Artificial Intelligence (AI) is a rapidly developing technology with a widespread impact. Its implementation can be seen all around us from chatbots to autonomous vehicles. AI developments are on course to improve operational efficiency, bring labour productivity, and boost economies. However, the negative environmental impacts of AI are often hidden from public view and the technology can have socioeconomic and sustainability implications and hinder the clean energy ambitions of tech giants. Google and Microsoft, both have reported a significant surge in their carbon emissions in the last five years, largely attributed to increase in their data centre energy consumption as a consequence of infusing AI into their core products.
AI has been used for decades but the question around its impact on the environment has arisen in recent years since the techniques or AI models developed now use a vast amount of data, significant computational power and are becoming far more complex and widespread. This was triggered by the generative AI boom and mass adoption of large language model-based products like ChatGPT, DALL-E, and others in the second half of 2022 . These recent AI models are fed with diverse and large data sets to bring more accuracy and robustness, and work with low latency. This requires extensive storage and powerful processing, thus escalating electricity demand and carbon emissions. Microsoft, for example, had a stable increase of carbon emissions for many years but has reported a significant increase of 29% in carbon emissions since 2020, mostly due to its data centre expansions that were designed to support AI workloads. Another tech giant, Google, has also been facing challenges due to increased electricity demand driven by AI. It saw a rise of 48% in its carbon emissions since 2019, driven by data centres and supply chain emissions. This is clearly a challenge to the ambitions of any AI company that is aiming to becoming carbon neutral.
In this report, we summarise our findings on the environmental impact of AI and we highlight some of the initiatives that are being taken by AI companies and cloud providers to offset related carbon emissions.
