The Nuclear Comeback: Why Governments and Tech Giants Are Betting Billions on Atomic Energy
Nuclear power has been declared dead, dangerous, and uneconomical so many times in the past three decades that the industry's current renaissance carries an almost paradoxical quality. The same technology that was being written off following Fukushima in 2011 — with Germany accelerating its phase-out, Switzerland voting to prohibit new plant construction, and the global nuclear pipeline shrinking to a handful of projects — is now the subject of the largest surge in government and private investment since the 1970s. The drivers are different from prior nuclear eras: not energy independence from OPEC, not Cold War strategic signalling, but the collision between three forces that were not simultaneously present in any prior decade. Decarbonisation commitments that require large-scale zero-carbon baseload generation. Data centre energy demand growing at a rate that intermittent renewables cannot practically satisfy. And small modular reactor technology that is, for the first time, genuinely close to commercial deployment at a cost structure that changes the economics of nuclear power fundamentally.
What Is Actually Driving the Investment Surge
The headline numbers are striking. The US Department of Energy's loan guarantee programme for advanced nuclear has committed over USD 15 billion in conditional loan guarantees since 2023. The UK government's Great British Nuclear programme is funding the development of SMR technology through a competitive process that has selected Rolls-Royce SMR as the preferred domestic technology. France has reversed its nuclear phase-down trajectory and announced EUR 1 billion in new nuclear R&D investment under the France 2030 programme. Japan has restarted 12 nuclear reactors since 2022 and is permitting operational life extensions that effectively restore a significant portion of the pre-Fukushima nuclear fleet. South Korea has returned to nuclear as the centrepiece of its long-term power generation strategy, reversing the previous administration's phase-out commitment. These are not marginal policy adjustments — they represent a structural reversal in the political economy of nuclear energy across the world's largest industrial economies.
The corporate dimension is equally significant. Microsoft's 20-year power purchase agreement with Constellation Energy for Three Mile Island Unit 1 — the first US nuclear plant restart driven by corporate power demand — was the transaction that made the connection between AI data centres and nuclear power explicit for markets. Google's announcement of agreements with Kairos Power for 500 MW of high-temperature gas-cooled reactor output, Amazon's nuclear agreements with Energy Northwest and Dominion Energy, and Meta's Request for Proposals seeking 1–4 GW of nuclear capacity collectively represent the largest corporate nuclear energy procurement in history. The common driver across these transactions is not environmental principle — though the carbon-free attributes of nuclear power are commercially valuable under the voluntary renewable energy frameworks these companies use — but load matching. Data centres operate at high utilisation continuously, 24 hours a day, 365 days a year. Solar and wind generate intermittently. Nuclear generates continuously. The physics of AI infrastructure creates a structural demand for baseload zero-carbon generation that no other technology currently provides at the required scale.
The Small Modular Reactor Thesis: Where It Stands in 2026
The investment narrative around nuclear's comeback rests significantly on small modular reactors — factory-built, standardised nuclear units with generating capacity of 50–300 MW per unit, compared to the 1,000–1,600 MW typical of conventional large-scale nuclear plants. The SMR value proposition is compelling in theory: factory manufacture of standardised components eliminates the one-off construction complexity that has driven cost overruns at large nuclear projects; modular deployment allows capacity to be added incrementally as demand grows rather than requiring a decade-long construction programme before generating a single watt; and smaller unit size makes nuclear economically accessible to markets — island grids, industrial clusters, remote industrial sites — that cannot absorb the output of a large conventional reactor. NuScale Power's VOYGR design received the first-ever SMR design approval from the US Nuclear Regulatory Commission in 2022, and Rolls-Royce SMR's 470 MW modular design is targeting UK regulatory approval by 2028 with the first deployment by 2035.
The honest assessment in 2026 is that SMR technology is closer to commercial reality than at any prior point but not yet there. NuScale's flagship Carbon Free Power Project in Idaho was cancelled in November 2023 after the power purchase agreements required to finance it failed to attract sufficient subscribers at the projected electricity price — a cautionary signal about the gap between the SMR cost projections that populate investment presentations and the actual project economics that utility procurement officers will commit to. Rolls-Royce SMR has not yet entered the regulatory review process in the UK. GE Hitachi's BWRX-300, Ontario Power Generation's chosen technology for Canada's first SMR, is in early engineering and the 2029 construction start target carries significant schedule risk. The SMR thesis is real, but the timeline expectations embedded in current investment narratives are optimistic by 2–4 years in most cases, and the cost uncertainty at first-of-a-kind commercial scale is substantially larger than the levelised cost projections from technology developers acknowledge.
