At the recent Consumer Goods Forum Global Summit in Vienna, IBM hosted a panel discussion exploring how retail and consumer goods firms can establish governance frameworks to ensure AI is deployed in ways that are ethical, transparent, accountable and aligned with business objectives.
In addition, a separate recent IBM report, IT Sustainability at a Crossroads: Choosing a Future of Responsible Computing, highlighted the challenges in balancing the environmental benefits of digital transformation with the growing energy demands of AI and other next generation technologies, suggesting that a holistic approach to sustainable IT will become increasingly important as businesses continue to scale AI adoption.
‘Generative AI is a game changer for more sustainable IT,’ it noted. ‘How it changes the game, though, is up to every organisation that uses it.’
Responsible AI
At the Vienna summit, SustainabilityOnline caught up with Mark Kardos, who leads IBM’s Sustainability Consulting Centre of Competence, to discuss the business case for AI as a tool to deliver both sustainability gains and cost efficiencies.
We started by talking about what ‘responsible AI’ looks like from an IBM perspective, and how businesses can align themselves with that thinking.
“Security and safety are obviously a big part of that,” he says. “Responsibility around data is really, really important, particularly when organisations are thinking about what they’re leveraging in terms of frontier models. A lot of cloud-based technology also brings significant security considerations.
“But a lot of it really does come down to application and use, and trying to think about what those long-term impacts are, both within an organisation and for the wider communities they serve.”
Developmental approach
While integrating AI into a business’s governance strategy can seem daunting – prompting many firms to ‘wait and see’ – Kardos believes that firms should adopt a developmental approach, with regular reviews that allow businesses to assess outcomes, identify risks and adapt governance as AI develops.
“Many firms think about it in terms of ‘let’s come up with a really comprehensive strategy as to how this will help us be responsible’,” he says. “That’s just too vague, and almost impossible to implement in that way. So, it’s more of an iterative process rather than an explicit step back and pause.”
As a recent IBM study found, 84% of consumer products and retail executives say AI will significantly enhance their ability to respond rapidly to market disruptions and evolving customer needs. Sustainability is a big part of that, Kardos notes.
“There are almost no conversations that I have seen around AI that don’t have a sustainability component – whether its clients asking about energy usage, the resilience of data centres, or how they optimise token counts and things like that. It’s so directly impactful to both the results they’re able to achieve, as well as, obviously, their environmental footprint.”
Moving beyond reporting
Where AI has been able to move the needle the most when it comes to sustainability is in assisting organisations with moving beyond reporting and compliance, he adds.
“Even in the early days of generative AI, we were quick to ask how we could leverage it to automate and drive insights around the reporting process.
“Obviously, a number of years ago, CSRD created huge pressure on businesses across much of the world, so that was a really clear use case. Now, we’re coming back to the idea of value, and where it’s driving that value. I think it’s been a huge unlock for a number of organisations.”
AI also enables firms to analyse supplier data at a level that was previously impractical, providing greater transparency and supporting supplier-specific interactions. Unlike other technologies, such as blockchain, that require all participants to adopt common systems, AI can act as a “translation engine”, enabling parties to exchange information in a format that’s relevant to their needs.
“That transfers to all sorts of different areas, whether you’re talking about supply chain applications, communications, or even manufacturer-to-customer interactions,” says Kardos. “Take auto manufacturers, for example. Even those segments of the market that aren’t necessarily sustainability-minded are still ultimately being driven towards EVs and more sustainable decisions simply because, once you understand driving patterns, it’s more cost-effective for them. Having AI as that translation engine, which is able to say, ‘Let’s speak your language’, has been huge.”
Improving performance
With more than 90% of most businesses’ environmental impact located in their supply chains, Kardos believes that AI can address longstanding challenges around data collection and supplier engagement, in turn enabling firms to focus on improving performance.
“The lack of transparency, driven by the need for collaboration and communication, has always been a massive challenge,” he adds. “And, let’s face it, sustainability functions are often underfunded and understaffed. So I think this is an opportunity to have tools at your fingertips that allow you to drive initiatives while taking care of much of the data chasing that tends to consume a huge chunk of the year.
“Now it’s a case of saying, ‘We’ve got the data, we have a way to process it, and now we can focus on performance’.”
IBM and NVIDIA recently announced a collaboration to help organisations move AI from pilot to production – learn more here.
