AI climate tools draw COP31 focus amid carbon concerns

Artificial intelligence is set to be a focus at next month's COP31 climate summit in Turkey, where the hosts say they will push for “climate-responsible AI” as concerns grow over the energy use of data centres.

The debate has centred on the carbon footprint of the data centres behind AI systems, including general-purpose language models. But experts cited in the discussion say AI is also being used to support climate action, including by improving weather forecasts, monitoring emissions and helping electricity grids handle solar and wind power.

Siddharth Singh, an AI expert at the International Energy Agency, said data centres raise “genuine concerns” for energy systems and climate, but said AI is also a tool whose effect depends on how it is deployed. The IEA says AI use across transport, industry and buildings could save 13.5 exajoules of energy by 2035, equivalent to Indonesia's annual energy demand.

AI can help forecast cloud cover, sunlight and wind, allowing operators to better predict renewable power output and reduce curtailment when grids cannot absorb electricity. WattTime co-founder Gavin McCormick said AI is useful for forecasting when power is cleaner, helping companies time the charging of electric vehicles, phones and other devices.

Florian Pappenberger, director-general of the European Centre for Medium-Range Weather Forecasts, said AI has made weather forecasting faster while remaining grounded in physics-based models. Google says one of its systems can forecast river floods up to seven days in advance in more than 150 countries.

AI is also being used with satellites to identify greenhouse gas emissions from sites such as oil and gas facilities, mines and landfills. McCormick said Climate TRACE, an initiative he co-founded, uses AI to build emissions inventories for cities and other users.

The technology also has climate risks. Maria Joao Sousa, executive director of Climate Change AI, said oil and gas companies have long used AI to analyse geological data, find reservoirs and streamline fossil fuel production. She said AI “will never be a silver bullet” and that technology alone cannot solve climate change.