The Political Economy That Has Changed
Understanding why nuclear is succeeding politically in 2026 where it failed in 2012 requires understanding what has changed in the political economy of energy, not just the technology. The most significant change is the energy security awakening that Russia's Ukraine invasion produced. European governments that had structured their energy systems around the assumption that geopolitical adversaries would remain reliable commercial energy suppliers spent the winter of 2022–2023 managing an energy crisis that vindicated every energy security argument that nuclear proponents had made for decades and that had previously been dismissed as alarmist. The willingness to accept the costs and complexities of domestic nuclear power generation — costs that were previously judged politically unacceptable — shifted measurably when the alternative was acute dependence on imported gas from an adversarial state.
The second political economy shift is the changed status of the anti-nuclear environmental movement. The mainstream environmental argument that nuclear power is incompatible with decarbonisation — an argument that dominated policy debate through the 2010s and contributed to Germany's energiewende and the closure of functioning zero-carbon plants in the US and Europe — has fractured. The IPCC's 2022 Working Group III report explicitly identified nuclear power as a key mitigation option. A cohort of prominent environmentalists, including Stewart Brand and the founders of the Breakthrough Institute, have made pro-nuclear arguments that have achieved mainstream visibility. Young climate activists, surveyed repeatedly about energy policy preferences, express significantly more openness to nuclear power than the environmental movement's institutional leadership. The political cost of pro-nuclear positioning has declined substantially, reducing the veto that opposition groups previously exercised over nuclear policy in democratic systems.
The Waste, Cost, and Timeline Realities That Persist
The nuclear revival's political momentum should not be confused with the resolution of the technical and economic challenges that constrained the technology before the current investment cycle. Nuclear waste remains an unresolved political problem in virtually every country operating nuclear plants, with permanent geological disposal repositories existing only in Finland (Onkalo, scheduled to begin receiving waste in the late 2020s) and under construction in Sweden. The US has not sited a permanent repository since the Yucca Mountain project was cancelled in 2010 for political rather than technical reasons. The cost of large-scale nuclear construction in Western regulatory environments has not declined: EDF's Hinkley Point C in the UK has seen costs escalate to GBP 35+ billion for 3.2 GW of capacity, and Vogtle Units 3 and 4 in the US came in at approximately USD 35 billion — roughly double original estimates. These projects are the most recent data points on the actual cost of conventional nuclear construction in Western environments, and they are not encouraging for the economics of future conventional-scale projects.
The nuclear industry's response — that SMRs change the cost trajectory through manufacturing standardisation and factory production — is theoretically sound but empirically unproven at commercial scale. The first SMR projects will face the same first-of-a-kind cost premiums that plagued early offshore wind projects, early utility-scale solar plants, and every other energy technology at its commercial debut. The crucial question is whether the SMR learning curve is steep enough — and the commercial pipeline deep enough — to drive costs down to the levelised cost projections that currently make the economics of SMR power purchase agreements attractive to corporate and utility buyers. The nuclear comeback is real, but the decade from 2026 to 2036 will be the test of whether the technology can deliver on the economic promise that the political investment has been made in anticipation of.
What the Nuclear Comeback Means for Energy Markets and Investors
For energy markets, the nuclear revival creates demand and supply dynamics reshaping long-term power price expectations. The corporate nuclear PPA market — currently dominated by hyperscalers but expanding to industrial energy users with net-zero commitments — is creating a new buyer segment structurally absent from nuclear economics in prior decades. Plant life extensions, new Asian and Middle Eastern construction, and the SMR pipeline collectively add zero-carbon baseload supply that alters capacity adequacy calculations for grids retiring fossil fuel capacity and integrating renewables. For investors, the most direct exposure is in uranium — where supply-demand fundamentals have tightened as utilities rebuild strategic inventories — and in the specialist engineering and component manufacturing companies serving the nuclear sector regardless of which reactor designs achieve commercial deployment.