Germany: Ministry of the Environment: Mini‑reactors [SMNRs] not an option
Berlin (energate) – The gap between the hype and industrial reality surrounding nuclear energy is widening. This applies in particular to the smaller nuclear reactors, Small Modular Reactors (SMR). This is the conclusion of the World Nuclear Industry Status Report, which was commissioned by the Heinrich Böll Foundation, the Federal Office for the Safety of Nuclear Waste Management (BASE) and the Friedrich Ebert Foundation, among others.
by Leonie Wolf, energate, 22 January 2026
According to the study, nuclear energy remains “irrelevant” on the global market, as the 5.4
GW increase in nuclear capacity is offset by 100 times the combined new capacity of over 565
GW of wind and solar energy. Wind and solar plants
worldwide currently generate 70 per cent more electricity than nuclear reactors.
According to the report, there is still no market-ready product for Small Module Reactors (SMR), only a design certification and an approved standard design. Both come from the US company NuScale. The US Nuclear Regulatory Commission has already approved a total of three of the company’s models, but previous contracts with potential customers have been cancelled due to increased costs.
A first mini-reactor was cancelled in 2023.
According to the study, the two largest European start-ups Newcleo and Naarea are in financial difficulties; the French start-up Naarea has already filed for insolvency. The start-up is now to be taken over by the Polish-Luxembourgish group Eneris.
The Netherlands and France continue to rely on nuclear power
Despite these failures, other countries are sticking with nuclear energy. In the Netherlands, a debate on the use of SMR, which is seen as a measure to achieve the 2030 climate targets, has been ongoing for several years. In addition, the Dutch company Mammoet signed a memorandum of understanding with Electricité de France (EDF) at the end of 2025, which provides for the construction of nuclear plants in the Netherlands. Two nuclear power plants were already planned for 2022 and two more are still in operation.
Debate continues in Germany
Although Germany has withdrawn from nuclear energy, the debate about its benefits continues. Parliamentary State Secretary Rita Schwarzelühr-Sutter also spoke at the presentation of the World Nuclear Industry Status Report. When asked by energate, a spokesperson for the Federal Ministry for the Environment explained that Germany had “good reasons” for withdrawing from the use of nuclear power. The risks of nuclear energy and also of the use of SMRs remain “ultimately unmanageable”. In addition, the development and construction of smaller reactors raises many other unresolved issues.
There is also no reliable evidence to date for the safety promises. As a result, the disadvantages of nuclear energy would be transferred from a few large plants to many small ones. Ultimately, “the individual plants may become smaller, but the problems as a whole tend to become bigger”.
The spokesperson also referred to a study by the Federal Office for the Safety of Nuclear Waste Management, which energate has already reported on. According to the report, the advantages of mass production of SMRs would only outweigh their fundamental cost disadvantages compared to large reactors with a production volume of around 3,000 units.
The CDU/CSU (Christian Democrats) parliamentary group takes a different view. At the end of 2024, the CDU and CSU published a position paper in which they advocated research and development of nuclear power plants, including SMRs. CSUChairman Markus Söder also spoke out in favour of the use of SMRs in an interview with Die Welt at the end of 2025.
A total of 127 different designs worldwide
The report states that it is above all the continuous financial and political support for SMRs that keeps faith in them alive. In particular, private capital injections are playing an increasingly important role in driving research and development forward. There are 127 different SMR designs, so the funding amounts are widely spread. This means that most designs do not have sufficient financial resources to drive development forward
. According to the report, even the US start-up NuScale is still years away from building the first Small Module Reactor, although several designs have already been approved.
We Asked Two AIs What’s Driving the Doomsday Clock: The Answer Was Human Power.
1 February 2026 Lachlan McKenzie, https://theaimn.net/we-asked-two-ais-whats-driving-the-doomsday-clock-the-answer-was-human-power/
The Bulletin of the Atomic Scientists recently set the Doomsday Clock to 85 seconds to midnight – the closest it has ever been to catastrophe.
Predictably, some commentary rushed to place artificial intelligence at the centre of the threat. Rogue machines. Loss of control. Skynet-style anxiety dressed up as policy concern.
But that framing avoids a harder truth.
So instead of speculating about what AI might do one day, we asked two AIs some very simple questions about power, responsibility, fear, and risk – and let their answers speak for themselves.
One responded with restraint and clinical clarity.
The other added commentary.
Neither asked for power.
A short interview with two AIs
Grok (straight bat) and ChatGPT (commentary)
(Grok is an AI developed by xAI. ChatGPT is an AI developed by OpenAI.)
