In this Masterclass hosted with the World Business Council on Sustainable Development, Professors Michael Wade and Julia Binder discuss the benefits of bringing together AI and sustainability thinking....
The World Economic Forum’s Global Risks Report 2025 identifies two forces that currently dominate business conversations about future risks: the accelerating urgency of the climate crisis and the race toward digital transformation driven by artificial intelligence. But what happens when these two megatrends collide?
This question framed IMD’s recent masterclass, “Greening AI: What every business leader needs to know,” hosted in partnership with the World Business Council on Sustainable Development (WBCSD). The session, led by Michael Wade and Julia Binder, unpacked how AI, a technology that comes with a substantial environmental cost, might also become one of the most powerful tools for driving sustainability.
“AI and sustainability are the two great transformations of our time, yet most companies treat them as entirely separate,” said Wade. Digital teams typically sit on one side of the organization with sustainability teams on another. The opportunity lies in bringing them together.
This integration begins by understanding AI’s environmental footprint. Training advanced models consumes staggering amounts of power and water. As these models become embedded in everyday operations, inference (i.e., using the models) is now starting to rival training in energy intensity. Last year, data centers consumed about 1.5% of the world’s electricity demand, a share set to nearly double by 2030, according to the International Energy Agency.
“The first step is to acknowledge that AI isn’t automatically good,” said Wade. “Its impact depends on how we deploy it.”
Introducing the presentation, the panel moderator, WBCSD’s Maarten Dirks asked participants what they feared most about AI’s footprint. The consensus was clear: energy consumption loomed largest ahead of other risks like “water,” “uncontrollable,” or “inefficiency.” A second poll revealed that most participants rated their organizations at the “intermediate” and “novice” stages of integrating AI with sustainability, a telling sign that awareness is high, but maturity remains low.
“You can’t reduce what you can’t see,” said Binder. “Most corporate emissions hide in Scope 3, those indirect emissions across sprawling supply chains.” Here, AI can act as a powerful flashlight.
She pointed to AI-enabled transparency tools that can map complex value chains, track materials, and spot hotspots invisible to traditional reporting. One striking example is the HolyGrail initiative, which embeds digital watermarks on packaging to enable recyclers to instantly identify materials and sort them for higher-value recycling rather than wasteful downcycling.
The food industry also offers some telling insights. When the ethical chocolate brand Tony’s Chocolonely used AI to analyze its “bean-to-bar” supply chain, it discovered that the largest contributor to its footprint wasn’t packaging or logistics, but dairy. That discovery led to a reformulation toward darker chocolate and milk substitutes. “AI helps you find the counterintuitive levers,” Wade noted. “That’s where the real progress happens.”
Once you can see impacts clearly, the next challenge is acting on them.
Once you can see impacts clearly, the next challenge is acting on them. Binder highlighted how AI already powers meaningful reductions in resource use, from optimizing shipping routes at companies like Maersk to enabling precision agriculture that targets water, fertilizer, and pesticides only where needed.
That principle extends beyond operations to behavior. “When people can see their energy use in real time, they change how they act,” Wade said, citing studies where real-time feedback from smart meters led households to cut consumption significantly. He compared it to the way wearables like the Oura ring change personal habits: “Visibility leads to action.”
The speakers also urged companies to dematerialize wherever possible, replacing atoms with bits. “Every time you substitute a physical process with a digital one, you’re likely cutting emissions,” said Wade. Even data centers, notorious for their energy use, can be part of the solution when their waste heat is redirected to warm municipal buildings or swimming pools, for instance.
If seeing and acting are the first two steps, scaling is the true test.
If seeing and acting are the first two steps, scaling is the true test. “We already have the tools to reduce emissions,” said Binder. “The question is: how do we make what works scale fast enough?”
She pointed to examples of AI-driven scaling: drones that use machine learning to restore forests by identifying optimal planting sites and seed mixes; digital platforms like Too Good To Go that match surplus food with local demand, now preventing millions of tons of waste globally; and AI applications that help balance renewable-heavy energy grids in real time.
For Wade, this system-level view is essential. “We tend to think of sustainability in silos – energy, transport, agriculture – but AI can coordinate across those boundaries,” he said. “That’s how small improvements turn into real transformation.”
Both Wade and Binder cautioned against what they called “AI solutionism. “Ask first: do we even need AI for this?” Wade urged. Simpler tools, traditional analytics, or organizational change can sometimes yield faster and cleaner results. “The greenest computation,” he said, “is the one you don’t run.”
Binder emphasized materiality, adding, “It’s easy to optimize the wrong thing because it looks good in a report. You can’t let the shiny pilot distract from where your biggest impacts lie.”
Both agreed that data quality is still the biggest bottleneck. “Everyone wants fancy models,” Wade said. “But the hard work is building clean, trustworthy data streams. Once you have that, the rest is easy.”
Finally, there’s the industry’s own reckoning. Despite ambitious climate pledges, many technology giants are seeing absolute emissions rise as AI scales. “The pressure to decarbonize won’t only come from investors or consumers,” Wade observed. “It will come from the sheer cost of energy and water.”
AI is a double-edged technology: energy-hungry and powerful in equal measure.
The professors closed the session with practical advice for the months ahead. Companies should start by mapping their true hotspots: not just where emissions are reported, but where they originate. They should then pick a small number of AI-powered interventions (route optimization, energy load management, or precision resource use, for example) and test them with clear baselines and measurable outcomes.
Most importantly, they should bring their digital and sustainability teams together, with shared KPIs that measure both performance and impact. “When these agendas finally converge,” said Binder, “we’ll unlock the real value of AI, not just for productivity, but for the planet.”
AI is a double-edged technology: energy-hungry and powerful in equal measure. The challenge for business isn’t whether to use it, but how to use it wisely. As Wade put it, “Technology doesn’t solve the sustainability problem by itself. But applied with purpose, it can accelerate solutions faster than anything else we have.”
To boost company performance, executives must support the competitiveness of their countries and embrace public-private collaboration to help articulate and implement national strategy, say IMD Professors Arturo Bris and Phil Rosenzweig....
YouTube
Rather than a long volume on one country or another, this book provides what The Wall Street Journal calls “a valuable and thoroughly researched analysis.” This selection will not only be highly...
YouTubeOur book for May is a very important book that has received considerable attention and praise, all fully justified. It speaks to business, of course – work design, product design – but...
YouTube
Explore first person business intelligence from top minds curated for a global executive audience
Share