The Laboratory on AI, Climate and Society
The Laboratory on AI, Climate and Society at Lamont-Doherty Earth Observatory, Columbia University is an interdisciplinary research group focusing on the responsible use of artificial intelligence in climate science, environmental research, sustainability, and its impacts on our society.
The lab brings together scholars, students, NGO affiliates, practitioners, and collaborators working across climate, justice, data science, artificial intelligence, ethics, governance, and transparency. Our work is grounded in the concept that AI can support scientific discovery and decision-making but only when it is developed with transparency, accountability, and care for the communities and ecosystems most affected by climate change.
We develop applications via model and database development, data tools, and governance frameworks as well as critical theories that are scientifically rigorous, openly documented, and designed for equitable outcomes.
We welcome anyone interested in contributing to our work, regardless of professional background or knowledge. It is indeed the diversity of the people who contribute to our group that makes the Laboratory a place to discover, discuss, build, and promote a more equitable future in which AI is a tool that addresses social injustice and other forms of digital and non-digital discrimination.
We focus on AI systems, theory, research practices, and applications that are not only technically powerful, but also transparent, auditable, inclusive, and socially accountable. This means asking not only what AI can do, but also how it is built, whose knowledge it reflects, what assumptions it carries, and who may benefit or be harmed by its use.
Through research, collaboration, public engagement, and training, the lab supports a more responsible AI ecosystem for climate science and society.
Our Mission
The mission of the Laboratory on AI, Climate and Society is to explore how artificial intelligence can be applied as a tool for responsible climate and Earth science.
Current Research Focus
The lab works across several interconnected areas of research and practice. Below, you will find some of the research areas we are currently exploring. Please feel free to reach out for collaboration or to work on new research themes. We want everyone to lead their own projects and succeed!
AI for Earth and Climate Science
We develop and apply machine learning methods to support climate and Earth system research. This includes work on climate system components, geophysical data, environmental datasets, and computational tools that can help scientists analyze complex systems. We have focused most of our attention on the polar regions, but our tools can be broadly extended to other fields, and we would love to hear from you.
One specific area of focus includes developing neural network emulators for components of the climate system. These tools can help approximate complex model behavior, reduce computational burdens, and create new ways to study Earth system processes.
Climate Justice and Environmental Data
We examine how AI can be used to understand climate impacts, infrastructure systems, environmental risk, and justice concerns. This includes developing and using environmental justice datasets and addressing questions about how climate burdens are distributed across communities.
AI systems trained on incomplete, biased, or poorly documented data can produce flawed or inequitable results. For this reason, the lab approaches environmental data not only as a technical challenge, but also as a social and ethical one.
AI Infrastructure, Energy, and Society
Artificial intelligence depends on physical infrastructure, including data centers, energy systems, minerals, water, cooling systems, and large-scale computational networks. We study how AI infrastructure intersects with climate change, energy demand, environmental justice, and public policy.
This work includes research on the social and environmental implications of AI infrastructure and the growing energy demands associated with large-scale computation.
Responsible AI Governance
We develop approaches for building AI systems that can be interrogated, audited, and trusted. This includes attention to documentation, transparency, model limitations, uncertainty, data provenance, and governance frameworks.
Responsible AI in climate science requires more than technical accuracy. It requires clear processes for accountability, public communication, ethical review, and interdisciplinary collaboration.
AI, Feminism, Equity, and Ethics
We also examine how digital systems interact with feminism, climate equity, ethics, and communities most affected by both algorithmic and environmental harm.
This work asks how AI systems can reproduce existing inequalities, whose perspectives are excluded from technical design, and how climate and AI research can better account for lived experience, community knowledge, and justice-centered approaches.
Methods and Tools
The Laboratory of AI, Climate and Society works across technical, social, and governance methods.
Our work may include:
- machine learning for climate and Earth system data;
- neural network emulators for climate system components;
- environmental justice datasets and mapping tools;
- database development and documentation;
- AI-assisted analysis of complex climate and infrastructure systems;
- research on data centers, energy systems, and AI infrastructure;
- governance frameworks for responsible AI;
- interdisciplinary methods connecting climate science, social science, and ethics.
The lab emphasizes reproducibility, documentation, and transparency. We aim to build tools and methods that can be evaluated, improved, and used responsibly by researchers, institutions, and collaborators.
Responsible AI Principles
We aim at developing, supporting and discussing AI systems that are:
Transparent
AI tools should be documented clearly, including their data sources, assumptions, limitations, and intended uses.
Auditable
Models and systems should be open to review, critique, testing, and improvement.
Scientifically rigorous
AI methods should support strong scientific reasoning, uncertainty analysis, and reproducible research practices.
Socially accountable
AI systems should be evaluated not only by technical performance, but also by their social, ethical, and environmental impacts.
Equity-centered
The benefits and risks of AI should be considered across different communities, especially those most affected by climate change, environmental harm, and algorithmic bias.
Interdisciplinary
Responsible AI for climate and Earth science requires collaboration across science, engineering, social science, humanities, law, policy, and community knowledge.
Collaboration
The Laboratory on AI, Climate and Society welcomes collaborators from any discipline who share a commitment to rigorous, responsible science and artificial intelligence.
We are especially interested in collaborations that connect AI with:
- climate modeling and Earth system science;
- geophysical data and environmental sensing;
- climate justice and environmental justice;
- data centers and energy infrastructure;
- responsible AI governance;
- AI ethics and public policy;
- interdisciplinary climate research;
- education, training, and public engagement.
Researchers, students, institutions, and community partners interested in collaboration are encouraged to reach out.
Training and Public Engagement
The lab contributes to training and public engagement around responsible AI and climate science. Through events, workshops, symposia, and collaborative research activities, the lab creates spaces for researchers, students, and practitioners to learn about AI methods, ethical risks, governance challenges, and emerging applications in Earth and climate science.
Past and ongoing activities include AI training at Lamont, public conversations on AI, climate, and society, and interdisciplinary programming that brings together scientific, technical, policy, and justice-centered perspectives.
Diversity, Equity, Inclusion, and Anti-Racism
The Laboratory on AI, Climate and Society is committed to building a research community that is diverse, inclusive, anti-racist, and anti-casteist in its practices and culture.
This commitment means going beyond representation alone. It requires examining the structures, norms, and power dynamics that have historically excluded communities of color, women, Indigenous peoples, people from caste-oppressed communities, and other underrepresented groups from scientific leadership and technological decision-making.
Our commitments span research, mentorship, collaboration, authorship, data practices, and public engagement. We aim to ensure that our datasets, case studies, models, and research questions reflect the needs and realities of communities most affected by climate change and environmental harm.
We understand this work as ongoing. We are committed to creating a research environment where harm can be named and addressed, where mentorship and authorship opportunities are distributed equitably, and where collaboration is grounded in respect, accountability, and care.
Contact
For collaborations, inquiries, or participation in the Laboratory on AI, Climate and Society, please contact:
Marco Tedesco
Founder & Director, Laboratory on AI, Climate and Society
Lamont-Doherty Earth Observatory
Columbia Climate School, Columbia University
Email: [email protected]
David Sathuluri
Founder & Co-Director, Laboratory on AI, Climate and Society
Lamont-Doherty Earth Observatory
Columbia Climate School, Columbia University
Email: [email protected]