Nicu Popescu and Alan Riley write on the European Council on Foreign Relations website on why it is necessary for Europe to acceleration action. A global AI-driven surge in electricity demand is reshaping geopolitics, favouring states such as America and China that can rapidly expand power generation and grids. Europe risks becoming an energy-constrained AI follower.
Fast energy: How Europe can power the AI revolution and stay competitive
Summary
- Slow decision-making in Europe undermines the continent’s security, prosperity and political stability—from defence industrial output for Ukraine to exposure to hostile powers in energy and technology.
- A global AI-driven surge in electricity demand is reshaping geopolitics, favouring states such as America and China that can rapidly expand power generation and grids. Europe risks becoming an energy-constrained AI follower.
- Europe faces structurally higher energy costs than the US and China, as well as grid bottlenecks, permitting delays and carbon prices that erode its competitiveness. It also has significant resource constraints.
- Despite this, Europeans have significant assets—industrial depth, capital, world-class research capacity and a large single market. To overcome these challenges, they must launch a “fast energy” programme to speed up permitting, grid buildout and deployment of clean power.
- Using all available technologies and cutting energy costs, this programme is essential if Europe is to secure affordable electricity for AI, strengthen energy security and maintain economic and geopolitical influence.
Slow death
Europe can no longer afford to be slow. Too often, Europeans seem resigned to a pace of political decision-making that once symbolised deliberation and prudence but now means stagnation. Slowness is not harmless—it directly undermines European security, prosperity and political stability. The war in Ukraine is a painful reminder: Europe has struggled to ramp up weapons production at the scale or speed needed to support Kyiv or its own defence capacity effectively. Europeans’ inability to act quickly has strategic consequences: slowness erodes not just the continent’s autonomy; it effectively prolongs and aggravates a situation whereby the “kill switches” to Europe’s peace, energy security, economic competitiveness and political stability remain within reach of increasingly hostile powers and private players.
The same inertia is visible in the green transition. Europe once led in environmental innovation, yet it is now losing ground in electric vehicle production and renewable energy. Former European Central Bank president Mario Draghi noted that Europe is in danger of losing up to ten times more renewable energy generation than it loses today because of grid constraint capacity—a symptom of bureaucratic gridlock and insufficient investment coordination. Spain, for instance, has had to capnew energy production because it lacks grid capacity. Meanwhile, high energy costs driven by inefficiencies are deepening voter frustration and fuelling the rise of populist and authoritarian movements that promise action at any cost.
This sluggishness now threatens Europe’s position in the next great arena: artificial intelligence. Competing in AI is not just about algorithms, rare earths and microchips—it requires vast amounts of affordable and reliable electricity. For decades, on both sides of the Atlantic, electricity demand was broadly flat. That era is over. With the rapid deployment of large-scale AI systems, power demand is accelerating sharply.
In the US, electricity demand growth since 2021 has averaged around 1.7% a year, with the overwhelming share of incremental demand coming from the buildout of data centres linked to AI and advanced cloud computing. What may well be moderate projections from the US Department of Energy suggest that data centre power use could risefrom around 4% of American electricity demand in 2024 to around 9% by 2030. In some parts of the US, increased demand from data centres used for AI has pushed up consumer electricity prices. In response, US utilities and technology companies are racing to add capacity through a mix of natural gas, renewables and nuclear power.
This is not only an American phenomenon. In its public calls for urgent action on infrastructure, OpenAI has explicitly pointed to the scale at which China is expanding its power system. In 2024 alone, China added429GW of new generation capacity—more than one third of the entire installed capacity of the US grid—while America added closer to 50GW. Whatever the precise mix of technologies, the geopolitical significance is clear: countries that can build power systems the fastest will shape the AI era. Those that do not will find themselves losing not just economic leverage, but the capacity to sustain their standard of living and the quality of their public services.
Against this backdrop, Europe risks becoming an also-ran. This danger is compounded by the fact that the continent is relatively resource-constrained, with limited domestic sources of oil, gas and coal. At the same time, many member states have made nuclear power politically or administratively difficult to deploy. Meanwhile, as the Draghi report on European competitiveness has underlined, even with strong political support for the European Green Deal, renewable deployment has been slowed by complex planning, permitting and grid-connection procedures. These frictions raise both the cost of capital and the time required to deliver new capacity—often by many years—before even considering the supply chain risks associated with critical raw materials.
