Electricity has powered human progress for centuries, yet it still makes up only about 20% of global energy use.[1] Despite climate urgency, electrification has lagged behind the Paris goals. A new force is stepping in, one that feeds on electrons and is growing at breakneck speed: Artificial Intelligence.
Could this finally ignite the long-awaited electricity revolution? Or will it derail the energy transition?
Electrification is central to the energy transition because it is the most easily decarbonized form of energy. Progress is real but still insufficient for Net Zero -- electricity’s share of global energy consumption has inched up from 15% in 2000 to around 20% today. Yet decarbonization overall is slow, particularly in industry and transport, where geopolitical and competitiveness concerns still outweigh climate goals.
Although falling costs for renewable electricity generation and supportive policies are driving gains, the story is mixed. Passenger cars are rapidly electrifying, with EV sales expected to surpass 20 million units in 2025 (about 25% of global sales).[2] At this pace, car-related emissions could align with Net Zero by 2030, one of the few areas on track. Aviation, shipping, and long-haul freight lag far behind. Hard-to-abate sectors such as cement, steel, and glass face technical and economic hurdles.
With climate diplomacy losing momentum, a new wild card has emerged: Artificial Intelligence (AI). Its rapidly rising electricity needs could disrupt the energy landscape — potentially accelerating the transition or complicating it.
AI is reshaping global electricity systems. Its rapid growth is increasing power demand — especially from data centres — and transforming how energy is produced and consumed. This poses challenges for energy markets and may hinder progress toward climate goals, with long-term impacts still uncertain.
The rapid expansion of AI is driving an unprecedented surge in global electricity demand, requiring massive and accelerated investments not only in new power generation but also in power grids and energy storage to maintain system resilience and flexibility. Unlike the past two decades — when energy demand growth came mainly from emerging markets — AI-driven data centre construction is now pushing demand sharply higher in developed economies, especially the US.
In 2024, data centres used around 415 terawatt-hours (TWh) of electricity (about 1.5% of global consumption).[3] The IEA projects this demand to more than double to 945 TWh by 2030, and their estimate is conservative compared to those of other forecasters. Longer-term estimates suggest data centre demand could soar to 3,500 TWh by 2050, comparable to current total electricity consumption in India and the Middle East combined. AI is expected to be the fastest-growing driver of electricity demand in the early 21st century.
The race by (AI) hyperscalers to secure huge amounts of electricity through long-term power purchase agreements (PPAs) is pushing up power prices and accelerating renewable deployment. About half of the power capacity additions these AI hyperscalers over the next decade is expected to come from renewables.[4] These investments will have cost impacts. While policy choices will determine how much households pay, the IMF warns that if renewable and transmission expansion are insufficient to cope with the growth of data centres, US electricity prices could increase as much as 8.6% by 2030.[5]
Now the trillion-dollar – and trillion-watt! -- question: will AI accelerate, or further derail, the energy transition?
Depending on how carbon-intensive the electricity generation choice is, these expansions could come with a significant carbon price tag. Yet AI could also improve efficiency and reduce overall energy use — counterbalanced by potential rebound effects (Jevons paradox), where efficiency gains lead to even higher consumption. Ultimately, the net impact will depend heavily on political decisions, regulatory frameworks, and infrastructure development.
Utilities are transitioning from traditional power suppliers to key coordinators of the electrified economy, generating power, expanding and managing grids, and balancing supply and demand over the short and long-term. Once largely overlooked in equity markets, the utilities sector is now viewed as a critical pillar of the AI growth story.
Within utilities, subsectors will not benefit equally. Future share price benefits will depend on value chain and regional positioning. Europe is unlikely to encounter the level of power shortages emerging in the US but could face peak-demand stress, boosting the need for flexible generation (including gas) and for energy-storage or battery solutions. We see the potential of continued earnings and dividend generation in the sector over the period to 2030. We could imagine growth to continue for regulated utilities and integrated companies, ( and integrated companies, likely driven by their regulated power network activities), led by power network developers.
We still see three building blocks of the energy transition:
The surge of AI and the accelerating rollout of renewable energy mark a pivotal moment in the global energy transition. Together, they have the potential to shift electrification from a slow structural trend to a true industrial revolution. But unlocking this potential will require overcoming significant constraints, from grid bottlenecks to energy security and financing. Among the most pressing is the secure supply of critical minerals, essential both for AI hardware and for renewable technologies, that will continue to reshape both geopolitics and industrial strategies. Ultimately, the success of the electricity revolution and its ability to deliver economic progress while delivering on climate goals will hinge on the governments building a long-term vision and steering the AI revolution towards the better good.
[1] World Energy Outlook 2025 – Analysis - IEA
[2] World Energy Outlook 2025 – Analysis - IEA
[3] Energy and AI – Analysis - IEA
[4] Energy and AI – Analysis - IEA
[5] Power Hungry: How AI Will Drive Energy Demand
[6] Renewable capacity statistics 2025, IEA
Get information faster with a single click