Some 48% of business executives believe that the use of Gen AI has led to an increase in greenhouse gas emissions, a new study by Capgemini has found.
Capgemini’s study, Developing sustainable Gen AI, which surveyed 2,000 retail executives from 15 countries in North America, Europe and Asia Pacific, found that adoption of Gen AI is accelerating rapidly – as of October 2024, 24% of firms had integrated Gen AI into their processes, up from 6% at the end of 2023.
However, the computing power required to fuel Gen AI capabilities has led businesses to consider optimising their AI structure, or turning to more renewable energy sources, Capgemini noted.
Gen AI emissions
Businesses that currently measure their Gen-AI related footprint expect the share of emissions generated by the technology to increase from an average of 2.6% to an average of 4.8% over the next two years.
Currently, just 12% of the firms surveyed track the environmental footprint of Gen AI, while just under two fifths (38%) are ‘generally aware’ of the technology’s environmental impact.
While a fifth of executives surveyed consider the environmental footprint of Gen AI to be an important factor when selecting what model to use, more than half acknowledge that environmental impact could be reduced if sustainability were a central criterion in selecting and developing Gen AI tools.
In addition, close to one-third (31%) of organisations have begun integrating sustainability measures into their Gen AI lifecycle, such as using smaller models and renewable energy sources.
‘Immense potential’
“Gen AI, when used correctly, has immense potential – it can help companies find solutions that go far beyond human imagination,” commented Dr. Dorothea Pohlmann, global chief technology officer, sustainability, at Capgemini Engineering. “This applies to innovations in the field of materials research, more sustainable products or even more efficient manufacturing processes. Almost all branches of industry see this potential and hope for acceleration and cost reduction to the same extent.
“In order to ensure the sustainable use of Gen AI, companies are faced with the challenge of being able to calculate and thus minimise the ecological footprint of this technology. This is particularly relevant in view of being able to meet ambitious sustainability goals.” Read more here.
