factories.” The idea that this could somehow be made sustainable? I don’t see how we could believe that at this point in history.

There was a period where [tech] companies were positioning themselves as something “dematerial,” something ethereal and blue, which was always a lie. They wanted us to imagine a cloud that was off in the sky and fluffy. And it was always not that. It was always these very real, material data centers, and there were always many, many exploited people fueling the work in the background.

So, there was a “greenwashing” phase—but I do think that [tech companies] saw themselves as something futuristic and greener and definitely not in the dirty industry of the past. [For example], Amazon positioned themselves very aggressively as a climate leader. It was always an empire built on cardboard boxes, but they did invest in huge ways in renewables.

The catch [now] is there’s just no way for you to chase this bubble and hold on to the things that you claimed about climate—not if you’re going to do the arms race model. We’ve seen this most clearly with Elon Musk. Most people thought of him as the highest-profile green tech entrepreneur on the planet. And this is the guy who decides he’s going to power Grok with methane and poison a city. They had to choose and they’ve made their choice. So we really shouldn’t play along with this idea that it can be green.

The “three-legged stool” powering AI consists of tech, fossil fuels, and the state, creating new markets where each “leg” was faltering.

We now have “a three-legged stool” of this emergent fossil-tech economy. One leg is tech, one leg is fossil fuels—looking for their new market—and then the other leg of the stool is the state. And that includes the police state and the military state, gobbling up these technologies. The public is really not part of this at all. But those three are reinforcing each other at the moment.

I think it’s really important for us to understand that the fossil fuel sector was panicking about the fact that solar was for the first time cheaper than gas. We weren’t all the way there, by any means, but we were just starting to see some of the numbers go in the right direction around the price of renewables, and around big economies like India, Pakistan, and China really getting serious about electrification and renewables. And at this very moment when markets are starting to collapse, what do you know? This lifeline is extended to the fossil fuel sector and to build all of these data centers and power them with fossil fuels.

Donald Trump […] can do a lot of things. He can roll back all kinds of climate regulations and he could pull the US out of the Paris Accord—but he can’t actually create new markets for these companies. And that is what AI has done.

For every country it’s different. In Canada, we’re on the verge of making this mistake because we had built a lot of LNG infrastructure and Asia was not as interested in buying it because there’s a renewable transition that is happening. So, they’ve got a whole bunch of liquefied natural gas and they have to figure out who they’re going to sell it to. So now they’re talking about building data centers to justify the fossil fuel infrastructure that they’ve just built.

They are afraid to introduce something that we’ve been calling for for years, which is a Green New Deal. We need a plan for our economy to get off fossil fuels and create jobs for everyone, meaningful work in this transition for everyone who wants a job. But that’s harder than just accepting a contract from Google or some big LNG company in Malaysia. So I think it’s a failure of imagination, a failure of leadership in the way they think about how to build an economy.

[And] because these companies have built out all of this capacity, they need customers really badly. And the market is not liking their product. Individuals love using free ChatGPT, or paying a little bit, but certainly not the actual price. And we’re hearing more and more that businesses aren’t seeing the kinds of efficiencies that they promised. So there is no market for what they have just built. There’s a huge gap. And unfortunately, governments are “marks.” They’re buying faulty tech because they are afraid of being left behind.

And there is an offer being made to political leaders around the world: We’ll cut you in and at the expense of your people. These are leaders who come to power with hypernationalist promises and they’ll use phrases like “data sovereignty” and make it sound like they’re standing up for their countries, but their loyalty is to this very global class that is banking on the machines at the expense of life.

Reframe

Klein argues that smaller, more curated versions of AI could be useful, but we are funding the wrong things: massive build-outs. Hyperlocal resistance and collectively dreaming of a better future are the antidote, she says.

Positive uses for AI depend on smaller, more cared-for datasets and investments in research—areas that are ironically being cut. We need to surface these trade-offs.

I don’t see how [the current trajectory] can be made sustainable, which is not the same as saying that I don’t see potential positive uses for AI in the climate sphere. But they’re not these massive build-outs.

We actually know the things we need to do. And we have the tech. So let’s use the tech. And let’s not build an artificial mirror world of AI. We don’t have the tech for that. Because even green tech has a cost. We can’t do both. We will either transition our economy off of fossil fuels and lower our emissions in time to avoid maybe two degrees—it’s getting harder and harder—or we will build this parallel world covered in data centers in order to summon a digital god.

Those are separate issues that are being conflated very deliberately. They’re using the good things that we want. We want to use this for medical research, for more efficient use of renewables, to bring back languages that are going extinct. There are things that we can do, but all of those good uses of AI really depend on good data going into the systems. And these are smaller, more curated, more cared-for data sets.

The irony is that at the same time as this is being rolled out, we have attacks on that kind of research in postsecondary education. So any government that is serious about having a valuable niche would be funding science. They would be funding postsecondary education. They would be valuing people who are creating the good data.

Localized resistance is emerging, and it is the way we can collectively come up with better visions for humanity than the tech-led version.

I think we’re in a latent phase, where the movement is finding itself in localities. Specific places have been placed on the front lines of these fights, like Memphis and [ … ] Spain. In water-stressed regions people are really clear: “Your cloud is not more important than our river.”

What gives me hope is that I actually think this is the most unifying struggle I have ever seen in my life. I have never seen a fight that has the potential to bring people together across partisan lines, to bring workers together with environmentalists instead of being pitted against each other.

I think it makes sense that it begins hyperlocal, because really it’s about saying that this world is worth protecting. And that isn’t an abstract concept. We never save the whole world. We save our places that we’re connected to that we love.

What really moves me is when I talk to local organizers who really understand that this is a kind of a spiritual work that is really about connecting to their place and to each other, and coming from a place of love. I’ve seen this in struggles against fossil fuel infrastructure in the past. In Canada, they were able to stop a pipeline called the Northern Gateway Pipeline. These incredible Indigenous youth organized in northern communities, and they were writing love letters to their communities—to the orcas and the bears and the salmon. It was not just “we hate your pipeline,” but “we love our place.”

Our power is in our ability to think with each other, to genuinely generate new ideas. That happens when we sit down with each other and bounce ideas off of each other, and where we’re not just a sycophant machine or regurgitating things that other people have already said. Go for a walk with your friends and try to dream a more beautiful future than that.

Research Areas