Managing the sustainable AI paradox: Harnessing AI while hitting sustainability goals

This report looks at the sustainability challenges that suppliers and tech user organisations face when deploying AI, given the technology’s energy-intensive nature. It explores a range of potential responses to the “sustainable AI paradox”, including data centre design and operations, energy-efficient coding, and right-sized model selection; as well as making the case for AI lifecycle environmental assessments that can identify areas where sustainability could be improved (and provide insight into unintended consequences).

Ways to access this research