Estimating the growth in emissions from AI data centres

2026-08-14

AI Summary

This study addresses the surging carbon emissions driven by the expansion of agent AI-powered data centers. Leveraging authoritative scenarios and life cycle assessment models, this work rigorously quantifies the dynamic emission trajectories of both operational and embodied carbon in AI data center operations. The findings confirm that industry expansion precipitates a dramatic increase in total emissions, with embodied carbon constituting a significantly rising proportion. By filling the critical gap in full-life-cycle carbon footprint quantification for the agent AI era, this research unveils previously overlooked embodied carbon risks. Consequently, it provides essential scientific evidence to inform precise low-carbon transition strategies for AI infrastructure development.

Abstract

The advent of agentic AI is driving an unprecedented growth in data centre expansion. Using state-of-the-art life cycle assessment models for operational and embodied emissions from AI servers, we estimate the ensuing growth in overall CO2 emissions of AI data centres. The main contribution of this work is a rigorous quantification of emissions arising from projected AI data centre expansion using scenarios by the IEA and McKinsey. We show that the scenarios promoted by the AI industry would result in a dramatic rise in overall emissions and that the embodied carbon component is considerable.

Problem

• AI data centres
• CO2 emissions
• embodied carbon
• life cycle assessment
• agentic AI

Innovation

• Life Cycle Assessment
• Embodied Carbon
• AI Data Centres
• Emissions Quantification
• Agentic AI