KPMG sees role for AI in sustainability

A KPMG report says that AI, despite its large energy burden, can still be a positive tool for corporate sustainability efforts. 

The report, "AI for the Chief Sustainability Officer: Understanding the Intersection of AI and Sustainability," notes there are many ways entities can use AI to reduce their environmental impact and advance their sustainability goals. 

AI-driven analytics, for example, can help a company gain deeper insights into their carbon footprint as well as identify inefficiencies for target emission reduction measures. It could also be used to optimize energy and water consumption in buildings and industrial processes, as well as supply chain logistics, via analysis of real-time use patterns. The report also explains that AI can be used for sustainability reporting, which often draws on many different data sources, both financial and nonfinancial. KPMG noted that AI can be an innovation tool that can assist in designing sustainable products and services, as well as forecast extreme weather events and analyze historical and real-time market data to predict future trends. 

KPMG noted that it is using AI for these purposes itself. For clients, the firm uses AI to identify its most impactful decarbonization pathways for target reduction, offers AI-guided solutions to accelerate reporting and compliance with sustainability standards, provide optimized AI tools that can reduce manual efforts within the sustainability data management and reporting process, as well as offer ongoing guidance on emerging AI technologies. 

And for itself, the firm said it is actively working to integrate AI and sustainability into its larger environmental strategy. It is currently exploring the development of AI tools that will help enhance its sustainability professionals' efficiency and accuracy. Beyond that, it's also working with international teams to assess the impact of their own AI use, especially on data centers they own, as well as within the context of Scope 2 emissions. KPMG is working with its key technology partners to understand the impact of AI use outside its direct control. The firm sees sustainability as a core component of its trusted AI framework. 

Despite these measures, there is the matter of AI being highly energy intensive. For instance, in Google's most recent environment report, it revealed that its emissions have increased 13% from last year and 48% from their 2019 target, which the tech company mainly attributed to a rise in data center energy consumption and supply chain emissions, which it said was at least partially due to AI. The company conceded that as it further integrates AI into its products, reducing emissions may be challenging due to increasing energy demands from the greater intensity of AI computing, and the emissions associated with the expected increases in its technical infrastructure investment. For example, another estimate says that one query to ChatGPT uses approximately as much electricity as lighting one lightbulb for about 20 minutes. The KPMG report acknowledged this can be a challenge but is hopeful that technological advances can address the issue. 

"The computational power required for AI can lead to significant resource use and an increase in emissions, potentially offsetting sustainability gains," said the report. "However, recent advancements in energy-efficient AI technologies and renewable infrastructure are promising in reducing energy consumption, carbon emissions and water usage. As the AI landscape continues to rapidly evolve in cost and energy efficiencies, companies may focus on emissions from owned data centers and cloud computing providers, in order to create a clear path to decarbonize." 

Tegan Keele, KPMG US climate data and tech leader, who co-authored the report, said in an email that while AI does consume a lot of energy, it's not the whole story when it comes to emissions. 

"While companies should be mindful of AI's energy footprint, focusing on AI computing alone won't move the needle on emissions. We need to look holistically at overall Scope 2 consumption and value chain impacts," said Keele. 

Maura Hodge, KPMG US's sustainability leader and another of the report's authors, added that KPMG's own efforts to help clients reduce their carbon footprint, in turn, can be useful in creating a net environmental benefit for AI solutions. 

"This is why at KPMG, we're actively working to maximize AI's immense potential to help drive decarbonization, while simultaneously mitigating the impacts of its energy and water consumption. It's about finding a way to strike the balance, where AI ultimately delivers net positive environmental impact," said Hodge. "We recommend that companies work closely with their technology partners to understand the full impact of their AI usage and development, especially for operations outside their direct control."