AI data centres have turned nuclear power from a legacy grid asset into a strategic priority on both sides of the Atlantic. Global data centre electricity consumption is set to more than double by 2030, and 24/7 baseload power isn’t something wind and solar can reliably supply — not without storage that doesn’t yet exist at the scale required.
This article is part of our series on the nuclear power renaissance reshaping AI infrastructure. The US and UK are both responding, but with very different policy architectures that reflect very different starting positions. This article compares what each country is actually doing — through the US Genesis Mission and the UK Advanced Nuclear Framework — looks at the real obstacles, and draws out what these frameworks mean if your infrastructure planning extends to the early 2030s.
There’s also one proposal worth covering that tends to get overlooked: a Texas startup petitioning the DOE to repurpose retired US Navy nuclear reactors for an AI data centre at Oak Ridge National Laboratory in Tennessee. It sits squarely inside the Genesis Mission framework, and it reveals both the creative possibilities and the hard limits of where nuclear policy currently stands.
What is driving governments to make nuclear power a policy priority for AI?
AI data centres need continuous, reliable baseload power — the kind that can’t be switched off when the wind drops. That makes nuclear attractive in a way it simply wasn’t during an era of steady electricity demand and steadily falling renewable costs.
The hyperscalers have given governments political cover to act. Amazon, Google, Microsoft, Meta and others have all signed nuclear power purchase agreements or made public commitments. When the world’s largest technology companies are publicly competing for nuclear power contracts, governments can frame their policy response as meeting demonstrated private sector demand rather than backing an unproven bet.
Both the US and UK have framed advanced nuclear as an AI infrastructure play — not just a clean energy initiative. That framing is relatively new, and it changes the urgency of both policy frameworks considerably.
The US enters this period with the world’s largest operating commercial nuclear fleet — roughly 93 reactors — and seven decades of regulatory infrastructure. The UK enters with fewer operational plants, a declining legacy fleet, and a regulatory system that needs to adapt to novel small modular reactor designs. Both are racing to unlock private capital. Neither has delivered new commercial SMR capacity.
What is the US government’s nuclear strategy for AI infrastructure, and what is the Genesis Mission?
The DOE’s Genesis Mission is the Trump administration’s umbrella AI infrastructure initiative. It positions the DOE as both policy enabler and financier, with the loan guarantee programme as the practical financing mechanism — reducing the cost of borrowing for qualifying projects, but not providing direct grants. The Genesis Mission creates the policy framework and signals political priority. It doesn’t guarantee funding for any specific project.
The demand-side complement is the emergency PJM auction proposal. The Trump administration and a bipartisan group of governors — including Josh Shapiro in Pennsylvania and Wes Moore in Maryland — called on PJM Interconnection, the US’s largest electricity market, to hold an “emergency” power auction offering 15-year electricity contracts. Long contract terms are what makes long-lead-time projects like nuclear economically viable. This proposal is the grid emergency auction as a direct policy response to interconnection queues running five to seven years in key data centre markets — the infrastructure context that makes the policy urgency legible.
Two caveats worth noting: neither the White House nor the governors can mandate PJM to hold this auction, and Bloomberg reported PJM wasn’t even invited to the announcement. As of early 2026, it remains a proposal. Nuclear benefits from the administration’s pro-fossil-fuel energy environment, but that environment isn’t guaranteed to persist.
What is the NRC licensing process and why does it constrain SMR timelines?
The Nuclear Regulatory Commission licenses all civilian nuclear reactors in the US. No commercial SMR design can begin construction without completing its review process — safety evaluations, environmental reviews, public comment periods — a multi-year undertaking for novel designs, which all current SMRs are.
Even with loan guarantees and political urgency in place, the NRC timeline is the rate-limiting constraint on any new US nuclear project.
TerraPower‘s Natrium reactor — 345 MWe, developed with GE Hitachi — illustrates the smoother end of this spectrum. The NRC approved its construction permit in March 2026, the first commercial-scale reactor approved in nearly a decade and five months ahead of schedule. TerraPower used a risk-informed, performance-based licensing methodology endorsed by the NRC in 2020. The Kemmerer, Wyoming project is still targeted for 2030 — three years beyond the original DOE target. Faster licensing doesn’t collapse construction timelines.
Oklo‘s path has been different. Its Aurora Powerhouse design is pursuing authorisation under the DOE’s Reactor Pilot Programme — a pathway created by executive order in May 2025 — rather than going through NRC commercial licensing first. The DOE approved the Nuclear Safety Design Agreement for Aurora-INL in March 2026, the first step under that pathway, not a commercial licence.
The practical implication: a nuclear power purchase agreement with TerraPower carries different regulatory risk from one with Oklo — not because the underlying technology is fundamentally better or worse, but because their licensing pathways are at different stages of certainty. That’s a vendor risk factor worth understanding if you’re tracking nuclear supply commitments. For a detailed look at the SMR technologies at the centre of these policy frameworks — including cost per MWh and capacity comparisons across Oklo, TerraPower, Kairos and X-Energy — see our SMR technology comparison.
