In 2024, it was virtually impossible to attend a Procurement conference that didn’t host talks on Sustainability and on AI. You would have done very well indeed to find a roundtable discussion or procurement seminar where these two topics weren’t front-and-centre or an article where they weren't at least mentioned. What I haven't seen as much of are instances of the two topics being discussed together, which I find surprising given how entwined they are becoming.
I want to emphasise the importance of Procurement leaders being able to reconcile these two concepts as they continue to climb up their priority lists. AI has the potential to revolutionise how sustainable procurement strategies are devised and deployed, but there is a catch. A big one. The energy consumption of AI threatens to dampen progress made on sustainability and whether you’re a CPO or a Buyer, you need to ask yourself if you’re thinking enough about this contradiction…
“Global data centre electricity demand will have doubled from 2022 to 2026”
Artificial Intelligence (AI) is a catch-all term for a group of technologies that have the capability to process enormous amounts of data and information, as well as mimic ‘human’ thinking. The impact of AI, despite being a relatively novel development, has been vast. According to Boston Consulting Group, 87% of organisations believe AI will deliver competitive advantage, which means if your procurement team is yet to embrace AI, you can be assured it’s headed your way. The AI market is set to grow by 26% in 2025 and the International Energy Agency forecasts that global data centre electricity demand will have doubled from 2022 to 2026. Yes, doubled. In four years.
For Procurement leaders the possibilities are tantalising. Spend classification, guided buying, strategy creation, supplier analysis and contract drafting are all examples of procurement processes that can all be augmented by the use of AI tools. Let’s illustrate this with two potential use cases. We know that AI is excellent at pattern detection, with both similarities and anomalies. An agriculture business could deploy AI to recognise weather patterns, enabling them to reduce water usage and vastly cut expenditure on chemicals. This would have sustainability and bottom line benefits. Another attribute of AI is the ability it has to pour over historical knowledge and existing data in order to predict future outcomes. This can help procurement leaders to offset climate or geopolitical risk, or optimise their global supplier base, which ultimately improves the resilience and continuity of supply.
“…a single ChatGPT conversation uses 500ml of water”
The downsides of harnessing AI to support sustainable procurement goals predominantly centre on power consumption, water usage and the use of rare earth minerals. Existing data centres are increasingly becoming obsolete, they simply lack the infrastructure required to handle the power demands of today’s users as AI adoption skyrockets. This increased demand for new-build construction increases emissions and raw material usage, including rare earth minerals in the boards of the computers themselves. CBRE estimates that in the US, data centre inventory increased by almost 25% last year, a significant shift. These data centres also happen to use an enormous amount of water to cool the computers they house. It’s estimated that a single ChatGPT conversation uses 500ml of water. Given the scarcity of water in many regions, this is a considerable issue. Whilst on the topic of ChatGPT – for many of us the gateway to the world of AI – it uses 10x the energy of a Google search and consumes as much energy everyday as 18,000 homes in the US alone. Closer to home, the International Energy Agency estimates that in Ireland, AI could account for 29% of all energy usage by 2026.
If a Procurement leader is working to reduce emissions and climate impact with a robust sustainability strategy, it must include a cost/benefit analysis of using AI and the energy consumption that comes with it. The need for clarity is paramount, as is the need for honesty. Does AI need to be used for everything, or are some activities not worth it when examining them through the prism of excessive energy consumption?
Every usage of AI carries an energy cost and procurement leaders need ensure they’re not taking one step forwards, then two back, when it comes to delivering sustainability targets.
So what are you doing?
This article was written by Alex McDougall, Principle Consultant. Alex has 7 years' experience of placing procurement specialists all over the UK and Europe - to reach out to Alex regarding the article, a job search or any hiring needs, you can reach him at alex@pod-talent.com.