
The Signal Insight
AI data centers’ insatiable hunger for energy is now widely understood, prompting unprecedented investments in power infrastructure by some of their biggest users. But Christophe Beck warns that we are ignoring their similar thirst for water. Addressing that challenge has “become an imperative” to sustaining the AI boom, he says.
Beck is chairman and CEO of Ecolab, an $80 billion water treatment, hygiene, and infection protection business whose customers range from Microsoft to Marriott. The century-old business was once called Economics Laboratory, and Beck is telling his fellow CEOs they should make water central to their strategy — not just to meet their sustainability goals, but for their bottom line.
There has been little evidence that boards are prioritizing water strategies, with just 5% of large global companies putting an internal shadow price on their water use, compared to the close to 25% that use an internal price for carbon emissions.
Ecolab has developed “direct-to-chip” technology to help cool microchips inside servers, which he says is more efficient than air conditioning entire data centers. But their growth, along with a boom in microchip manufacturing “fabs,” could create as much new demand for scarce water resources by 2030 as the entire drinking needs of the US today, Beck says — making “smart water management” a necessity and a financial opportunity.
This interview has been condensed and edited for clarity.
Andrew Edgecliffe-Johnson: What’s your message to CEOs who may not be thinking of water as central to their strategy?
Christophe Beck: I meet many of those. The core message is in two parts. The first one is that there is no life without water, there is no growth without water, there is no future without water. We’ll have to find ways to use and recycle water, as nature has always done. And the second message, which is the most important one, is that it’s going to help you make more money because you’re going to reduce your energy usage at the same time.
Especially in industrial setups, up to 75% of the power that’s being used is to manage water. It’s to heat water, to cool water, to transport water, or to treat water. So when you reuse and recycle water, you reuse and recycle energy, and you reuse and recycle carbon, and you reduce your cost at the same time.
Isn’t it going to be tempting for companies to ignore the water side of the AI challenge as they’re so focused on the energy aspects?
Totally, until it becomes a decision point. The moment that you want to build a data center, people really face the truth. And it turns really bad if you’re not ready for it.
What do you think the return is going to be like on these data centers, and do you think this boom is sustainable?
It’s going to be a very good return, but probably not as good as we hope right now. Most of those companies [building AI data centers] were software companies, they were not infrastructure companies. Today, they’re managing huge infrastructure. It’s a lot of capex, it’s a lot of things to bring together, it’s a lot of things that can go wrong. I mean, it’s all infrastructure stuff that they hated. [That dramatic] change from software to hardware doesn’t happen in a perfect way from the get-go. Water and energy usage was not ideal over the last five years, [but] it’s becoming much better now. So there will be an optimization phase, but the demand is going to keep growing dramatically.
What does AI mean for your own IT budget? Are you spending radically more on technology?
We are, and I’m convinced it’s the right thing to do. There are probably two [types of] companies: the ones who are saying, “My God, this is a train that has left the station. I’d better get on that train as quickly as I can, because if I miss it, I’m never going to catch it.” And you have a big group saying, “No, it’s not that dramatic. I’m going to take my time. I’m going to see what others are doing, and I’m going to get on the train when the time is right.” I’m afraid that [latter] group will not like where it ends up, because that train will be so far ahead that it’s going to be really hard to catch.
How do you see the climate discussion changing in companies?
The naming is changing dramatically, there’s no doubt. It’s way harder to talk about climate. It’s way harder to talk about sustainability. [But] for too long, companies focused on the sustainability component of the climate [agenda], and forgot value creation, growth, and making money. For us, it has always been a means to an end. The less water, the less energy you use to do something, the more money you’re going to make, the faster you can grow, the better your story is going to be. Some companies kind of got confused [about] what their true role is and how they’re supposed to be delivering value, and that’s getting a bit corrected now.
What’s happening with data centers, with microelectronics manufacturing, and the fact that AI will require so much water and so much power in the years to come, is going to help [us rise to] the climate challenge, because it’s become a must. We have no choice. We’re building a data center every week, a fab every month. They both use a huge amount of water. We need to find solutions to use less water in order to make it work.

Notable
- A bill awaiting the signature of New Jersey Gov. Phil Murphy would require data center operators to publicly disclose each quarter how much water and electricity they consume.
- Inside Ecolab, Beck is applying AI to its data to optimize the operations of its customers around the world, from steel mills to restaurants.