The Net Climate Impact of Artificial Intelligence (AI): Balancing Current Costs with Future Climate Benefits

Abstract: This chapter asks a deceptively simple but unresolved question: what is the net impact of artificial intelligence (AI) on climate change? Research has largely focused on the environmental footprint of AI, with less attention paid to its potential climate benefits or trade-offs between the two. This chapter develops a framework to evaluate both the greenhouse gas (GHG) emissions generated by AI systems and the emissions reductions that AI may enable across the economy, explicitly accounting for AI's potential to speed economic growth, delays in realizing the climate benefits of AI, and rebound effects. The framework offers a methodology for evaluating AI projects holistically rather than through emissions accounting alone. 

We examine the rapidly rising energy demands of AI and data centres, which increasingly threaten corporate net zero commitments, alongside AI's potential to support emissions reductions through applications such as energy system optimisation, demand response, grid management and accelerated electrification. We also note that AI is used extensively by the fossil fuel industry to find and produce hydrocarbons more cheaply, slowing the transition to a low-carbon economy. Although most current AI applications are emissions-intensive, we account for potential hardware and algorithmic efficiency gains, and targeted deployment in energy-and climate-critical systems that could deliver climate benefits for many use cases. However, absent complementary climate policies, efficiency gains enabled by AI are likely to trigger direct and indirect rebound effects that erode or reverse emissions savings. 

Critically, AI emissions are growing today, while AI-enabled emissions reductions will take time, creating an AI "carbon debt" that accelerates warming. Even if that debt is eventually repaid by AI climate solutions, the additional warming and climate damage will persist for decades to millennia: carbon neutrality does not imply climate neutrality. We conclude with policy recommendations to stimulate AI's climate benefits while limiting its environmental harms.