How the data centre boom proves we must rewire global energy systems

The AI revolution is reshaping the very foundations of our energy systems. As hyperscale data centres proliferate to power machine learning and generative AI, they are driving an unprecedented surge in electricity demand. Our recent research found that global data centre power demand will hit 700 TWh in 2025, exceeding that of EVs. By 2050, data centres could consume 3,500 TWh, equivalent to current power demand from India and the Middle East combined. This acceleration is both a technological milestone and a wake-up call for energy infrastructure worldwide. Success requires system-wide flexibility to manage both supply and demand through integrated solutions.

At the heart of this transformation lies a paradox: the digital economy, built on the promise of efficiency and intelligence, is placing immense strain on legacy energy systems that were never designed for such complexity. The rise of AI and rapid growth in power-hungry data center facilities is reshaping the evolving energy landscape and placing a strain on grid connections.

Grid resilience under strain

Data centres are among the most energy-intensive facilities on the planet. The biggest challenge for the power industry is predicting the future scale of data centre electricity demand. Unlike traditional industrial loads, they require constant, high-quality power 24/7, 365 days a year. And they’re not alone. The electrification of transport, heating, and manufacturing, combined with the rapid deployment of renewables and flexible loads, is creating a dynamic and unpredictable demand landscape. Legacy grids, built for linear and predictable consumption patterns, are struggling to keep up. The result is a growing risk of instability.

The integration of intermittent renewables like wind and solar adds further complexity, requiring real-time balancing and fast-response capabilities. In many regions, grid operators are already facing challenges in maintaining reliability, especially during peak demand or extreme weather events. Without evolving through significant upgrades, grids will increasingly face bottlenecks, outages, and inefficiencies that threaten not just digital infrastructure, but economic productivity and societal resilience.

Infrastructure investment is urgent

Meeting the needs of a digital, electrified economy will require trillions of dollars in investment. Transmission and distribution networks must be expanded and modernised to handle higher loads and more distributed generation. Energy storage systems – both short-duration and long-duration – will be critical to smoothing out supply and demand fluctuations.

But investment alone isn’t enough. Policy reform is essential to unlock private capital and accelerate deployment. Planning and permitting processes must be streamlined, and regulatory frameworks must evolve to support innovation and flexibility. Governments, utilities, and investors must work together to create an environment where infrastructure can be built at the pace and scale required.

The countries and regions that move fastest to upgrade their energy systems will be best positioned to lead in the digital economy. Those who lag risk falling behind, both economically and environmentally.

AI is transforming grid operations

Ironically, the very technology driving the surge in demand – AI – is also part of the solution. AI is opening up new opportunities for faster, more accurate and democratised modelling scenario development and forecasting. Advancements in sensor technology, IoT devices, Satellite Imaging, and real-time monitoring are exponentially increasing both the quantity and quality of data, enabling AI systems to derive more accurate insights and predictions. Revolutionising how grids are operated and managed, enabling faster, smarter decision-making across the energy value chain.

From real-time forecasting and scenario modelling to autonomous load balancing and predictive maintenance, AI is helping grid operators respond to volatility and build resilience. By using machine learning algorithms to analyse vast datasets to anticipate demand spikes, optimise generation schedules, and detect faults before they cause disruptions, AI can turn hidden shifts in demand into actionable intelligence for traders, grid operators and policy makers alike. 

This vision is not the work of science fiction. During a recent heatwave a hyperscale data centre rerouted its computing load to avoid grid congestion – an AI-driven move that prevented price spikes and stabilised the local market. These kinds of invisible shifts are now visible, quantifiable, and actionable.

A connected future

The knock-on effect of the AI and the data centre boom accelerating innovation and amplifying disruption is a symptom of a much-needed, deeper transformation: an integrated way of thinking about energy insights. That is the key takeaway of the book I co-authored, Connected. 

The energy industry's traditional siloed approaches to decision making are inadequate for today's interconnected, rapidly changing markets and struggle with today's dynamic environment where multiple variables interact simultaneously.  Viewing the energy system as interconnected is more critical than ever because electrification, flexibility and decarbonisation increasingly bind together markets that once operated separately. 

Fossil fuels, renewable sources and power can no longer operate in silos: EV adoption shifts electricity demand, gas shortages ripple into power and heating, and investment in shared infrastructure affects multiple sectors at once. This convergence is reshaping how we think about infrastructure, investment, and innovation. It’s creating new opportunities for collaboration across sectors and new imperatives for action.

To thrive in this new era, we must embrace a systems-level approach. That means breaking down silos between energy, technology, and policy. As the energy system grows more complex, three capabilities will define success: seeing the complete picture, responding instantly, and adapting continuously. This requires a trifactor approach combining trusted, real-world data and AI-augmented decision-making capabilities to enable decision makers to respond instantly – whether reallocating investment, adjusting procurement, or rebalancing portfolios. 

It means investing not just in hardware, but in intelligence based on reliable, real-world data, software, analytics, and platforms that enable smarter, more adaptive energy systems. 

The AI revolution is here. The question is not whether our energy systems will change, but how fast - and how well - we can adapt. The data centre boom is a powerful signal that the time to act is now.