- The Capgemini report finds that nearly half of executives say rising greenhouse gases are due to artificial intelligence
- Only 12% measure their company’s AI carbon footprint
- Limited reporting makes it difficult to measure
Businesses are becoming increasingly aware of the environmental impacts of artificial intelligence when it comes to the power and natural resources required to run data centers, new research has found.
A surge in AI interest and use has put AI data centers in the firing line in recent years – their hunger for large amounts of electricity as well as water for cooling has left critics questioning the effectiveness of the developing technology compared to its environmental costs .
As a result, research by Capgemini revealed that nearly half (48%) of executives surveyed attributed increasing greenhouse gas emissions to their generative AI projects.
Companies know that GenAI is bad for the environment
Although companies are increasingly aware of the environmental impacts of AI, quantifying them is more of a challenge. Only 12% of organizations measure their GenAI carbon footprint, and sustainability ranks low when it comes to influencing a company’s decision to choose a model.
There is also a notable reliance on third-party suppliers, reducing the amount of control companies have over their emissions. More than three in four use pre-trained models compared to just 4% who build their own.
Some organizations are looking to use smaller models to reduce their environmental impacts and others are exploring renewable energy options to further reduce emissions, but the key challenge is that sustainability is currently a low priority for management – only one in five see it as a key factor.
Model providers also do not disclose data regarding their environmental impact, with bundled datasets making it difficult to identify certain impacts.
In addition to addressing sustainability positioning as a decision factor, Capgemini says companies should consider using smaller, task-specific models to lower energy consumption.
Companies can also improve the sustainability of their infrastructure by choosing efficient hardware and green data centers.
The report also calls for further governance of the ethical and sustainable use of generative AI.
“If we want Gen AI to be a force for sustainable business value, there needs to be a market discussion around data collaboration that creates industry standards around how we account for AI’s environmental footprint, so business leaders are equipped to make more informed, responsible business decisions and mitigate these impacts,” noted Capgemini Head of Global Sustainability Services and Corporate Responsibility, Cyril Garcia.