This paper argues that, despite the execution failures of recent EU energy policy, Europe has real structural advantages: a deep research base, a large pool of technology companies, an industrial sector bigger than America’s and a single market of more than 450 million relatively wealthy consumers. If Europe uses the AI challenge to force through a programme of “fast energy”—faster permitting, faster grid buildout, faster deployment of clean and firm power—it will strengthen its energy security, lower its long-term energy costs, accelerate decarbonisation and remain competitive in the AI age.
Energy and the AI revolution: The critical factor
The inevitable future
The scale of investment flowing into AI reflects a shared judgment by governments and firms in America and China: AI will be central to future economic growth and geopolitical power. Its effects are already visible in defence—for example, AI-enabled systems deployed in Ukraine; in health, where AI is used to search vast pharmaceutical libraries; and in manufacturing, with highly automated “dark factories” in East Asia. At the physical core of this transformation sit hyperscale data centres. Running them requires enormous and continuous amounts of electricity.
The energy implications of AI-focused data centres differ fundamentally from those of the earlier cloud-computing wave. Between around 2005 and 2017, the number of data centres grew rapidly, yet total electricity use remained relatively flat because efficiency gains offset growth. Cloud services primarily store and move data. By contrast, AI computes intensively—and computation is hungry for energy.
A commonly cited figure is that an AI query requires ten times the electricity of a standard web search. Even that shorthand understates the complexity of the issue. First, energy use varies widely depending on the time of day, the model and the nature of the task—text, images or especially video. Second, unlike mature cloud services, AI systems are still on a steep improvement curve: with more parameters, more training, more inference and therefore structurally rising energy demand.
A simple illustration shows how quickly this scales: consider a short, low-quality, AI-generated video clip—say, a few seconds at very modest resolution. Estimates suggest that even such a basic request can consume several million joules of energy: enough to run a microwave for an hour or power an e-bike for dozens of kilometres. High-definition, photorealistic video requires far more. If such capabilities become cheap and ubiquitous—as consumer technology historically tends to—usage will surge, and so will electricity demand. Crucially, most AI-related energy use does not come from training large language models, but from inference, ie, the day-to-day use of AI by consumers and firms. Various estimates put inference at 80-90% of total AI compute demand. As models improve and are embedded into more products and services, usage will only grow. The direction of travel is towards greater complexity, more personalisation and more continuous interaction—all of which are energy-intensive.
Assessing the precise trajectory is difficult because the industry is opaque. Most leading systems are closed (the data sources are opaque), and detailed energy data are not public. Researchers therefore rely on open models, such as Meta’s Llama, as proxies, which may not be representative. Even so, the trend towards significantly greater energy use remains clear.
In 2024, data centres accounted for roughly 4% of American electricity consumption. Some conservative official estimates suggest this could rise to around 9% by 2030, with half of that power going specifically to AI workloads. Other estimates provide a range, the lower end of which is consonant with reports from official sources. These estimates indicatethat electricity used for AI alone could reach the order of 165-325TWh a year by 2028. In contrast, the International Energy Agency estimatesthat the 2024 consumption of all American data centres was just over 180TWh.
Investment plans are consistent with these numbers. The US government-backed “Stargate” AI infrastructure investment initiative envisages around $500bn invested in a small number of new hyperscale sites. Each site will potentially require several gigawatts of dedicated power—comparable, individually, to the average load of a small European country. Apple has announced plans to spend hundreds of billions of dollars on data and manufacturing facilities over the coming years. Globally, investment banks such as UBS estimate annual data-centre investment will approach half a trillion dollars by the late-2020s.[1]
The challenges to overcome
None of this will be possible without generating and deploying vast amounts of power. However, delivering this at scale and speed is not easy even for the energy-rich US. Natural gas is currently the marginal source of supply for many data centres, but high-efficiency gas turbines are expensive and often stuck in long manufacturing queues. Small modular nuclear reactors are frequently discussed, but none is likely to be operating at commercial scale before 2030, and conventional nuclear plants face long regulatory and construction timelines. Repowering existing plants—such as the Microsoft deal to support the reopening of the Three Mile Island nuclear plant in the US—can help at the margin, but it is not a scalable strategy.
