Rebecca Scottorn, Partner at L.E.K. Consulting, explains why AI-driven data centre growth is changing how the energy sector thinks about grid capacity, resilience, and investment priorities.

AI-driven data centre growth is no longer a future pressure on the energy sector; it is already reshaping investment decisions, infrastructure priorities, and thinking around reliability. As power demand rises and project timelines compress, utilities, gas infrastructure operators, and equipment providers are being pushed to respond faster than traditional planning cycles have typically allowed.

To better understand how the sector is responding, we recently conducted a survey and held industry discussions with energy executives across O&G majors, refining and petrochemical companies, oilfield services and equipment providers, utilities, and players across the power and renewables value chain. Hearing directly from decision-makers working in the sector every day provides a useful view of the pressures they are seeing, the trade-offs they are making, and the strategies now taking shape.

All respondents had at least five years’ experience in the energy industry, as well as responsibility for strategic initiatives and investment decisions. Now in its seventh year, the study combines quantitative and qualitative insight into key energy trends, with a particular focus on North America and Europe. This year’s edition also examined two themes that are becoming increasingly difficult to separate from the wider energy conversation: the role of natural gas and the use of AI.

The findings point to a clear shift. While oil and gas producers remain disciplined amid global uncertainty, power utilities and pipeline operators are accelerating investment to keep pace with data centre-driven load growth and to modernise ageing networks. The scale and urgency of data centre demand is becoming one of the defining forces in energy investment.

So what does that mean in practice? Here are the key takeaways:

Data centres are reshaping energy investment priorities

The expansion of AI-driven data centres is reshaping energy infrastructure planning, capital allocation, and overall power market dynamics. As demand for technology continues to accelerate globally, utilities, gas infrastructure operators, and equipment providers are being forced to respond at a speed not seen before.

Against this backdrop, the study shows spending patterns diverging sharply. While oil and gas producers are remaining disciplined amid global uncertainty, power utilities and pipeline operators are accelerating spending to keep up with data centre-driven growth and modernise outdated grids. The scale and urgency of data centre demand is emerging as one of the defining forces in energy investment today.

Grid modernisation and interconnection delays are shaping data centre strategies

Grid modernisation is now central to the energy transition. Investment priorities are shifting from supply-side expansion to system-level resilience, with capital focused on assets and technologies that ensure reliability, create capacity for large-load customers, and integrate renewables without compromising stability.

Across North America and Europe, utilities recognise the unprecedented pressure AI-driven data centres are placing on electricity systems, representing one of the largest sources of incremental load growth. In response, many are increasing investment in infrastructure modernisation, including substations, grid lines, digitalisation, and smart meters, while also reinforcing networks against climate-driven stress.

E.ON, for example, is committing approximately $47 billion from 2024 to 2028, with around 81% going towards network expansion and digitalisation, and the remainder supporting generation and storage. This reflects the dual focus on meeting hyperscale demand and replacing ageing infrastructure.

Regional dynamics: Different pressures, same conclusion

The US faces the steepest load increases, propelled by AI-driven data centres and industrial reshoring, while European utilities face a different set of challenges. Demand growth is more moderate, but high renewable penetration and increasingly interconnected markets place sustained strain on networks. Companies such as E.ON are directing the majority of capital towards network expansion and digitalisation to manage congestion, integrate renewables, and maintain system stability.

Across regions, the objective is consistent. Whether driven by load growth or system complexity, utilities are investing to modernise networks, improve resilience, and unlock additional capacity, reinforcing the grid’s role as a critical enabler of reliable power systems.

Gas and behind-the-meter power emerge as a reliability backbone

The rise in data centre demand is intensifying pressure on already constrained grids. Utilities report multi-gigawatt clusters requiring near-continuous power and full redundancy, while interconnection queues and equipment bottlenecks continue to slow grid access. Survey responses point to a growing dependence on temporary behind-the-meter (BTM) solutions as a result.

This is consistent with broader market activity, which suggests that BTM use for data centres is set to more than double between 2026 and 2030 as operators seek reliable, controllable power sources. In the survey, 63% of O&G respondents cited a lack of confidence in the grid and said they were adopting BTM solutions to maintain project certainty.

This dynamic is creating a two-speed system: grid upgrades take years, while data centres scale in months. Against this backdrop, natural gas has emerged as a direct solution to data centre power needs, particularly through behind-the-meter generation. Behind-the-meter power allows operators to secure both primary and redundant supply, bypass interconnection bottlenecks, and stabilise operating costs.

Respondents expect behind-the-meter solutions to power close to half of all data centre megawatts by 2030, reflecting near-term necessity rather than long-term preference, as grid capacity gradually catches up. Companies view BTM generation as a short-term workaround of less than two years rather than a structural solution, with reinforced and modernised grid capacity remaining the long-term answer.

This strategy is not, however, without its hurdles. Gas turbine lead times have stretched to three to five years or longer, creating a significant bottleneck for both front-of-meter and behind-the-meter additions. Supply chain pressures, cost inflation, and equipment availability also risk slowing deployments just as demand peaks. As one respondent said, “Data centre players don’t want to wait – you do what you have to do.”

Storage, flexibility and renewables move from optional to essential

Looking ahead to 2026, data centre energy strategies are becoming more hybrid and sophisticated. While gas provides reliability, storage and other flexible, dispatchable assets are increasingly essential for smoothing loads, managing peak demand, and improving system efficiency.

Battery storage is increasingly being paired with on-site generation to provide instant backup, load shifting, and resilience during grid disturbances. At the same time, renewables, whether contracted through power purchase agreements or integrated locally, are becoming a structural part of data centre power portfolios, driven by both sustainability targets and long-term cost considerations.

Utilities such as E.ON have also highlighted the need to expand and digitise grids to integrate renewables, EVs, battery projects, and large new loads such as data centres simultaneously. This is pushing data centres towards multi-source power architectures that combine grid supply, gas, storage, and renewables rather than relying on a single solution.

AI is both driving and supporting industry demand

Ironically, the same technology driving data centre growth – AI – is also becoming a critical tool for managing the energy systems that support it.

AI is already being deployed across utilities and infrastructure operators for load forecasting, predictive maintenance, and performance optimisation. Advanced forecasting can help grid operators anticipate data centre-driven demand spikes, while predictive analytics can identify equipment failures before outages occur, directly improving resilience for mission-critical facilities.

For data centre operators, AI-enabled energy management systems can optimise fuel usage, dispatch storage more efficiently, and reduce operating costs while lowering emissions. These capabilities are becoming central to meeting uptime requirements alongside sustainability and cost pressures.

The future of power is hybrid

With all this in mind, data centre growth is accelerating faster than traditional power infrastructure can adapt. In response, operators are pursuing flexible, hybrid energy strategies, anchored by gas and behind-the-meter solutions today, but increasingly complemented by storage, renewables, and AI-driven optimisation.

As grid modernisation continues and new generation comes to fruition, reliance on temporary solutions may ease. But for now, data centres are pushing the energy sector into a new era of speed, flexibility, and innovation – one where reliability, resilience, and intelligence matter more than ever.