Search...

Type above and press Enter to search. Press Esc to cancel.

November 16, 2026 | 8 Mins Read

The Word Is Prudent 

November 16, 2026 | 8 Mins Read

The Word Is Prudent 

Share

By Berend Booms | Associate Editor | Future of Assets

On September 17, the New York Public Service Commission started work on an energy affordability index, which would link part of what utility executives earn to how affordable their customers’ bills are. It also ordered the state’s major electric, gas and water utilities to file, within sixty days, a written inventory of each artificial intelligence system running in their operations, along with whether a person can override it. 

The message about what regulators now expect from utilities is clear: spend carefully, and make sure you are ready to explain every decision behind that spending, whoever, or whatever, made it. It also brings to mind some of the predictions I made at the start of this year. For instance, I predicted that asset management would be pulled into enterprise risk, and that the conversation around AI would move from accuracy to accountability, specifically for regulated sectors such as energy and utilities. While most of the shifts I presented in my predictions article are set to play out over several years, the rapidly approaching end of this calendar year feels like a good opportunity to check some of my homework. 

Checking My Homework 

Some of the predictions held up better than I expected. For example, I predicted that efficiency by default would give way to resilience by design. I briefly covered the April 2025 blackout in Spain and Portugal last year, and in March this year, ENTSO-E published its final report on the outage. The expert panel found no single cause, pointing instead to many interconnected contributing factors. There’s one detail I did not have at the time of writing my initial LinkedIn post, which I think is crucial: at the critical moment, only about 58 percent of Spain’s shunt reactor capacity, the equipment that helps keep voltage in check, was connected. The rest could not be switched in on time, because doing it manually took longer than the grid could wait.  

Not all of the predictions aged as well. I wrote that the disclosure requirements around sustainability were tightening, at least across Europe. But then, in February, the European Union adopted its Omnibus package, which narrowed the Corporate Sustainability Reporting Directive to companies with more than one thousand employees and €450 million in turnover. Still, I feel the underlying constraint remains. The IFS Global Utility Survey 2026 reflects this: around a quarter of executives surveyed (26 percent) name sustainability tracking and carbon footprint management across their asset portfolio among the most important asset lifecycle management capabilities to invest in over the next two to three years.  

Where accountability is concerned, the direction of the prediction was right, but the route wrong. I had assumed it would imminently arrive in the form of AI regulations; in Europe, the obligations for high-risk AI in critical infrastructure have since been pushed back to December 2027. Accountability is still playing a major role; just think of the oldest question in all of utility regulation: can you justify what you spent? 

Competing for the Same Capital 

In order to justify what you spend, you have to spend sensibly: in the US, a utility that wants to recover an investment from its customers has to show regulators that the spending was prudent, or reasonable, given the knowledge the utility had when making the decision. The money that’s at stake keeps growing: the Edison Electric Institute (EEI) projects that US investor-owned electric utilities will spend around $239 billion this year, and around $1.4 trillion in total between 2026 and 2030. 

Looking at the latest IFS Utilities survey, it becomes clear how incredibly crowded the agenda behind that spending has become. The survey asked 850 senior executives at utilities across the United States, the United Kingdom, France, the Nordics, the Middle East and Australia about their most critical operational challenges. In first place, 27 percent of respondents named grid reliability and resilience against extreme weather events and cyber threats. Right behind reliability and resilience sit balancing capital between maintenance and expansion, which was identified as a challenge by 25 percent of respondents, and managing aging infrastructure while meeting new capacity demands, raised by 24 percent of executives surveyed. Three more challenges also land at 24 percent, including keeping control of costs to meet customer bill expectations and rate case requirements. 

More interesting than the order these challenges appear in is their closeness. There is not a single challenge that pulls clear of the rest, which also means every single one of these is competing for the same capital. 

Showing Your Work 

There are numerous valuable insights coming out of the survey. When asked which emerging technology will have the greatest impact on their operations, 24 percent of respondents picked integrated infrastructure planning models to demonstrate prudent, auditable investment decisions; if we filter down to just US respondents, that figure rises to 36 percent. “Prudent” and “auditable” are the words that stand out to me: they describe a decision that has to survive questioning by someone outside the company. 