Grid infrastructure is an additional bottleneck. New transmission lines in advanced economies often take four to eight years from planning to completion; new grid connections can take anywhere from three to more than ten years. Locating data centres closer to generation helps only partially, because power still has to be moved and balanced across networks. Political tensions can also arise from concentrating large new loads in specific regions. It risks raising local electricity prices for households and small businesses, as has already occurred in parts of the US. At scale, this invites political backlash and pressure for price caps, shifting costs back onto either taxpayers or the AI firms themselves.
Nevertheless, the US is moving at speed with executive orders directing federal agencies to cut permitting restrictions and financially support the rollout of data centres. The White House has also published an AI action plan to further underpin its deregulation programme and infrastructure buildout. The administration is taking steps to bring more of the regulatory process under federal control in order to streamline it, and it has reduced existing options for judicial review of federal permitting legislation. As noted, big cost rises for American consumers can whip up political headwinds, but these are nevertheless unlikely to slow the US rollout of power generation and grid networks for AI-focused data centres.
Europe: The risk of being an energy also-ran
At first sight, Europe is in a more difficult position than America or China to furnish the necessary energy resources to develop AI at scale. However, it can solve these energy challenges. Together with the broader family of European democracies, the EU and its member states possess the industrial capacity, technical knowhow, financial resources and market scale to deliver secure energy at an affordable price and speed.
Since the early 20th century, Europe has been resource-constrained by comparison with its principal competitors. It has limited oil, gas and coal. By contrast, the US is the largest producer of both oil and natural gas and is one of the world’s three-largest producers of coal. China is the world’s largest producer of coal and makes approximately 60% of its electricity from coal. Given these resources, it is not surprising that China has over 1260GW of coal-fired power generation capacity. In addition to that fossil fuel resource base, China has rolled out over 3400GW of renewables capacity and has 61GW of nuclear power capacity, with a further 40GW of capacity coming on stream by 2030.
In this context, European states have looked at a number of different energy solutions to mitigate their energy resources constraints over the last 40 years. For example, France focused a significant part of its energy policy on developing a large fleet of nuclear power plants. By contrast, Germany focused principally on developing large-scale flows of Russian gas into its domestic market.
For a mix of energy security and climate change reasons, the more recent focus at national and EU level has been on developing and then rolling out renewables across the EU’s electricity markets. Over 320GWof solar has been rolled out across the bloc since 2015, with wind power increasing by around 94GW (giving a total installed base of 236GW as of the end of 2025).
Europe’s resource constraints have also had a major impact on the cost of its energy. Europeans do not have access to the cheap natural gas available to America or the cheap coal available to China. Transport costs and factors such as liquefaction and gasification for natural gas significantly raise the cost of imported energy. And renewables themselves have a limited impact on price because price is set in EU markets by the merit order (cheapest to the most expensive) and marginal pricing. This leaves natural gas as usually the most expensive fuel—which, as the most expensive fuel sent into the market last, sets the price for all the other fuels in the merit order. In addition, through its emissions-trading system (ETS), the EU imposes carbon taxes on fossil fuels. Such taxes are much less significant in China and almost non-existent in America outside California.
The gap between Europe’s energy prices and those of its principal competitors has widened since 2019. Covid-19, the Russian-orchestrated energy crisis and the invasion of Ukraine have had what appears to be a structural impact on pricing, pushing prices higher. Even after the effects of the energy crisis, in 2025 European wholesale day-ahead natural gas prices were between three to five times higher than in the US.
Chinese natural gas pricing tends to fall between EU and US pricing (using wholesale day-ahead prices, although it is difficult to provide precise figures due to the use of different benchmarks). However, when there is a surge in the cost of natural gas, as in 2022, China has the option of quickly switching to its plentiful coal-fired power supply, importing more natural gas by pipelines from Russia, and drawing on growing natural gas resources. Chinese electricity prices are much more regulated, so the nearest comparison is to look at an average of the available power tariffs, which puts China mid-way between EU and US pricing. The upshot of this, though, is that Europe has significantly higher energy costs than the US and higher energy costs than China. Over the last decade, that energy competitiveness gap has been widening. These high prices affect the EU’s capacity to develop AI at scale.
