Assessing the Climate Impact of Global AI development

Project ID: 2531bd1611

(You will need this ID for your application)

Research Theme: Artificial Intelligence and Robotics

Research Area(s): Engineering
Information and Communications Technologies

UCL Lead department: Bartlett School of Sustainable Construction

Department Website

Lead Supervisor: Zhifu Mi

Project Summary:

Artificial Intelligence (AI) is expanding rapidly, driving exponential increases in computational demand and electricity consumption. While AI has transformative potential across multiple sectors, its environmental implications remain poorly understood. In particular, the carbon footprint of AI training is escalating at a rate that threatens to undermine global climate mitigation goals, with emissions projected to reach gigaton scales by 2040 if left unmitigated. In addition to direct emissions, the social and economic costs of AI’s carbon emissions cannot be overlooked.

This project seeks to develop a comprehensive modelling framework to quantify the CO₂ emissions of AI training and assess pathways to align AI development with carbon neutrality targets. Specifically, the project objectives encompass:

  1. Quantifying historical and present-day AI-related CO₂ emissions to establish a monitor metric for AI’s climate impact, drawing on macro and micro level datasets.

  2. Developing global and regional emission projections for AI training up to 2050 under alternative scenarios, integrating computational demand growth, hardware efficiency trends, and electricity decarbonisation trajectories.

  3. Estimating the social costs AI-related emissions, using climate change integrated assessment models.

  4. Exploring mitigation strategies, including renewable energy adoption, institutional innovations, carbon pricing mechanisms, and algorithmic efficiency breakthroughs, to determine how AI can evolve without derailing climate goals.

The student will be based at The Bartlett School of Sustainable Construction, supervised by Prof. Zhifu Mi and Prof. Xi Liang.

We are seeking a highly motivated candidate with a master’s degree in computer science, environmental sciences, energy, economics, or other relevant disciplines. The candidate should have experience with at least one programming language such as Python, MATLAB, or R. Familiarity with any of the following areas would be advantageous but not essential: machine learning, carbon accounting, input-output analysis or climate change integrated assessment modelling.