It’s not a one-off mention either, as the same thread runs through the investment priorities: the priority most often cited as receiving the greatest investment attention is looking at the asset portfolio to establish where to prioritize investment and modernize existing grid infrastructure, at 26 percent, while 24 percent point to using asset investment planning tools to develop objective and defensible investment plans.  

Still, a defensible plan stands or falls on having the right operational pieces supporting the right strategy. About one third of the executives rated their asset lifecycle management as “advanced,” and just over half say their asset management systems are in place but operate in silos. Moreover, fewer than three in ten describe their data across operational systems such as SCADA, EAM, ERP and GIS as fully integrated. Asked what would most improve data-driven decision-making, 19 percent of the fully integrated group chose “incorporating a single decision model tool to calibrate risk against capital investment,” compared with around one in ten of everyone else. 

If we follow the responses of the most mature organizations, we soon find out where that leads. Looking at those rating themselves “advanced,” the capability they most often name as most important to invest in over the next two to three years is supply chain optimization and inventory management for critical spare parts, at 38 percent. The utilities that rate themselves most mature want the right part on the right shelf. 

Bottom of the List 

When I read a report, I’m always intrigued to see what ends up at the bottom of the list. An answer that ranks lower than you expected can say as much as the one at the top. Two questions piqued my interest this time.  

The first asked the respondents what primarily prevents them from adopting or expanding modern asset lifecycle management solutions, and “a lack of organizational alignment and change management capability” came last, picked by 8 percent of respondents. Having spent half a decade working in change management, I’ve come to believe that accountability lives in the organization. You can have a fantastic planning model explaining the trade-off between an upgrade and a replacement in as much detail as you like; deciding who signs off on that trade-off, and who has to answer for it when a regulator asks, is work for people. 

When asked how the success of digital transformation is measured, “reduction in infrastructure investment risk” was the least common answer at 13 percent, followed closely by “return on investment (ROI) and payback period” at 15 percent. For an industry under immense pressure to justify its spending, those stand out to me as surprisingly low numbers. 

Who Will Be Answering 

There’s one more tension I would like to address, because it forces me to confront another prediction I made at the start of the year. The survey puts workforce shortages and skills gaps, especially in data science and technical roles, near the bottom of the operational challenges at 22 percent (only scaling pilot projects ranks lower), and places field workforce enablement at the bottom of the list for asset lifecycle investment areas at 21 percent. Yet when asked what would most improve data-driven decision-making, “building internal capabilities and data literacy across the organization” shows up at the top of the list.  

Back to my prediction: I thought workforce risk would soon eclipse technology risk, but I framed this mostly as a retirement problem. My reading of the risk has always been that the older people get, the more likely they are to retire, and the more likely it is that tribal knowledge walks out the door. The survey doesn’t measure age or tenure, so it can’t settle that question. My own view has shifted, though: more than anything, I think we are dealing with a knowledge capture problem first and a knowledge transfer challenge second. Looking at my home country of the Netherlands, it becomes obvious how big that challenge really is going to be: grid operators and their contractors estimate they will need up to 30,000 additional technicians by 2030. 

That has huge implications for accountability. Making a choice between keeping an aging transformer in service for another year and freeing up budget for new capacity will increasingly be left to someone who has not seen many bad days on the network. Not only that, we are also asking them to defend their choices later. A regulator can and will ask for the paperwork, but knowing which risk is acceptable in the first place is learned through experience.  

Which brings me back to New York. By the middle of November, every major utility in the state will have filed its inventory, and for every AI system within its operations, it will have to be made clear whether a person can override it. It’s the right question to ask, and Spain shows why the answer is only half the story. The equipment that could have held the voltage was there; the grid just couldn’t wait for it. Checking for overrides on AI systems is great, but an override only protects anyone if the person holding it knows when to use it, and has the time to act. As the IFS survey shows, the most mature utilities are making sure the right part sits on the right shelf. Someone still has to know when the time is right to reach for it.

The survey numbers tell part of the story. To hear how utility leaders are acting on them, watch the webinar recording of "The New Shift in Utility Transformation."
Watch on demand