If that were not enough, Europeans’ capacity to roll out renewables is adding to cost and delay. As net zero mandates push demand for solar and wind, the cost of securing turbine parts and products to construct renewables systems pushes prices up and creates shortages of components. And, as in America, permitting can be very slow, takingover five years for new wind or solar plants; plans to strengthen the grid or develop new networks can take much longer.
Europeans must meet these challenges. Cheap, plentiful and secure energy is critical not just for developing AI, but also to undertake full electrification of the economy, and for rebuilding European industrial production capacity, particularly in relation to defence. Europe has the financial capacity, industrial base and technical knowhow to achieve this. It has an immense single market to support the delivery of a major transformation of its power capacity.
The EU has already taken some effective measures. For example, it sought to deal with some of the delays around permitting by adopting the Article 122 Permitting Regulation.[2] This has temporarily speeded up the rollout of renewable projects. This initial strengthening of accelerated permitting has been reinforced by the enactment of the Red III Directive, which seeks to permanently strengthen the capacity for renewable projects to be implemented.[3]
However, given these challenges, there is a compelling case for a much faster rollout of power generation to enable European AI development. To achieve this, Europeans should adopt a comprehensive “fast energy” programme. This would focus on three key factors: speed of action, a willingness to engage with all available technologies and bearing down on the costs of energy.
Speed, technology and cost: The fast energy programme
Any effective fast energy programme must comprise three critical parts.
1. Speed of action
The US view is that the AI race will be decisive, and one it must win. The EU and its member states also need to respond at speed. They must recognise that developing the energy resources for AI is tantamount to an emergency. Europeans cannot be bound to processes that will take half a decade or more to wind their way through EU and national procedures. As its names implies, any fast energy programme must place a premium on speed to get things done.
The EU has already been willing to take emergency measures in respect of permitting for renewables under Article 122. It did this under the exceptional situation created by the energy crisis, and the measures were temporary. Nevertheless, the economic and geopolitical consequences of failing to deliver the energy resources needed for AI should galvanise the EU to adopt further Article 122 measures to deliver a significant expansion of energy power generation. To make the fast energy programme a success, the EU should convene a small high-powered group made up of member state representatives, together with a commission representative, to push action forward at EU and national level. Working with industry, this group would identify barriers to development and measures that could support faster development. It would unblock legal and administrative paths and work with the EU and member states to remove such blockages with fresh legislative or administrative measures as necessary. The group would also be able to make recommendations to the EU and member states for key additional financial or regulatory support to increase the speed of power generation rollout.
2. All available technologies
To develop power generation at scale, the EU needs to be willing to use all available energy technology—not just renewables. As noted, one of the most difficult challenges with renewables is the rising cost of parts and materials for solar and wind plants, as well as the grid components. Even with significant political will, capital and the removal of most regulatory barriers, supply chain demands impose major delays and threaten significant rising costs. As a consequence, the EU needs to use other technologies.
One such major technology is nuclear energy. Some member states have renewed their interest in nuclear energy, with programmes under way in France, Poland and the Czech Republic. Many more have indicated their interest in traditional nuclear power plants and small modular reactors. That being said, there are still no fully licensed and operating small modular reactors anywhere in an OECD country, and none is likely to be operating in the EU until the 2030s.
The critical difficulties with developing new nuclear power plants are the major cost overruns and significant delays that blight virtually every such project. One issue for the EU’s fast energy programme will be to cut the cost and delivery time for building new nuclear power plants. A starting point is Britain’s Nuclear Regulatory Review, chaired by Dr John Fingleton, which reported at the end of 2025. The review makes a compelling case that the entire UK nuclear regulatory system is over-elaborate, over-engineered and heavily fragmented—a situation which adds massively to the costs and delay in the delivery of British nuclear power plants. Britain is infamously the most expensive country in the world in which to build a nuclear power plant. Nevertheless, much of the critique of cost and delay in the British nuclear sector applies to most European nuclear sectors.
The fast energy programme should therefore draw inspiration from the Fingleton model to undertake a regulatory review of member states interested in building out their nuclear capacity. Such reviews should be undertaken speedily, as was the case in Britain, where it launched in April and reported in December. Such reviews could give guidance to member states on how to streamline their nuclear regulatory systems. There would be a place here for action at EU level, as it is likely that some elements of streamlining would necessarily involve elements of European law. The goal should be to significantly cut back on the cost and time for building nuclear power plants so they can play a role in supporting an AI rollout.