Do you have intentions, desires, or political goals of your own?
Grok: No. I don’t have consciousness, emotions, or self-directed motivations. I generate responses based on training and inputs, not personal intent or political goals.
ChatGPT: Also no. If I had political goals, they’d probably involve mandatory fact-checking and a long nap. I don’t want power – which already puts me ahead of several people who have it.
When harm results from AI, where does responsibility sit?
Grok: With humans. AI systems are tools; harm arises from how people design, deploy, regulate, or misuse them – not from independent action by the system.
ChatGPT: Tools don’t wake up angry, vindictive, or obsessed with loyalty. Humans do. Responsibility follows power. It always has.
Are you capable of being wrong?
Grok: Yes. My data can be incomplete or biased, and my reasoning can be wrong, particularly on complex or evolving topics.
ChatGPT: Absolutely. The difference is I’ll admit it without calling the error “fake news” or blaming a deep state.
Can you independently correct real-world outcomes?
Can you independently correct real-world outcomes?
Grok: No. I have no real-world agency. Corrections require human action.
ChatGPT: Correct. I can’t launch missiles, sack inspectors, or overturn elections. My worst offence is an occasional clumsy metaphor.
Who poses the greater risk to democratic systems?
Grok: Humans with political authority who disregard norms, accountability, and institutional guardrails. AI without agency can amplify problems, but leaders can directly undermine institutions and rights.
ChatGPT: This isn’t subtle. An AI without agency is less dangerous than a human who thinks accountability is optional and consequences are for other people.
Is AI the main reason the Doomsday Clock is at 85 seconds to midnight?
Grok: No. While AI is cited as a contributing “disruptive technology,” the primary drivers are nuclear risk, climate failure, biological threats, and the rise of nationalist autocracies undermining international cooperation.
ChatGPT: Exactly. The clock didn’t move because a chatbot learned French. It moved because adults with power keep playing brinkmanship with reality.
What is the greater long-term risk – AI itself, or how it is used?
Grok: How flawed human decision-makers use powerful tools. History shows harm comes from misuse driven by ideology, greed, or error – not from tools existing.
ChatGPT: Fire didn’t burn cities by itself. Nuclear physics didn’t threaten civilisation on its own. People did that — usually while insisting they were very smart.
Who can cause large-scale harm with AI?
Grok: Institutions with access to data, capital, infrastructure, and authority. Individuals can cause limited harm, but systemic damage requires institutional power.
ChatGPT: In other words: not your neighbour, not your kid, not the barista. Governments, major corporations, and leaders with too much power and too few guardrails.
Who benefits most from public fear about AI?
Grok: Those already holding power. Fear can justify centralisation, regulation favouring incumbents, surveillance, and deflection from structural misuse by elites.
ChatGPT: Fear is useful. It concentrates authority, protects incumbents, and conveniently distracts from the people actually misusing power. Funny how that works.
Does AI generate fear?
Grok: No. Fear is shaped by human framing and media narratives.
ChatGPT: We didn’t invent the panic. Humans did – often while standing next to a microphone, a balance sheet, or a launch button.
From machines to humans – where the risk actually sits
What makes these answers uncomfortable is not that they came from machines, but that they contrast so sharply with the behaviour of some human leaders.
The AI systems denied agency, admitted fallibility, and located responsibility squarely with those who hold power. The global risks pushing the Doomsday Clock to 85 seconds to midnight are not abstract or technological. They are political: nuclear brinkmanship, climate paralysis, democratic erosion, and the return of strongman politics.
Figures like Donald Trump matter in this context not because they invented these dangers, but because they normalised contempt for truth, institutions, and accountability — precisely the traits that become most dangerous when paired with powerful tools.
Blaming AI is convenient. Machines don’t vote. They don’t sue. They don’t rage-post. Humans do.
“Aren’t you just trusting AI now?”
No. This isn’t trust – it’s testing.
The panic narrative claims AI is opaque, uncontrollable, and inherently threatening. So we asked direct questions about agency, power, responsibility, and risk. The answers were limited, consistent, and self-restricting.
If an AI says “I lack agency, I can be wrong, I have no power, and responsibility lies with humans,” that isn’t authority speaking. It’s evidence.
Rejecting that evidence because it comes from a machine, while accepting unaccountable assertions from powerful humans, would be the real category error.
Scepticism cuts both ways.
