Far from just powering chatbots or trading algorithms, AI has the potential to cut global emissions by a quarter across energy, food, and mobility. It can transform not only how we power our world, but also how we live within it.

Executive Summary 

Artificial intelligence may become one of humanity’s most powerful allies in confronting climate change. A peer-reviewed study estimates that AI could cut 5 gigatonnes of greenhouse gas emissions annually by 2035 across just three sectors: power, food, and mobility. That equals nearly a quarter of projected emissions in those sectors. Crucially, these reductions outweigh AI’s own projected emissions of approximately 1.6 gigatonnes from data centres. Yet beyond the numbers, AI’s deeper significance lies in its power to reshape systems, accelerate discovery, nudge behaviour, and model futures. The true question is not whether AI can help, but whether societies can direct it wisely. 

The Double-Edged Algorithm   

For most of history, humanity has relied on ingenuity to overcome the limits of nature. Fire gave warmth and energy, engines gave speed and scale, and electricity lit up the night. While each breakthrough reshaped the world, it also introduced new risks. Artificial intelligence is the next leap, albeit a technology unlike any other. AI is different not because it is stronger or faster, but because it learns. It can uncover hidden patterns, anticipate outcomes, and act at a pace no human can match. 

Applied to climate change, this adaptive power could slash emissions in energy, food, and mobility by as much as 25%, offering a chance to prevent ecological collapse. Yet the vast computing power that enables such progress also consumes enormous energy, threatening to fuel the very problem it seeks to solve. AI stands as a paradox: a technology with the potential to save us, but only if we can control the cost of its ambition. 

Seeing the Unseeable  

The global economy is a tangled web of cities, energy grids, farms, and transport networks. Climate change frays these threads. No single human mind can track such interactions in real time. But AI can. 

In energy, AI predicts demand and supply with precision, making it possible to integrate intermittent solar and wind at scale. DeepMind demonstrated that AI forecasting could raise wind energy’s economic value by 20%. In cities, AI can optimise transport flows and infrastructure planning. In finance, it can reduce risk by aggregating diverse data, unlocking capital for clean energy projects. 

The shift is not just technical; it is philosophical. For the first time, societies can outsource the governance of complexity itself. 

Discovery at the Speed of Thought 

Reaching net zero requires more than scaling current solutions. It requires technologies that don’t yet exist. The International Energy Agency estimates nearly half of needed emissions cuts by 2050 depend on innovations still in prototype. 

AI unlocks discovery at such an exponential rate that decades collapse into months. A case in point is DeepMind’s GNoME, which identified two million potential crystal structures for new materials, 45 times more than scientists had catalogued before. Similarly, AlphaFold predicted the structures of 200 million proteins, a leap that could accelerate the discovery of novel drugs to treat once incurable diseases. 

In the past, humanity relied on gradual evolution to invent solutions. Today, algorithms are shaping evolution itself. 

Nudging Behaviours towards Sustainability 

Even a hypothetical perfect technology cannot solve climate change if human behaviour resists change. Consumption patterns, after all, account for up to 70% of possible emissions reductions. AI has a unique role here. It can quietly, almost indiscernibly, shift how we live, shop, travel, and eat. 

Google’s Nest learns household heating habits and quietly trims energy use without demanding conscious effort. Google Maps steers drivers along fuel-efficient routes, shaving emissions journey by journey. These tools do not command obedience. Instead, they work by aligning sustainability with convenience. 

Each intervention may seem trivial. But when multiplied across billions of daily decisions, they accumulate into systemic change. This is AI’s subtle power: not to lecture humans into sacrifice, but to weave new habits into the fabric of everyday life. By making the greener choice the easier choice, AI transforms sustainability from an act of willpower into a default mode of living. 

Harnessing Intelligence for a Cooler Planet 

AI will not solve climate change single-handedly. But it could become one of our most powerful allies. Across power, food, and mobility, it offers the potential to cut emissions by as much as 25%. These are significant reductions that far outweigh the tech’s own energy demands. More than just numbers, AI gives us new ways to see complex systems, discover solutions faster, and shape everyday choices. The real opportunity is clear. If we direct this intelligence toward sustainability, we can move from merely reacting to crises to actively building a cooler, more resilient planet. 

By making the greener choice the easier choice, AI transforms sustainability from an act of willpower into a default mode of living