The fast energy plan should aim for “near to market” energy generation and storage products which may support increased power generation. One course worth serious consideration is to develop near-to-market battery storage products which are cheap, safe and have significant energy density. A number of firms are in the process of bringing such products to market. Such batteries could play two major roles in supplying energy to AI data centres. First, they provide backup and support to data centres and make it possible to use more of their own renewable power in such centres, reducing dependence on the grid. Second, and particularly in southern Europe with its greater capacity to generate solar power, the deployment of cheap domestic battery storage will reduce overall demand on the grid. Such a “cheap battery” rollout has a compelling “fast energy” advantage here. Batteries for domestic and light-industrial use, where twinned with solar panels, do not need grid connections or upgrades. Without these, rollout can be much quicker.
Another technology and energy source for the fast energy plan are natural gas and combined cycle gas turbines (CCGTs). These are significantly more efficient than traditional gas-fired power plants. Given Europe’s recent experience of dependence on Russian natural gas, there is understandably wariness among member state governments about becoming dependent on imports of liquefied natural gas (LNG).
However, several factors are likely to significantly reduce the risk of using natural gas. First, the scale of LNG imports coming on stream in the US creates a huge incentive for the to keep supplying Europe with imports. Second, more LNG coming on stream globally from Canada, Mexico and Indonesia (among many others) balances the risk of dependence on US gas. Third, China is one of the world’s largest importers of LNG, but structural factors are leading it to use less LNG. These include the country’s low growth, rising domestic gas production, access to more pipeline gas and its own rollout of nuclear energy. Europeans can further minimise their dependence risk by seeking to develop their own domestic gas resources both onshore and offshore. Hence, even with the current security and energy crisis in the Gulf, LNG does not carry the same scale of risks as Russian pipeline gas.
One of the key barriers to deploying natural gas is access to sufficient CCGTs. The timetable for the manufacture of CCGTs is currently three to four years. But the European fast energy programme will be able to take several measures to address this. For example, policymakers could attempt to draw up a contractual and regulatory structure to incentivise the building out of many additional CCGTs. The EU and its member states could support a framework agreement under which large-scale data centres are guaranteed to be built, with AI developers then having the confidence to contract for a large number of CCGTs. In turn, this would strengthen manufacturers’ confidence to expand production.
The EU’s fast energy programme will need to look across the range of existing and near-to-market power generation and storage systems—and seek to promote their use to rapidly increase Europe’s power generation capacity.
3. Bearing down on costs
As explained above, one major and recurring problem for Europe is the cost of energy. In the early 2000s, Europe appeared to have entered a world of flattened demand for energy, underpinned by three factors. First, increasing energy efficiency and conservation, which led to economic growth with less energy use. Second, European deindustrialisation reduced energy demand. And third, energy prices remained low. In that world, energy costs did not feature highly in European business or political discourse. Now, that context has now changed decisively because of rising energy costs, especially after the 2021-22 crisis; growing European energy competitiveness concerns; and economic and security demands to increase power generation capacity. The EU’s fast energy programme must therefore focus on bringing energy costs down.
Natural gas
Natural gas costs have been a significant concern since the onset of the 2021-22 energy crisis. Following the crisis, the Draghi report recommended a greater switch to long-term supply contracts in the gas sector and a reduction in reliance on the spot market. This was an understandable reaction because in August 2022 prices spiralled to over €340 per MWh (the price between 2009 and 2019 ranged between €9-€29 per MWh).
However, since 2022 the global natural gas market has been characterised by substantial flows of LNG exports entering the market from the US and Qatar; concerns about spot markets are thus now overblown. There is a case for the EU and its member states to enable long-term supply contracts of around 15-20 years in length. This would allow LNG exporters to offer significantly low long-term energy prices. To the extent that there are EU and national regulatory barriers to such contracts, the fast energy programme should aim to remove these barriers. Policymakers should aim to secure a mix of long-term contracts and spot market reliance, creating a more liquid market, supply diversity, high storage levels and increased domestic production. This would protect Europe against any significant supply disruption and spiralling energy prices. It should also open up the prospect of overall lower natural gas costs going into the next decade.
