Artificial intelligence is opening new opportunities to improve our understanding of the global carbon cycle, from enhancing Earth system models to advancing satellite-based carbon monitoring and climate predictions. These opportunities were at the centre of the Workshop on Artificial Intelligence for Carbon, held on 1 July 2025 in Como, Italy, where leading researchers gathered to discuss the latest advances, identify research priorities, and shape a shared vision for the future of AI in carbon science.

Organised within the CMCC Foundation's Integration of the Planetary Biogeochemical and Industrial Carbon Cycle (ICC) Programme, the workshop explored the application of artificial intelligence across a wide range of topics, including Earth system modelling, climate prediction, carbon cycle monitoring, remote sensing, ecosystem dynamics, biodiversity, and climate policy. Through six thematic sessions, participants examined how AI can help improve carbon accounting, enhance climate projections, support environmental monitoring, and contribute to the development of effective climate mitigation and adaptation strategies.

Among the invited speakers was Amirpasha Mozaffari (Barcelona Supercomputing Center, BSC), who presented "Reconstruction of Historical Land Surface Boundary Conditions with Machine Learning." His presentation demonstrated how machine learning techniques can be applied to reconstruct historical land surface boundary conditions, providing valuable datasets for Earth system modelling and improving our understanding of terrestrial carbon cycle dynamics. The research closely aligns with CONCERTO's ambition to integrate advanced artificial intelligence, Earth observation and modelling approaches to reduce uncertainties in climate science.

The workshop also featured CONCERTO Project Coordinator Dr Manuela Balzarolo (CMCC Foundation) among its speakers, highlighting the strong scientific links between the event and the project's objectives. Together with experts from leading universities and research organisations worldwide, participants discussed emerging challenges, identified priorities for future research, and contributed to defining the building blocks of a position paper on the role of artificial intelligence in advancing carbon science.