A study conducted by the London School of Economics and Systemiq indicates that artificial intelligence (AI) has the potential to dramatically reduce global carbon emissions, projecting cuts of 3.2-5.4 billion tonnes of greenhouse gases annually by 2035. This significant reduction is expected to far exceed the carbon footprint generated by AI itself. Titled ‘Green and intelligent: the role of AI in the climate transition,’ the research posits that AI can facilitate a transition to a sustainable and inclusive economy, reframing the shift to net-zero as an opportunity for innovation rather than solely a cost. The study focused on power generation, meat and dairy production, and passenger vehicles, concluding that AI-driven savings in these sectors alone would more than compensate for the emissions from AI data centers.
The study identifies five primary mechanisms through which AI can advance climate action: enhancing the intelligence of complex systems like renewable energy grids, accelerating discovery processes and minimizing waste (e.g., in new material development), empowering individuals to make more environmentally conscious choices (e.g., through optimized routes or smart homes), predicting climate changes and assessing policy impacts, and safeguarding populations during extreme weather events via early warning systems. The researchers underscore the necessity of an “active state” where governments proactively incentivize green AI research, regulate its environmental impact, and invest in infrastructure to ensure the equitable and sustainable deployment of AI for maximum climate benefits.
Reference: https://www.artificialintelligence-news.com/news/study-finds-ai-slash-global-carbon-emissions/
