
The rise of artificial intelligence (AI) is transforming our societies, but its environmental impact raises major questions. How can we reconcile technological innovation and the preservation of natural resources? It is to address this issue that the Forum for Sustainable AI was held, a structuring event aimed at guiding the development of AI towards a more eco-responsible trajectory.
During this Forum, several major announcements were made to coordinate international efforts and promote a minimalist and beneficial AI for ecological transition.
Launch of the Coalition for Sustainable AI

The Ministry of Ecological Transition, in partnership with the United Nations Environment Programme (UNEP) and the International Telecommunication Union (ITU), has officially launched the Coalition for Sustainable AI.
🔹 This initiative brings together governments, companies, and research institutions around the following goals:
- Improvement of the energy and ecological efficiency of AI technologies
- Development of standards to measure environmental impact
- Promotion of responsible innovation and sharing of best practices
Among the first members of this Coalition are IBM, Hugging Face, Mistral, Nvidia, INRIA, Climate Change AI, OVHCloud, and many others.
🔹 Discover the coalition: https://www.sustainableaicoalition.org/
An international roadmap to harmonize the assessment of AI’s environmental impact
The Forum also provided an opportunity to present a global roadmap aimed at establishing a transparent and harmonized assessment of the ecological footprint of AI.
🔹 The objectives of this roadmap:
- Harmonization of standardization efforts at the international level
- Development of robust indicators based on Life Cycle Assessment (LCA)
- Integration of best practices in eco-responsible design
- Strengthening cooperation between governments, businesses, and researchers
Led by the Ecolab of the General Secretariat for Sustainable Development, this initiative was designed in collaboration with experts from ISO, ITU, and IEEE, in partnership with OECD and UNESCO.
The Global Observatory on AI and Energy
A new tool to deepen the analysis and management of AI’s energy consumption
The International Energy Agency (IEA) announced the creation of the first global Observatory on AI and Energy.
🔹 Its missions:
- Centralization of data on AI’s energy consumption
- Continuous updating of trends and developments in the sector
- Provision of decision-making tools for governments and industry
- Assessment of challenges and opportunities for a more resource-efficient AI
This observatory aims to become a reference for better regulation and energy optimization of AI infrastructures.
The Frugal AI Challenge: a hackathon to promote a minimalist AI
The Forum also hosted the Frugal AI Challenge, an innovative hackathon co-organized by Hugging Face, Data for Good, and the Ministry of Ecological Transition.
🔹 The challenges addressed by the participants:
- Prediction of fire outbreaks – Development of analysis tools to identify at-risk areas
- Climate misinformation detection – Identification of environmental fake news
- Monitoring illegal deforestation – Acoustic analysis to spot destructive activities in forests
A distinctive feature of this hackathon: the models developed were assessed both on their accuracy and their energy consumption, thus highlighting the importance of digital sobriety.
An unprecedented public debate on AI and ecology at Sorbonne
One of the central questions of the Forum was: Does AI constitute an opportunity or a risk for ecological transition?
To answer this, 500 participants attended a Tribunal for Future Generations, a debate organized by several public institutions in partnership with Usbek & Rica.
🔹 The themes discussed:
- The real energy impact of AI
- The potential of AI to accelerate ecological transition
- The necessary regulatory and framework mechanisms
The discussions provided essential insights and fed reflections on the ecological issues of AI.
Towards a more responsible and sustainable AI
These announcements mark a major step forward in structuring a more resource-respecting AI truly committed to ecological transition.
The Coalition for Sustainable AI and the new standardization and observation initiatives represent significant advancements for better measuring, framing, and optimizing the environmental impact of AI.
Initiatives like the Frugal AI Challenge or the public debate at Sorbonne illustrate the capacity of innovation and collective dialogue to foster the development of more resource-efficient and effective models.
Can AI be a climate ally? The decisions made today will shape the response of tomorrow, with the involvement of public, private, and academic stakeholders.