AI could help cut up to 5.4 Giga tonnes of CO₂e annually by 2035 research says, with examples from Amazon, Google, Oracle and more driving climate action
AI is increasingly recognised as a key enabler of global climate action.
In a 2025 study, Green and intelligent: the role of AI in the climate transition, Stern et al. argues that AI could cut global emissions by up to 5.4 Giga tonnes CO₂e annually by 2035 – just from its applications in the power, food and mobility sectors.
This saving outweighs projected emissions increases from AI-driven data centre energy use, suggesting a strong net-positive climate contribution.
As environmental pressures mount, the report presents AI not as an added risk but as a critical tool to unlock system-wide efficiencies, drive innovation and support sustainable economic growth across both developed and emerging markets.
Transforming complex systems with AI
AI can revolutionise how businesses manage and optimise large-scale systems across energy, cities and transportation.
For instance, grid instability from intermittent renewable sources like wind and solar can be mitigated by AI models that forecast supply and demand and manage distributed energy resources.
Singapore’s National AI strategy and Google DeepMind’s wind optimisation tool, which boosted the value of wind energy by 20%, are already proving this concept.
AI can also enhance access to sustainable investment in emerging markets by reducing information asymmetries.
“The paper discusses the ways AI can play a powerful role in supporting climate action while boosting sustainable and inclusive economic growth,” writes Kate Brandt, Chief Sustainability Officer at Google, on LinkedIn.
Kate Brandt, Chief Sustainability Officer of Google
“Lord Stern spoke about three key sectors, power, food and mobility, which collectively contribute nearly half of global emissions and the ways AI can reduce their emissions.”
By aggregating diverse data sources, AI supports better risk prediction, enabling more efficient capital flow into climate solutions.
Accelerating resource efficiency
Nearly half of the emission cuts needed by 2050 depend on technologies still in development.
AI is accelerating this innovation, with tools like DeepMind’s GNoME identifying 2 million new materials that could transform renewable energy storage.
In industrial sectors, AI also improves efficiency and reduces waste.
Amazon’s AI-powered packaging optimisation has saved more than 3 million tonnes of material since 2015.
Start-ups such as GreyParrot are also demonstrating how AI can radically increase recycling rates using computer vision.
Credit: Green and intelligent: the role of AI in the climate transition, Stern et al. Projected annual global emissions in AI scenario vs. BAU and ambitious emissions reduction scenario by 2035 for the sectors in scope (Power, Meat and Dairy, Light Road Vehicles). Note, the ambitious emissions reduction scenario is calculated using the IEA’s net zero emissions scenario42 for Power and Light Road Vehicles and UNEP’s 2050 Paris-aligned target3 for Meat and Dairy
These advances improve not just productivity but also circularity and sustainability across value chains.
Behavioural change for green lifestyles
AI can help reduce emissions by shaping consumer behaviour.
Smart technologies like Google Nest and Oracle Opower use real-time data and behavioural science to help users reduce energy use.
In the food sector, Winnow Vision’s AI-powered cameras identify waste patterns, helping chefs reduce food waste across more than 3,000 kitchens.
By offering personalised, low-impact choices, such as Google Maps’ fuel-efficient route suggestions, AI empowers individuals to make more sustainable daily decisions.
Credit: Green and intelligent: the role of AI in the climate transition, Stern et al. Total emissions and emissions savings from AI in 2035 for the sectors in scope (Power, Meat and Dairy, Light Road Vehicles). Note, the 2023 bar is the total 2023 GtCO2e emissions of power (15.3 GtCO2e), meat and dairy (8.7 GtCO2e), and light road vehicles (3.2 GtCO2e) sectors
These tools support shifts in consumption behaviour, crucial for long-term climate goals.
Enhancing climate modelling and adaptation
Accurate climate modelling is essential for informed policy.
AI-enhanced tools like IceNet outperform traditional forecasting systems in predicting Arctic sea ice, while Climate Policy Radar uses AI to help governments design better climate policies based on thousands of case studies.
On the resilience front, AI strengthens early warning systems.
Google’s FloodHub, for example, forecasts floods five days in advance in 80 countries, helping to prevent billions in economic damages annually.
AI-powered digital twins, such as NVIDIA’s Earth-2, are also redefining how we model extreme events and long-term ecological change.
Public leadership and governance
Despite AI’s promise, letting the market govern its development alone risks exacerbating global inequalities and energy use.
Governments could incentivise AI innovation for climate good, regulate data centre energy use, and invest in equitable digital infrastructure, especially across the Global South.
Public policy could also ensure AI aligns with inclusive climate strategies, prioritising public benefit over private interest.