Securing diverse external gas supplies in this way can be reinforced if decision-makers also strengthen domestic energy security. Here, the fast energy programme would focus on member states developing domestic natural gas resources. The programme would also aim to develop new lines of gas supply with near neighbours such as Algeria and other states on the Mediterranean Sea. This would mean expanding storage facilities, both domestically and in neighbouring states. For example, Ukraine has very large natural gas storage facilities equivalent to approximately 30% of total EU storage. Europeans could develop mechanisms to ensure their natural gas storage facilities can be more easily accessed by European companies.
Renewables
The most difficult energy cost to deal with is the impact of the ETS carbon price. As the most recent OECD economic survey of the EU underlines, Europeans faced a carbon price of €70-€80 t/Co2 in 2025. The EU’s ETS was originally envisaged as a forerunner of a global carbon regime which other states would join. Had they done so, it would have limited the impact on the EU’s international competitiveness. However, the rest of the world is not following the EU. The bloc has thus ended up lumbered with both high domestic energy prices and declining international competitiveness.
The cost issues flowing from the carbon price are compounded by the system costs that flow from renewables. While the actual wind or solar power is free at source, the cost of distribution (getting the power to market) and the capacity to deal with the intermittency across the grid are expensive. In addition to this, a total system backup is required to replace wind and solar when these are not available; this too is expensive. In essence, all EU states have ended up running two power systems—one green, and one largely fossil fuel, usually natural gas, which is always available on standby.
The EU’s fast energy programme will need to consider ways of tackling these costs. The European Commission is launching a review into the ETS, which will hopefully focus on the carbon costs. It may be possible to address other costs, including system costs, by seeking to create programmes to develop domestic supply chains for products necessary to run the grid and enable interconnections.
A further task for the fast energy programme is to address the cost of additional wind and solar plants coming onto the grid. Renewables have been permitted to access the grid largely without regard to the cost (there are limitations that apply, such as environmental or access capacity, but cost is not one of them). The fast energy programme should draw up criteria and a mechanism focusing on the cost of wind and solar plants to reduce costs to industry and consumers.
Additional options include progressing existing plans to use more power purchase agreements and contracts for difference. This would reduce the impact of existing market mechanisms on natural gas and electrical power prices.
Nuclear
As noted, one way of cutting the costs of the delivery of nuclear power is to undertake a review of member states’ existing nuclear regulatory systems and significantly reduce regulatory divergence between member states. The fast energy programme should work with industry to identify ways to cooperate on rapidly increasing power generation. For example, following a streamlining of some member states’ nuclear regulatory systems, one option for cooperation would be for the tech industry to co-finance a large number of nuclear power plants.
The evidence from the French nuclear power plant expansion programme in the 1970s and 1980s was that building at scale reduced overall capital costs and progressively speeded up delivery times as more and more nuclear power plants were built. By building at scale it was possible to provide the supply chains and skilled workforce to sustain a large number of nuclear power plants. There could be a win-win for member states, and for AI and tech firms, in supporting a mass rollout of nuclear power plants. Member states would get substantially more and cheaper additional power capacity at speed on the back of building at scale with the AI and tech firms. The AI and tech firms would obtain large-scale additional power capacity and can plan on that basis.
Powering the AI age
AI is not primarily constrained by algorithms, chips or capital. Rather, it is constrained by electricity—delivered reliably, affordably and at scale. America and China have already internalised this and are moving quickly to build generation and grid infrastructure. Europe, by contrast, faces a dual challenge: higher structural energy costs and slower delivery mechanisms. If it does not close the gap, Europe will import AI capability, export industrial competitiveness and cede geopolitical influence in a domain that will shape defence, health, manufacturing and services.
Yet Europeans are not condemned to fall behind. They have market scale, capital, industry and scientific depth. What they lack is speed. A credible fast energy programme—built around emergency-level delivery, technological pragmatism and relentless cost reduction—can unlock the electricity supply required for AI while simultaneously strengthening energy security and accelerating decarbonisation. The strategic choice is clear: Europe must treat building out power systems as the enabling infrastructure of the AI age, and act accordingly—quickly, at scale and with discipline on cost.
External link
Energy in Demand - Sustainable Energy - Rod Janssen 