Practices and approaches that minimise the environmental impact of developing and operating AI systems, particularly focusing on energy consumption and carbon emissions. Sustainable AI seeks to balance performance objectives with environmental considerations.

As AI models grow larger and more complex, their environmental footprint has become a significant concern. Training large models can consume substantial energy resources and generate considerable carbon emissions. Sustainable AI practices seek to address these impacts through more efficient algorithms, optimised infrastructure, renewable energy sourcing, and consideration of environmental trade-offs in model design decisions. Sustainability considerations are increasingly important as organisations face pressure to reduce their overall environmental impact.

‍

An organisation implementing green AI practices by selecting energy-efficient model architectures, scheduling training jobs during periods of abundant renewable energy availability, optimising inference to reduce computational requirements, and reporting on the carbon footprint of their AI operations.