At COP31, AI can help communities adapt to climate change—if professionals know its limits

Urgency meets new technology

Turkish President Recep Tayyip Erdoğan has called COP31, hosted in Turkey, the “implementation COP,” stressing that climate promises must become action. A United Nations Environment Programme report published in September said limiting warming to 1.5 C, or only slightly above it, is no longer achievable.

A later return below 1.5 C remains possible, but would require rapid, substantial cuts in emissions. Continued warming raises the threat of severe and irreversible impacts, adding urgency to efforts to understand climate risks and respond effectively. AI’s possible role is attracting attention, including through COP31’s Antalya AI Pledge.

Information is not a decision

In Canada, AI is used to improve weather forecasts, while researchers apply it to flood-risk assessment. New uses include infrastructure and adaptation planning, such as flood modelling in Vancouver. Yet additional data alone does not ensure sound choices.

A coastal community could use AI modelling to identify roads, homes and public facilities at risk from flooding. Choosing among shoreline reinforcement, wetland restoration, development limits or relocation also means weighing livelihoods, ecosystems and ties to place, as well as competing values, Indigenous rights and jurisdiction, and whose knowledge and authority count.

Skills for responsible adaptation

The Climate Action Competency Framework developed at Royal Roads University’s ResilienceByDesign Lab identifies systems thinking, critical information assessment, collaboration, communication and ethical leadership as important skills. AI adds further needs: professionals must understand how systems work, spot possible bias and data gaps, judge recommendations and consider transparency and accountability.