Responsible AI can help science tackle the planetary crises

Responsible uses of AI for sustainability research and governance is an opportunity the world cannot afford to miss, the report concludes.

Systematic analysis of 8,500 studies

The report draws on a systematic review of 8,500 scientific publications and deep expert analyses across eight sustainability themes, including preparing for future shocks, understanding climate change and other complex Earth systems, prospering urban areas, stewarding oceans, and securing freshwater for all. It shows how AI can uncover hidden patterns in environmental data, model risks more accurately, and support better collective decisions.

Drew Purves, Nature Lead at Google DeepMind, who contributed to two of the report’s chapters, notes: “AI has enormous potential to advance sustainability science. Realizing that potential requires deep partnerships across disciplines and across sectors.”

The report notes that AI’s environmental footprint, potential data biases, and unequal access to computational power needs attention. Addressing these challenges, they argue, requires governance frameworks, interdisciplinary collaboration, and ethical standards that put people and planet first.

“AI will not solve sustainability challenges on its own,” says Maria Schewenius, co-editor and PhD candidate at the Centre and the Beijer Institute of Ecological Economics. “But it can help us better understand complex systems, anticipate risks, and support decisions that safeguard both people and our living planet.”

Co-author Jonathan Donges, deputy-lead of the Earth Resilience Science Unit at PIK, highlights AI’s potential to illuminate the planet’s hidden patterns: “Earth is a deeply interconnected system. AI offers powerful tools to help detect feedbacks, thresholds, and tipping points that were previously invisible to science. That knowledge can help us deal with deep uncertainties and act in time to prevent irreversible harm.”

Transparency, inclusivity, and fair distribution

Elizabeth Tellman, Assistant Professor at the University of Wisconsin–Madison, adds a call for equity and inclusion in shaping AI’s role in sustainability: “We must ensure that AI systems serve the communities most affected by climate and environmental change. Equity, inclusivity, and shared governance must be at the heart of AI for sustainability.”

Complementing this perspective, Samuel T. Segun, Senior Researcher at the Global Center on AI Governance, Canada, underscores the role of governance in ensuring AI benefits are fairly distributed: “AI offers a powerful opportunity to bridge science and policy particularly in safeguarding the environment. However, if we want it to truly serve people and the planet, we must ensure that its benefits are equitably distributed and that governance frameworks prevent misuse.”

The report concludes that, responsible uses of AI for sustainability research are possible, and an opportunity the world cannot afford to miss. Breakthroughs in sustainability research driven by AI are not only within reach, they can also be essential for our collective future. The responsibility now lies with sustainability researchers, companies, technology entrepreneurs, funders, and decision-makers to shape AI development alongside human insight, guiding society toward better decisions for a safer, fairer, and more resilient planet.

Explore the full report AI for a Planet Under Pressure »

Watch a recording of the launch event below