PGGM’s Post

⤵️ My thoughts regarding responsible AI have just been published on PGGM's website. For those interested, find the link below!

I had a conversation with a brilliant environmentalist today that gave me pause... as a tech guy. 😅 We're all captivated by what AI can build, automate, and discover. But we spend far less time talking about what it consumes. Every model, every inference, every breakthrough relies on very real resources: -Energy -Water -Critical minerals -Massive computing infrastructure. AI isn't intangible; it's powered by the physical world. The question isn't whether we should advance AI. We absolutely should. The question is whether we'll build it responsibly. If sustainability isn't embedded into AI governance, infrastructure, and policy from the outset, we risk creating a future that's technologically extraordinary but environmentally impoverished, a world of remarkable intelligence built on depleted ecosystems. "A technologically advanced wasteland" as she put it. Innovation and sustainability shouldn't compete. They should reinforce one another. The greatness and legacy of AI is threatened solely by weather or not we ensure that progress doesn't come at the expense of the planet that makes it possible.

What’s your perspective – does AI ultimately help or hinder sustainability? The debate surrounding AI and sustainability is intensifying. On one side, AI requires substantial computing resources, energy, and infrastructure, raising valid concerns about its environmental impact as adoption increases. However, we might be focusing on the wrong question. Instead of merely asking, "Does AI consume energy?" we should consider, "What waste could AI help us eliminate?" For instance: - Reducing unnecessary rework through improved analysis and decision support. - Eliminating low-value manual processes and duplication. - Assisting organisation’s in monitoring emissions and identifying opportunities to reduce their environmental footprint. - Optimising energy consumption across transport, buildings, and supply chains. - Decreasing time spent in meetings by surfacing information automatically. - Enabling colleagues to dedicate more time to problem-solving rather than searching for information. For decades, the focus has been on optimising physical resources. The next challenge may involve optimising another finite resource: human effort. The most sustainable organisations of the future won't just be those that emit less carbon; they will be those that waste less time, energy, talent, and opportunity. AI isn't inherently good or bad for sustainability. The crucial question is whether we can use it to create more value than it consumes. This is the conversation we should be having.

AI is becoming one of the most powerful forces shaping the future, but it is also raising an important sustainability question that we can no longer ignore. What stood out to me while thinking about data centers and AI infrastructure is how much energy, water, and material use sits behind the systems we rely on every day. The conversation is no longer only about what AI can do. It is also about what it consumes in order to function. That makes the environmental footprint of AI a very real design and infrastructure issue, not just a technology issue. What I find most relevant is that sustainability at this level is not about stopping progress. It is about asking whether progress is being built responsibly. If data centers continue to grow at this pace, then energy efficiency, water use, clean power, and lifecycle thinking will need to become part of the conversation from the beginning. As someone who thinks a lot about sustainable design, this makes me believe that every system we build, whether physical or digital, should be shaped with long-term impact in mind. How can AI continue to grow while being designed, powered, and cooled in a way that respects planetary limits?

AI operates not only on code but also on significant amounts of electricity and water. As utility demands increase, the costs associated with running AI are expected to rise dramatically. The future will favor those who can create highly efficient, low-waste technological ecosystems. Efficiency has evolved beyond being merely a sustainability objective; it is now a critical metric for business survival.

While AI could do great things for scientific research, it's primarily being used for significantly less important purposes — and consuming enormous amounts of energy in the process. Hear Prof. Sami Kara FRSN FFCIRP FACATECH and Prof. Michael Hauschild discuss the importance of building a circular economy, and how technology could help us reach environmental sustainability instead of hinder us: https://lnkd.in/gS-V_TRm

AI is changing the way industries operate, including how we think about sustainability. We believe innovation comes with responsibility. As new technologies evolve, it's important to consider both the opportunities they create and the challenges they present. The future of sustainability will be shaped by the choices we make today.

I really enjoyed making this video because it reflects how I see the future we have the opportunity to build with AI. For decades, sustainability has often arrived at the end of the process, trying to mitigate the damage that had already been done. Today, we have the chance to change that. We can become part of the design process itself: deciding how we build products, how we use energy to support these systems, how we integrate nature into our decisions, and even how we consume. AI is an irreversible technology, much like oil once was. We are creating a world where almost everything will depend on it. That's exactly why I believe sustainability must be embedded from the very beginning, not added later. If we design AI with life cycle thinking, circular economy principles, and respect for natural resources from day one, we have a real opportunity to build something better. This is the vision I share in this video. I hope you'll watch it, and I'd love to hear whether you share the same vision. 🎥 Digital Product Passport: What Every Business Needs Know

3/4 of sustainability professionals don't know how to measure AI's environmental impact. If folk working in the sustainability space are not getting to grips with the data, what hope do business leaders more broadly have to understand the impact of the tools taking over many of our daily tasks and functions? The truth is, the tech supply chain remains hard to get to grips with, and there are no standard models yet around energy use per prompt or workflow. But we can't let the complexity stop us from moving. If we want to understand what responsible AI adoption looks like, here is where we need to focus: 👉 Progress over perfection: You don't need a flawless internal framework to begin. Start by establishing who your suppliers are, how you use the tools, and what your data gaps are. 👉 Demand transparency: You can't manage what you can't measure. Use digital procurement as a lever - ask suppliers about their Power Usage Effectiveness (PUE), water consumption, and whether their operations run on time-matched renewable energy. 👉 Be curious beyond carbon: A true assessment should look at the entire value chain, from the water used to cool data centres to societal externalities like labour displacement and biased algorithms. Responsible AI adoption is as much a human and sustainability challenge as it is a technology one. Let’s stop waiting for perfect reporting frameworks and start treating AI impact as a business decision that requires robust, responsible governance. ♻️ Follow me Polly Milne for more on responsible AI adoption through an ESG lens.

Every organisation chasing AI at scale is quietly making a sustainability trade-off — the question is whether they're making it consciously. Global data centre electricity demand is set to more than double by 2030. Companies are now signing nuclear energy deals just to keep up with what AI needs. Two groups are exposed, in opposite directions. Industries already carbon-heavy — aviation, manufacturing, construction, shipping, fast fashion — are now running more AI on top of emissions they haven't fixed yet. AI adoption without an energy strategy just compounds an existing liability. Industries built on a sustainability identity — ethical apparel, renewable energy, green banking, plant-based food — face the opposite risk: staying clean by staying analog. If competitors use AI to cut waste and optimise at scale, a sustainability-first brand that under-invests in AI could lose ground on efficiency, even while winning on principle. Here's what actually keeps this in balance, and it isn't a values debate. It's a reporting gap. Right now, the energy and water AI uses gets folded into a company's overall ESG numbers, which is exactly why it's easy to hide or dilute. If a company is using AI, that usage needs its own line in ESG reporting, not buried inside the rest of the business. No formal standard for this exists yet. That's exactly why it needs to become one. The same logic applies in reverse to sustainability-first companies. Stepping back from AI isn't the responsible choice, it's just a slower way to lose ground. What they need is a clear internal standard for when AI use is worth its footprint, and what compensates for it when it's used. The real prioritisation for 2026 isn't AI vs sustainability. It's: 1. Report AI-specific energy and water use separately, not blended into overall ESG numbers 2. Tie new AI infrastructure to new clean energy, not existing renewable supply 3. Set a clear standard for when AI use is justified, and what compensates for it 4. Treat sustainability credentials as a reason to adopt AI carefully, not a reason to abstain The AI build-out won't stay this intense forever. Whether sustainability survives it depends on whether companies report this honestly now, not on the hype slowing down.