What is the UK Advanced Nuclear Framework and what does it aim to do?
The UK Department for Energy Security and Net Zero launched the Advanced Nuclear Framework on 5 February 2026, with the explicit goal of attracting private capital to fund SMR and advanced modular reactor development in the UK.
The framework has two main pillars. First, a formal pipeline of assessed projects: developers apply, Great British Energy-Nuclear evaluates applications, and successful applicants receive government endorsement in principle — then go and secure private finance themselves. Second, a concierge-style planning navigation service to actively help approved developers through UK planning consents, regulatory approvals, and investment processes. The goal is to stop bureaucratic complexity killing projects before they start.
The National Wealth Fund can act as a catalytic investor — a public capital backstop to reduce first-mover risk for private co-investors, without committing to specific upfront funding.
The first designated SMR site is Wylfa on Anglesey, Wales, announced in November 2025. Wylfa previously hosted a Magnox reactor decommissioned in 2015, giving it existing nuclear site status — helpful for planning consent and community acceptance.
Two early project examples: X-Energy and Centrica plan 12 high-temperature gas-cooled reactors at Hartlepool; Holtec, EDF and Tritax are targeting the former Cottam coal site in Nottinghamshire.
On timelines: Omdia principal analyst Alan Howard assessed in late 2025 that SMR commercial viability in the UK is not expected until the 2030s. The framework creates the conditions for deployment — it doesn’t change construction physics or reactor manufacturing lead times.
How do US and UK nuclear policy approaches compare?
The structural differences between the two frameworks affect the credibility of each country’s deployment timelines. Here’s how they stack up.
Starting position. The US has roughly 93 commercial reactors and seven decades of operational and regulatory expertise. The UK is starting much closer to scratch — Wylfa is the first confirmed new nuclear site in years. The US is leveraging existing infrastructure; the UK is building new capacity from near-scratch.
Financing model. The US uses DOE loan guarantees and the IRA, with Genesis Mission adding political urgency. The UK relies primarily on attracting private capital, with government endorsement as the incentive and the National Wealth Fund as a limited backstop. Neither model guarantees project finance — both reduce risk for private investors rather than replacing them.
Regulatory approach. The NRC is established but historically slow. The Trump administration’s support for risk-informed, performance-based methodology and appetite for NEPA reform signals genuine acceleration, but even accelerated processes take years. The UK’s concierge service is designed for speed but remains untested at commercial scale.
Vendor overlap. The same developers are active in both markets. X-Energy has an Amazon deal in Washington State and a Centrica partnership at Hartlepool. TerraPower is exploring UK deployment alongside its Wyoming build. If either vendor hits regulatory or financial difficulties, timelines in both markets are affected simultaneously.
Political durability. Genesis Mission is a Trump administration initiative — its continuation depends on political continuity. The UK framework was launched under Labour and framed as cross-party national interest. That’s somewhat more politically durable, but it’s also new and unproven.
The bottom line: the US is leveraging existing infrastructure with political urgency overlaid. The UK is building new policy architecture from scratch with ambitious timelines. Both are racing to unlock private capital. Neither has commercial SMR capacity online. For the full picture of how these policy frameworks fit within the broader nuclear-AI buildout story, see our series overview.
What is the Navy reactor repurposing proposal and how does it fit the policy framework?
Texas-based HGP Intelligent Energy petitioned the DOE for a loan guarantee to repurpose two retired US Navy nuclear reactors for an AI data centre at Oak Ridge National Laboratory in Tennessee.
The reactor types proposed are the Westinghouse A4W — which powers Nimitz-class aircraft carriers — and the General Electric S8G, which powers Los Angeles-class submarines, with nearly a third already decommissioned. Proposed output: 450 to 520 megawatts from two reactors. Total project cost: $1.8 to $2.1 billion. Cost per megawatt comes in at $1 to $4 million — significantly lower than new-build SMR estimates.
The US Navy has operated more than 100 nuclear reactors for over 50 years without a radiological accident. HGP CEO Gregory Forero’s pitch is straightforward: “We already know how to do this safely and at scale.”
The proposal fits cleanly inside the Genesis Mission loan guarantee frame — it’s an AI infrastructure project using existing proven reactor technology rather than a first-of-a-kind commercial design. If approved, it would be the first time a military reactor has been repurposed for civilian use.
What it doesn’t resolve: there is no NRC licensing precedent for converting a military reactor to civilian use. Decommissioning liability — and HGP acknowledges that dealing with retired nuclear materials is insanely expensive — has to be accounted for, and the proposed decommissioning fund is untested at this scale.
This proposal is a useful test case for what Genesis Mission can and cannot do. It can create access to loan guarantee mechanisms. It cannot create regulatory precedents that don’t exist.
What do these policy frameworks mean for companies with UK or European operations?
If your cloud infrastructure includes European data centres, the UK Advanced Nuclear Framework is more directly relevant to your planning than Genesis Mission. The question is whether UK nuclear deployment can reduce power constraint risk in a region where that constraint is already real.
