News in the Channel - issue #37

AI SUSTAINABILITY

Contributors

emissions. “So, businesses have a responsibility to manage them,” he adds. “This was always true of businesses using cloud software, but AI has dramatically increased the energy use involved compared with standard cloud computing. For businesses, this creates risk and risk that is difficult to predict with any level of accuracy.”

you track compute emissions: deploy AI only where there’s a proven productivity or efficiency return,” he says. Businesses must take a long-term, architecture-first approach, he adds. “The most essential consideration is matching innovation with environmental management and long-term planning,” he says. “Businesses can reduce AI’s impact through energy-optimised infrastructure, such as platforms that avoid unnecessary data duplication and support modular scaling of compute and storage. Green software practices, zero-copy architectures and flexible scaling help reduce waste across the AI pipeline. “Traditional scaling often means adding more hardware, more power and more cost. Sustainable scalability is about growing intelligently using energy-efficient systems, optimising resource utilisation, and supporting future workloads without constant forklift upgrades. It’s about aligning IT growth with environmental and financial responsibility.” Adam says the starting point must be intentionality. “ESG and AI don’t have to be in conflict, but AI cannot be deployed indiscriminately,” he warns. “Businesses need to be clear on where AI genuinely adds value and where it simply adds complexity. From an ESG perspective, that means favouring targeted, use- case-driven AI over broad, experimental deployments, and being transparent about energy usage, data sources and outcomes. Responsible governance matters as much as the technology itself.” Improving efficiency Richard says businesses can help to reconcile ESG worries by using AI to optimise their own operations for sustainability. “For instance, AI can model the energy impact of different configurations in data centres, helping firms make informed decisions that reduce carbon output,” he says.

Jason Beckett

hitachivantara.com

Adam Herbert

go-data.com

What’s often underestimated is how quickly these demands scale when AI is deployed without clear constraints or purpose. “ ”

Squaring with ESG goals With AI use increasing, and the carbon produced by it falling under scope 3 emissions, this poses a potential problem for businesses that want to use AI but are also mindful of their sustainability targets. Simon says businesses square AI use with their ESG goals with great difficulty. “This is a nascent technology, so businesses can only make educated guesses about future energy demands for AI data processing,” he says. “ESG goals set before these assumptions are made are likely to need to be significantly adjusted. Water use is also a large-scale impact area that responsible businesses should be considering in their (excuse the pun) upstream processes.” AI should be treated like any other capital asset, Jason adds. “Make sure you set usage budgets and get on top of how

Richard Eglon

nebulaglobalservices.com

23 CONTINUED