– Advertisement –
Over 100 partners, including 37 tech companies, 11 countries and five international organisations, have joined forces under the Coalition for Environmentally Sustainable Artificial Intelligence (AI), aiming to ramp up global momentum to place AI on a more environmentally sustainable path. Spearheaded by France, the UN Environment Programme (UNEP) and the International Telecommunication Union (ITU), the Coalition brings together stakeholders across the AI value chain for dialogue and ambitious collaborative initiatives.
The Coalition was announced at the Artificial Intelligence (AI) Action Summit in Paris, where Heads of State and Government, leaders of international organisations, CEOs, academics, artists, and members of civil society gathered to discuss support for AI innovation, adequate regulation, and respect for rights to ensure development of these technologies in the interests of all, including developing countries.
The Coalition will encourage AI initiatives for the planet, including its role in decarbonising economies, reducing pollution, preserving biodiversity, protecting the oceans, and ensuring humanity operates within planetary boundaries. It will use a collaborative approach – bringing together governments, academia, civil society, and the private sector – to focus on standardised methods and metrics for measuring AI’s environmental impacts, comprehensive life cycle analysis frameworks for reporting and disclosure, and prioritisation of research on sustainable AI.
“We know that AI can be a force for climate action and energy efficiency. But we also know AI power-intensive systems are already placing an unsustainable strain on our planet,” UN Secretary-General, António Guterres, said in his remarks at the Summit. “So, it is crucial to design AI algorithms and infrastructures that consume less energy and integrate AI into smart grids to optimise power use.”
While AI may help tackle some of the world’s biggest environmental emergencies – for instance, it is being used to map the destructive dredging of sand and chart emissions of methane, a potent greenhouse gas – a growing body of research cautions that there is a negative side to the explosion of AI and its associated infrastructure, including the electronic waste produced – and high levels of electricity and water consumed – by the proliferating data centres that house AI servers produce electronic waste.
From data centres to training models, AI must run on sustainable energy that fuels a more sustainable future. The Coalition aims to build sustainable AI into the global discussion in much the same way AI security or AI ethics are studied.
“The AI Action Summit is a turning point: for the first time, the ecological transition has been at the core of the discussions in a international AI summit. I am very proud that France organised this first Forum for sustainable AI with 200 stakeholders present,” said Agnès Pannier-Runacher, France’s Minister of Ecological Transition, Energy, Climate and Risk Prevention. “Today, my Ministry, with ITU and UNEP, launched the Coalition for Sustainable AI – more than 90 members, including 37 companies, have joined this ambitious initiative on green AI and AI for green.”
More than 190 countries have adopted a series of non-binding recommendations on the ethical use of AI, which covers the environment. In addition, both the European Union and the United States of America have introduced legislation to temper the environmental impact of AI. However, the policy landscape remains sparse.
A number of recent initiatives are working to build the knowledge base around AI and the environment: through the National Institute for Research in Digital Science and Technology (INRIA) and the French Ministry of Ecological Transition, a community of 36 scientists, companies, public institutions and international organisations have published a position paper identifying the challenges that must be overcome to maximize the positive effects of AI systems while limiting their environmental impact.
The first hackathon combining AI and energy sobriety, the Frugal AI Challenge, also brought together more than 60 teams of data scientists around the unprecedented challenge of designing AI models related to environmental issues, including the detection of climate disinformation, analysis of regions at risk of fires, identification of illegal deforestation) while optimising their energy efficiency. In addition, the first international working group on the use of generative AI to provide access to environmental knowledge aims to develop a best practice guide for the effective and ethical use of generative AI to provide access to environmental documents, while a multi-stakeholder Green Digital Action initiative, convened by ITU, has launched a new thematic pillar on green computing with a dedicated Sustainable AI working group.
In 2024, UNEP released an issue note that explores AI’s environmental footprint and considers how the technology can be rolled out sustainably; it followed a major UNEP report, Navigating New Horizons, which also examined AI’s promise and perils. In 2025, UNEP will publish a guide to encourage public and private purchases towards energy-efficient data centres. These guidelines will be based on international best practices and established global standards (EU Code of Conduct, Energy Star, ISO/IEC 30134, etc.). They will inform investors, development banks and local authorities on the objective elements defining an energy-efficient data centre.
“The power of AI to solve complex global challenges is becoming ever clearer, but so too are its environmental impacts and the need for environmental guardrails to ensure the field grows sustainably,” said Golestan (Sally) Radwan, Chief Digital Officer for UNEP. “The new Coalition brings together critical stakeholders who have the power to work together and build systems that ensure the net effect of AI on the planet is positive as the technology continues to deploy rapidly.”