AWS has publicly stated that Britain needs more nuclear power to feed its AI data centre surge — a statement made in 2025, before the framework launched. This is a response to an existing bottleneck, not a precautionary measure.
If your team uses UK cloud regions — AWS eu-west-2, Azure UK South, GCP europe-west2 — the relevant risk isn’t immediate. Continued power constraints could slow capacity expansion, affect availability SLAs, or filter through into higher long-term compute costs. That’s a three-to-five year planning consideration, not an operational concern today.
The X-Energy cross-market risk is the most direct exposure for organisations tracking both markets. Any supply commitments depending on X-Energy’s delivery schedule carry shared risk across US and UK timelines simultaneously.
Power being committed to today won’t flow for seven to ten years in the best case. Keep making cloud region decisions on current capacity, latency and compliance requirements. Nuclear power availability is a medium-term variable to monitor, not one that warrants infrastructure decisions today.
Frequently Asked Questions
What is the Genesis Mission? The Trump administration’s AI infrastructure initiative, led by the DOE. It positions the DOE as policy enabler and financier, with the loan guarantee programme as the practical financing mechanism. It provides the policy frame for proposals like HGP’s Navy reactor project — but it doesn’t guarantee funding for any specific proposal.
What is the UK Advanced Nuclear Framework? A government programme launched by DESNZ on 5 February 2026 to attract private capital for SMR and advanced modular reactor development in the UK. It includes a formal project pipeline, a concierge-style planning navigation service, access to the National Wealth Fund as a catalytic investor, and assessment by Great British Energy-Nuclear. First designated site: Wylfa on Anglesey, Wales.
What is the NRC and why does it slow down SMR deployment? The Nuclear Regulatory Commission licenses all civilian nuclear reactors in the US. Because SMR designs are novel, the NRC must conduct extensive safety reviews before construction begins — a multi-year process. This is the primary constraint on how quickly US SMR projects deliver power, regardless of capital or political support.
Why has Oklo taken a different path from TerraPower with the NRC? Oklo is using the DOE’s Reactor Pilot Programme pathway — established by executive order in May 2025 — rather than NRC commercial licensing first. TerraPower used a risk-informed, performance-based licensing methodology and GE Hitachi’s regulatory experience, winning NRC construction permit approval in March 2026. Both paths are legitimate; they carry different timelines and levels of near-term certainty.
What are the Westinghouse A4W and GE S8G reactors? The A4W powers Nimitz-class nuclear aircraft carriers. The S8G powers Los Angeles-class nuclear attack submarines. Both are US Navy designs with decades of operational history. HGP Intelligent Energy has proposed repurposing retired examples for a civilian AI data centre at Oak Ridge National Laboratory — which would be the first military-to-civilian reactor conversion in history.
What is Wylfa and why is it significant? A site on Anglesey, Wales, designated as the UK’s first SMR construction location in November 2025. It previously hosted a Magnox reactor decommissioned in 2015, giving it existing nuclear site status — which helps with planning consent and community acceptance. It’s the symbolic anchor of the UK’s nuclear buildout programme.
What is the National Wealth Fund’s role in UK nuclear? The National Wealth Fund is a UK government-backed investment vehicle that can co-invest alongside private capital in Advanced Nuclear Framework projects. It doesn’t provide grants or guaranteed funding — it reduces first-mover risk to attract private co-investors.
Does the Genesis Mission guarantee funding for nuclear projects? No. It directs DOE attention and provides access to loan guarantee mechanisms. Loan guarantees reduce borrowing costs — they don’t replace private capital. Proposals still have to pass through DOE review and meet the programme’s eligibility criteria.
How does the PJM emergency auction relate to nuclear policy? The proposal would have PJM offer 15-year electricity contracts to incentivise new power plant construction, including nuclear. The long contract term is designed to make long-lead-time projects economically viable. However, PJM wasn’t included in the announcement, and neither the White House nor the governors can mandate the auction. It remains a proposal as of early 2026.
What is X-Energy and why does it appear in both US and UK nuclear policy? X-Energy is a US SMR developer whose Xe-100 high-temperature gas-cooled reactor has attracted interest in both markets. It has an Amazon deal in Washington State and a Centrica partnership at Hartlepool. The same vendor serving both markets means difficulties in one market affect timelines in both simultaneously.
When will UK nuclear power actually be available for AI data centres? Not before the early 2030s, based on current assessments. The Advanced Nuclear Framework creates the conditions for development — it doesn’t accelerate construction physics or regulatory review timelines. Don’t factor UK nuclear power availability into infrastructure decisions today unless your planning horizon extends to 2032 or beyond.
Is nuclear power under the Genesis Mission politically durable? The Genesis Mission is a Trump administration initiative; continuation depends on political continuity. DOE loan guarantees have been used across multiple administrations — the financing mechanisms may outlast the branding. The political energy is administration-specific and should be treated as a variable in any timeline assessment.
