Measuring What Matters: How to Assess AI’s Environmental Impact

The report Measuring What Matters: How to Assess AI’s Environmental Impact offers a comprehensive overview of current approaches to evaluating the environmental impacts of AI systems. The review focuses on identifying which components of AI’s environmental impacts are being measured, evaluating the transparency and methodology soundness of these measuring practices, and determining their relevance and actionability. Synthesizing findings from academic studies, corporate sustainability initiatives, and emerging environmental tracking technologies, the report examines measurement methodologies, identifies current limitations, and offers recommendations for key stakeholder groups: developers (producers), users (consumers), and policy-makers.

One of the most pressing issues uncovered is the widespread reliance on indirect estimates when assessing energy consumption during the training phase of AI models. These estimates often lack real-time, empirical measurement. Furthermore, equally important lifecycle stages remain significantly underexplored. This reliance on proxies introduces substantial data gaps, impedes accountability, and restricts consumers’ ability to make informed, sustainable choices about AI.