Strategic Intelligence: Artificial Intelligence for Climate Mitigation

Strategic Intelligence: Artificial Intelligence for Climate Mitigation

Summary

Artificial intelligence (AI) is a tool for autonomous decision-making that can enable efficiencies to reduce carbon emissions and ecosystem degradation. Some tech companies have claimed that AI will help fix the climate, but this would require AI-driven efficiencies to outweigh the environmental harm caused by AI development and use.

Key Highlights

• Predictive AI can help improve efficiency and reduce resource use across key sectors. AI can improve sustainability outcomes when used in specific ways to reduce emissions or pressure on ecosystems. Predictive AI is often the most appropriate form of AI for these specific purposes, such as renewable grid optimization, sustainable agriculture, building energy management, and supply chain optimization.
• Generative AI produces significant emissions, and there is little evidence that it mitigates climate change. The emissions and water use associated with data centers, as well as AI inference and training, mean that generative AI's environmental burden often cancels out any ESG benefits it creates. Claims about generative AI's climate benefits are also frequently exaggerated or unsupported by evidence.

Scope

• This report focuses on the relationship between AI and climate mitigation. It analyzes six specific AI use cases that can drive climate mitigation: renewable energy optimization, sustainable agriculture, building energy efficiency, supply chain optimization, and emissions monitoring.
• It also includes specific recommendations for chief sustainability officers who are incorporating AI into their company's decarbonization strategy.

Reasons to Buy

• Climate change is one of the gravest threats humanity has faced. Ever since AI entered the mainstream consciousness, there has been speculation around whether AI can help mitigate climate change.
• There are applications of AI that contribute meaningfully to climate mitigation, including renewable grid optimization, sustainable agriculture, and supply chain optimization. However, AI causes significant environmental harm. To successfully support climate mitigation, an AI application must generate efficiencies that outweigh the environmental harm it causes. In many cases, this is linked to the type of AI system used: predictive, generative, or agentic.