AI sparks debate at COP30 climate talks

At the United Nations climate talks in Brazil, artificial intelligence is being cast as both a hero worthy of praise and a villain that needs policing.

Tech companies and a handful of countries at the conference known as COP30 are promoting ways AI can help solve global warming, which is driven largely by the burning of fossil fuels like oil, gas and coal.

They say the technology has the potential to do many things, from increasing the efficiency of electrical grids and helping farmers predict weather patterns to tracking deep-sea migratory species and designing infrastructure that can withstand extreme weather.

Climate groups, however, are sounding the alarm about AI’s growing environmental impact, with its surging needs for electricity and water for powering searches and data centres. They say an AI boom without guardrails will only push the world farther off track from goals set by 2015 Paris Agreement to slow global warming.

“AI right now is a completely unregulated beast around the world,” said Jean Su, energy justice director at the Center for Biological Diversity.

On the other hand, Adam Elman, director of sustainability at Google, sees AI as “a real enabler” and one that’s already making an impact.

If both sides agree on anything, it’s that AI is here to stay.

Michal Nachmany, founder of Climate Policy Radar, which runs AI tools that track issues like national climate plans and funds to help developing countries transition to green energies like solar and wind, said there is “unbelievable interest” in AI at COP30.

“Everyone is also a little bit scared,” Nachmany said. “The potential is huge and the risks are huge as well.”

The rise of AI is becoming a more common topic at the United Nations compared to a few years ago, according to Nitin Arora, who leads the Global Innovation Hub for the United Nations Framework Convention on Climate Change, the framework for international climate negotiations.

The hub was launched at COP26 in Glasgow to promote ideas and solutions that can be deployed at scale, he said. So far, Arora said, those ideas have been dominated by AI.

The Associated Press counted at least 24 sessions related to AI during the Brazil conference’s first week. They included AI helping neighbouring cities share energy, AI-backed forest crime location predictions and a ceremony for the first AI for Climate Action Award – given to an AI project on water scarcity and climate variability in the Southeast Asian nation of Laos.

Johannes Jacob, a data scientist with the German delegation, said a prototype app he is designing, called NegotiateCOP, can help countries with smaller delegations – like El Salvador, South Africa, Ivory Coast and a few in the Association of Southeast Asian Nations – process hundreds of official COP documents.

The result is “levelling the playing field in the negotiations,” he said.

In a panel discussion, representatives from AI giants like Google and Nvidia spoke about how AI can solve issues facing the power sector. Elman with Google stressed the “need to do it responsibly” but declined to comment further.

The training and deploying of AI models rely on power-hungry data centres that contribute to emissions because of the electricity needed. The International Energy Agency has tracked a boom in energy consumption and demand from data centres, especially in the United States.

Data centres accounted for around 1.5 per cent of the world’s electricity consumption in 2024, according to the IEA, which found that their electricity consumption has grown by around 12 per cent per year since 2017, more than four times faster than the rate of total electricity consumption.

Environmental groups at COP30 are pushing for regulations to soften AI’s environmental footprint, such as mandating public interest tests for proposed data centres and 100 per cent on-site renewable energy at them.

“COP can not only view AI as some type of techno solution, it has to understand the deep climate consequences,” Su said.

AP

At the United Nations climate talks in Brazil, artificial intelligence is being cast as both a hero worthy of praise and a villain that needs policing.

Tech companies and a handful of countries at the conference known as COP30 are promoting ways AI can help solve global warming, which is driven largely by the burning of fossil fuels like oil, gas and coal.

They say the technology has the potential to do many things, from increasing the efficiency of electrical grids and helping farmers predict weather patterns to tracking deep-sea migratory species and designing infrastructure that can withstand extreme weather.

Climate groups, however, are sounding the alarm about AI’s growing environmental impact, with its surging needs for electricity and water for powering searches and data centres. They say an AI boom without guardrails will only push the world farther off track from goals set by 2015 Paris Agreement to slow global warming.

“AI right now is a completely unregulated beast around the world,” said Jean Su, energy justice director at the Center for Biological Diversity.

On the other hand, Adam Elman, director of sustainability at Google, sees AI as “a real enabler” and one that’s already making an impact.

If both sides agree on anything, it’s that AI is here to stay.

Michal Nachmany, founder of Climate Policy Radar, which runs AI tools that track issues like national climate plans and funds to help developing countries transition to green energies like solar and wind, said there is “unbelievable interest” in AI at COP30.

“Everyone is also a little bit scared,” Nachmany said. “The potential is huge and the risks are huge as well.”

The rise of AI is becoming a more common topic at the United Nations compared to a few years ago, according to Nitin Arora, who leads the Global Innovation Hub for the United Nations Framework Convention on Climate Change, the framework for international climate negotiations.

The hub was launched at COP26 in Glasgow to promote ideas and solutions that can be deployed at scale, he said. So far, Arora said, those ideas have been dominated by AI.

The Associated Press counted at least 24 sessions related to AI during the Brazil conference’s first week. They included AI helping neighbouring cities share energy, AI-backed forest crime location predictions and a ceremony for the first AI for Climate Action Award – given to an AI project on water scarcity and climate variability in the Southeast Asian nation of Laos.

Johannes Jacob, a data scientist with the German delegation, said a prototype app he is designing, called NegotiateCOP, can help countries with smaller delegations – like El Salvador, South Africa, Ivory Coast and a few in the Association of Southeast Asian Nations – process hundreds of official COP documents.

The result is “levelling the playing field in the negotiations,” he said.

In a panel discussion, representatives from AI giants like Google and Nvidia spoke about how AI can solve issues facing the power sector. Elman with Google stressed the “need to do it responsibly” but declined to comment further.

The training and deploying of AI models rely on power-hungry data centres that contribute to emissions because of the electricity needed. The International Energy Agency has tracked a boom in energy consumption and demand from data centres, especially in the United States.

Data centres accounted for around 1.5 per cent of the world’s electricity consumption in 2024, according to the IEA, which found that their electricity consumption has grown by around 12 per cent per year since 2017, more than four times faster than the rate of total electricity consumption.

Environmental groups at COP30 are pushing for regulations to soften AI’s environmental footprint, such as mandating public interest tests for proposed data centres and 100 per cent on-site renewable energy at them.

“COP can not only view AI as some type of techno solution, it has to understand the deep climate consequences,” Su said.

AP