In the Amazon rainforest in Brazil, an AI-powered drone sweeps through the canopy. It is elegant, almost beautiful, in its efficiency. It identifies illegal logging, capturing even slight changes in vegetation density.
Three thousand miles away, in Arizona’s drought-stricken communities, the data centres powering such AI systems drain local aquifers. This is today’s climate-tech paradox: Solutions that increasingly resemble the problems they claim to solve.
If 2024 marked humanity’s first breach of the 1.5° Celsius threshold, 2025 opens to a perfect storm: a fractured climate diplomacy still reeling from Baku’s acrimonious compromise on global climate action; the spectre of a second Donald Trump presidency; and a tech industry whose climate promises have proven hollow.
While the Silicon Valley script remains familiar — carbon-capture breakthroughs, a nuclear renaissance, AI-powered climate solutions — Microsoft and Google both reported higher emissions in 2023, driven by AI-linked data centres. Hundreds of others are way off their targets too.
It’s a precarious moment: the loudest champions of technological solutions, men such as Elon Musk, Marc Andreessen and David Sacks, are poised to influence a presidency historically hostile to climate action. Their accelerationist vision, which prioritises unfettered technological development over environmental concerns, threatens to turn climate tech from a tool of mitigation into one that will deepen global inequities.
{{/usCountry}}It’s a precarious moment: the loudest champions of technological solutions, men such as Elon Musk, Marc Andreessen and David Sacks, are poised to influence a presidency historically hostile to climate action. Their accelerationist vision, which prioritises unfettered technological development over environmental concerns, threatens to turn climate tech from a tool of mitigation into one that will deepen global inequities.
{{/usCountry}}This disconnect between promise and reality was starkly visible at the 2024 Conference of Parties (COP) in Baku.
There, a $300 billion climate finance deal, barely a third of what developing nations sought, was gavelled through despite protests from the Global South.
The tech industry’s approach mirrors this pattern. The Frontier coalition, backed by giants such as Alphabet, Meta and Shopify and until now regarded as a standout private-sector effort to advance carbon removal technologies, recently announced an $80 million commitment to carbon credits — a sum that would barely register against the adaptation needs of a single vulnerable nation. Meanwhile, energy providers delay retiring coal plants in order to meet tech giants’ voracious power demands for their AI infrastructure.
And so the corporate initiatives, much like the Baku finance framework, offer the appearance of action while carefully preserving existing power structures.
The implications unfold far from Silicon Valley boardrooms. In Taiwan, semiconductor manufacturers won water allocation rights over farmers during one of the country’s worst droughts in a century, in 2021.
In Ireland, data centres devour a fifth of all power generated, and have prompted a moratorium on new facilities until 2028.
In Indonesia’s nickel belt, entire communities are being displaced to feed the electric-vehicle supply chain.
Technologists such as Andreessen, Sacks and Sriram Krishnan believe unfettered technological progress, free from regulatory constraints and equity considerations, will solve the climate crisis, and dismiss multilateral climate frameworks as unnecessary impediments to innovation.
Even Musk, whose Tesla has arguably contributed to the goal of reducing emissions, has aligned himself politically with climate-action sceptics.
Where does this leave us?
Advocates of climate action say it is time for a sort of “Manhattan Project for climate”.
The parallel is seductive: Faced with an existential threat, we mobilise our technological might just as we did when seeking to defeat Nazi Germany. But this comparison, at best, illustrates how hard it is to talk about climate solutions. At worse, it reveals how much misunderstanding persists.
The Manhattan Project succeeded by creating a weapon of unprecedented power, consolidating technological superiority in the hands of a single nation and its allies. What followed was decades of nuclear proliferation as other nations rushed to achieve parity, creating new forms of vulnerability and dependence.
The nature of the climate crisis is such that any similar concentration of resources would be counterproductive. Wealthy nations developing proprietary green solutions while forcing developing countries to either accept dependence or undertake costly parallel development programmes cannot help.
The Manhattan Project needed only to produce a working weapon; addressing climate change requires unprecedented global cooperation — including among economic, geopolitical and even social rivals.
The choice ahead isn’t between embracing or rejecting technology. It’s between using technology to reinforce existing power structures or transforming how we do things, from the ground up, in ways that will undoubtedly be challenging for everyone. But will make the future of humanity more secure.
The real test in 2025 will be whether national climate plans can show a clear, walkable path to reduce fossil-fuel consumption while prioritising human dignity.
This goal is simple and it has no room for technological spectacle. The sooner the world recognise this — the sooner it sees through the techno-utopian illusion — the sooner the real work of securing a liveable future can begin. Not just for those who can afford the latest green technology, but for all.