Google has revealed that its greenhouse gas emissions have increased by 48% over the past five years, primarily due to the high computational power required by artificial intelligence (AI) systems. This surge in emissions is a significant setback to the company’s climate goals.
AI’s intensive computational needs have placed immense pressure on Google’s data centers globally. In its recent environmental report, Google acknowledged that reducing these emissions “may be challenging,” particularly as it continues to build new infrastructure to support its growing AI demands.
Earlier this year, Google announced a £788 million investment to establish a new data center in the UK, a move driven by the increasing demand for AI capabilities. This expansion comes as Google strives to meet its self-imposed target of achieving net zero emissions by the end of the decade.
The environmental impact of AI is a growing concern, with adoption rates continuing to rise. A study by the International Energy Agency predicts that the electricity consumption of data centers could double between 2022 and 2026, exacerbating the challenge.
Google’s data centers in Europe and the Americas predominantly use energy from carbon-free sources. However, this is not the case for sites in the Middle East, Asia, and Australia, where the proportion of energy from cleaner sources is significantly lower.
Google asserts that it is “actively working through” the “significant challenges” it faces in reducing emissions. The company admits that some initiatives to lower emissions might not yield immediate benefits. The report highlights that while progress has been made in advancing clean energy in many regions, there are still hard-to-decarbonize areas like the Asia-Pacific where carbon-free energy is not readily available. Additionally, the transition from investment to the construction of clean energy projects and the resulting reductions in greenhouse gas emissions often involves longer lead times.
4o