The Capgemini Research Institute has published a report titled “Developing Sustainable Gen AI,” which reveals the significant environmental impact of generative artificial intelligence (Generative AI). As this technology is adopted at an accelerated pace, many organizations are not adequately measuring their environmental footprint, putting their environmental, social and corporate governance (ESG) and sustainability goals at risk.
The adoption of Generative AI has grown rapidly. A previous Capgemini study showed that only 6% of companies had integrated this technology across all their functions by the end of 2023, a figure that increased to 24% by October 2024. Although Generative AI can drive business growth and improve Energy efficiency also requires processing large volumes of data and considerable computing power, which consumes large amounts of electricity, water and other resources.
Nearly half of executives surveyed (48%) believe that the use of Generative AI has increased their greenhouse gas (GHG) emissions. This carbon footprint is expected to continue growing, with companies currently measuring their environmental impact projecting an increase of 2.6% to 4.8% over the next two years. To mitigate this impact, many organizations are turning to renewable energy sources and optimizing their AI infrastructure.
Lack of Priority in Sustainability
Despite rapid innovations in Generative AI, companies’ sustainability reports have not kept pace with change. Only 12% of executives measure the environmental footprint of their use of Generative AI, and just 38% are aware of its environmental impact. Sustainability remains a marginal consideration in evaluating Generative AI models, with only a fifth of executives considering environmental footprint as one of the top five factors when selecting or building AI models.
With growing awareness of the environmental footprint of Generative AI, nearly a third of companies have begun to incorporate sustainability measures into the lifecycle of this technology. More than half are already using smaller models and powering their infrastructure with renewable energy sources, or plan to do so in the next 12 months. However, with more than three-quarters of organizations using pre-trained models and only 4% building their own models from scratch, executives rely heavily on their technology partners to address the environmental footprint of Generative AI.
Almost half of executives believe that the use of Gen AI has caused an increase in greenhouse gas (GHG) emissions, and 42% acknowledge that they have had to rethink their climate goals
Proposals for Sustainable AI
Capgemini’s report suggests that companies conduct a thorough assessment of both the financial profitability and environmental footprint of their Generative AI projects before implementing them. It is recommended to consider whether these energy-intensive technologies are really needed or whether more sustainable alternatives could be used. Additionally, sustainable practices must be applied throughout the AI lifecycle, including hardware, model architecture, data center power sources, and sustainable use policies.
Generative AI and Sustainability Goals
Despite its environmental footprint, Generative AI can be used to accelerate sustainability goals. The report highlights use cases such as ESG reporting and scenario planning, materials optimization for key industries, and sustainable product design. One-third of executives are already using Generative AI for sustainability initiatives, and two-thirds expect a more than 10% reduction in GHG emissions in the next 3-5 years as a result of these initiatives.
