When AI becomes critical infrastructure: a leadership responsibility in the energy sector

Artificial intelligence is no longer a topic for the future in the energy sector. AI is fast becoming a fundamental premise: not as a single tool, but as a control layer that increasingly shapes how the energy system operates, develops and is regulated.

While the debate still often revolves around specific applications and potential, the reality in Danish energy and utility companies is already changing. AI is increasingly being used for real-time control, forecasting and optimisation, while new large electricity consumers such as data centres and Power-to-X are putting pressure on the system’s capacity and raising demands for flexibility and speed.

‘AI has moved from being a digital experiment to being a prerequisite for handling the complexity of the energy system. It’s no longer a question of whether you want to work with AI, but of how you design your business for it,’ says Mikkel Eriksen, Senior Partner at Eve.

What is certain is that AI already matters a great deal to the energy sector and will matter even more. But how companies lead and organise themselves in a reality where AI becomes a prerequisite for running the business will be decisive for their future success.

Many recognise AI’s potential, but fewer are building their business on it

Today there is broad agreement that AI will be of decisive importance to the energy sector. Even so, many companies find that the road from ambition to real value is harder than expected.

‘AI features prominently in companies’ strategies and on the leadership agenda, but in practice it is often treated as a stand-alone initiative. That’s where we see the biggest challenge: the ambition is high, but the organisation isn’t designed to deliver on it,’ explains Mikkel Eriksen.

According to him, this is not down to a lack of will, but to the fact that AI is still thought of as an add-on to the existing business, rather than something that challenges and fundamentally changes the way people work and make decisions.

‘The difference in the years ahead won’t be between those who have tested AI and those who haven’t. It will be between those who have set up their business to use AI systematically and those who still regard it as an experiment,’ adds Jim Nielsen, Managing Partner for Data & AI at Eve Denmark.

"AI features prominently in companies’ strategies and on the leadership agenda, but in practice it is often treated as a stand-alone initiative. That’s where we see the biggest challenge: the ambition is high, but the organisation isn’t designed to deliver on it."

Mikkel Eriksen, Senior Partner
Mikkel Eriksen, Senior Partner

The technology is becoming routine, but can leadership keep up?

The technology itself is developing rapidly and becoming ever more widely accessible. This also means that AI alone will not necessarily be the decisive competitive factor.

‘The technology is increasingly becoming standard. What will set companies apart is not whether they use AI, but whether they manage to design the agentic organisation and thereby strike the right balance between people and machines, between human-led decisions and autonomous actions,’ says Jim Nielsen.

He points out that many companies will underestimate how much change this really requires, and that AI challenges not only the IT landscape, but also the way organisations work, prioritise and follow up. If this is not addressed at leadership level, AI efforts risk ending up as isolated solutions with no real impact.

An energy system that can no longer be managed in the same way

The question is highly topical in light of developments in the Danish energy system. More decentralised energy sources, more flexible consumption and new large electricity consumers mean that the number of possible system states is growing significantly.

‘The complexity of the energy system is starting to exceed the level at which traditional planning and manual control are sufficient. That’s where AI goes from being smart to being necessary,’ says Anders Nørgaard, Senior Partner at Eve, who has advised several Danish energy companies through mergers and major transformations for more than 15 years.

AI can contribute on many fronts, from better forecasting and maintenance to more dynamic grid and portfolio management. But according to several observers, the next leap is no longer about analysis, but about control.

‘The big change comes when AI doesn’t just support decisions but becomes directly involved in control, in real time and across systems. That changes both operations and employees’ roles,’ explains Anders Nørgaard.

Why do so many AI initiatives get stuck in the pilot phase?

Despite the clear pressure on the energy system, many AI programmes remain in an experimental phase. According to Jim Nielsen, this is often because AI is still treated as a classic IT project, where many start with a narrowly defined use case and a business case. This contrasts with the need for a far more long-term approach in the energy sector, in line with the way grid expansion and other critical infrastructure have traditionally been planned.

Perceptions of risk also play a part.

‘Leadership teams are often very aware of the risk of getting something wrong with AI. But far fewer discuss the risk of doing nothing, and of ending up with a system that is too complex to manage with the tools they have today,’ says Mikkel Eriksen.

The complexity of the energy system is starting to exceed the level at which traditional planning and manual control are sufficient. That’s where AI goes from being smart to being necessary

Anders Nørgaard, Senior Partner, Eve

Regulation, accountability and a distinctly Danish strength

As a heavily regulated and safety-critical sector, the energy sector faces particular requirements when adopting AI. For Jim Nielsen, regulation is not about changing the direction of how AI can be used in the sector, but about setting clear requirements for the framework. In Denmark, where the sector has a strong tradition of trust and accountability, AI can become a strength if the technology is built in transparently and in a controlled way, in terms of both resilience and trust.

At the same time, new regulations such as NIS2 and the AI Act make accountability and governance more complex to navigate. When AI is used in critical systems, incidents can trigger several parallel requirements for reporting and follow-up. This reinforces the need for clear decision-making structures and human oversight, even in a reality where AI increasingly operates autonomously.

According to Jim Nielsen, this underlines that regulation is not just a compliance issue, but a leadership responsibility. As AI moves from analysis to actual control, it becomes crucial to have clearly assigned leadership responsibility, clear governance and human oversight, including in situations where the systems in practice operate autonomously.

He adds:

‘The question is no longer whether AI can be used, but how to do it responsibly. Companies that take a position on governance and accountability early on will be in a stronger position in the long run.’

2035: when AI has become part of everyday operations

Looking 10 to 15 years ahead, many of the processes currently described as AI initiatives will probably be regarded as ordinary operations.

‘In the future, we won’t talk about “AI projects” in the same way. AI will simply be an integral part of how the energy system is managed, just as ERP systems today form the backbone of financial management in many companies,’ says Anders Nørgaard.

For leadership teams, this means the focus shifts from technology to the collaboration between people and systems. Lise Rasmussen adds that what will matter is not whether AI works, but how people and organisations work with it. That calls for new skills, new roles and a different way of thinking about leadership.

The competitive advantage will not be technology alone, but leadership maturity

Denmark is well placed. We have an advanced energy system, strong data, a high level of digital trust and a business community used to operating within regulated frameworks. But it is no longer enough to be interested in technology. Turning AI into real value requires leadership maturity.

From a leadership perspective, AI is therefore no longer about identifying the next use case, but about addressing a more fundamental question: how the energy company should operate in a reality where decisions are increasingly supported by data and AI.

‘The biggest risk today is not experimenting with AI. The biggest risk is underestimating how far-reaching the consequences will be,’ says Anders Nørgaard.

He points out that many organisations are in fact facing a choice:

‘Do you build your business for an energy system in which AI is an integral part of how it is controlled, or do you try to handle the complexity of the future with yesterday’s organisational logic? Ultimately, that is a leadership choice.’

Talk to our authors