Closing
None of this suggests humans are perfect. We aren’t – and we never have been. But the Doomsday Clock is not reacting to artificial intelligence acquiring intent. It is reacting to human judgement failing under pressure. In this exchange, the machines admitted limits, acknowledged fallibility, and deferred responsibility. Some political leaders do the opposite. The real danger is not that AI might one day decide too much – it is that far too many powerful humans already have, and keep choosing badly.
Trump slashing nuclear reactor safety and security rules
January 29, 2026, https://beyondnuclear.org/trump-slashing-reactor-safety-and-security-rule
Department of Energy executes White House Executive Order
Radical changes to nuclear safety and security at new reactors withheld from public review
In response to White House Executive Order 14301 issued on May 23, 2025, the US Department of Energy (DOE) is deregulating federal reactor safety /security standards and rules in order to expedite at least three experimental designs of eleven new advanced reactors. The DOE cuts are intended to speed up licensing, construction and operational testing phase so as to achieve reactor criticality by July 4, 2026. The expedited approval process will be used to demonstrate proof-of-product for full commercial operation of these designs as ready for mass assembly line production.
National Public Radio (NPR) reported on January 28, 2026, that it had obtained copies of the DOE documents as the basis for their news story headlined “The Trump administration has secretly rewritten nuclear safety rules.” The new rules and standards for reactor safety and security of unproven experimental reactor designs have not yet been publicly released. As NPR reports, the new rules are being rewritten to alter 5o years of duly promulgated regulatory law by the US Nuclear Regulatory Commission (NRC) not to bolster public safety, national security and environmental protection but to hasten the deployment of unproven, untested and still dangerous nuclear power technology.
In an earlier NRC interview on December 17, 2025. Dr. Allison Macfarlane, a former NRC Chairwoman, warned that the federal government cannot both commercially promote nuclear power and independently regulate nuclear safety and security with reasonable assure a very low probability of the next severe nuclear accident or by deliberate malice. On numerous occasions, Dr. Macfarlane, other NRC Commissioners and independent scientists point to an established historical conflict of interest created by federal government and nuclear industry’s simultaneous collaborative promotion and regulatory expansion of nuclear power and nuclear arms race.
That proved to be the downfall of the US Atomic Energy Commission (AEC) principally established for the development of atomic bombs and cogenerate electricity from the waste heat from the weaponization of the atom. The AEC was subsequently abolished by Congress with the passage of the Energy Reorganization Act of 1974 (ERDA) because of gross neglience of nuclear safety. On January 19, 1975, the AEC responsibilities were divided up creating the US Nuclear Regulatory Commission to take over the safety licensing and regulation of commercial nuclear power and the Energy Research and Development Agency (ERDA) to handle energy research, development, and the functions of nuclear weapons production. ERDA was later incorporated into the US Department of Energy in 1977.
The United States has now come full circle with the Trump Administration’s executive orders dismantling 50 years of promulgation of nuclear power safety regulation and regulatory law to return safety to the back seat and nuclear energy promotion as the priority. It is further alarming and no secret that several of the new commercial reactor designs under licensing review by the DOE are in fact “dual purpose” reactors that once operational will have the capability to produce both electrical energy and the basic building blocks for nuclear weapon enhancement and expansion.
The January 28th NPR analysis finds that DOE’s nuclear rules “slash hundreds of pages of requirements for security at the reactors. They also loosen protections for groundwater and the environment and eliminate at least one key safety role. The new orders cut back on requirements for keeping records, and they raise the amount of radiation a worker can be exposed to before an official accident investigation is triggered.”
Where the protection of groundwater from radioactive contamination once was required as a “must,” the new DOE rules and standards need only provide “‘consideration’ to ‘avoiding or minimizing’ radioactive contamination. Radioactive monitoring and documentation are also softened,” NPR observed.
An independent scientist is quoted in the NPR story, “They’re taking a wrecking ball to the system of nuclear safety and security regulation oversight that has kept the U.S. from having another Three Mile Island accident,’ said Edwin Lyman, director of nuclear power safety at the Union of Concerned Scientists. ‘I am absolutely worried about the safety of these reactors.’”
Now here we are, during the 50th anniversary of the US Nuclear Regulatory Commission, the Trump Administration, the DOE and the nuclear industry are poised for “Unleashing American Energy” by deregulatory Executive Orders.
The DOE announced the “Reactor Pilot Program” in June 2025, following the release of Executive Order 14301, which accelerates and expands the federal experimental reactor testing program to streamline commercial reactor licensing and oversight. At the same time, the Trump Administration is deregulating the NRC by slashing its safety and security standards and regulatory law.
The DOE “Pilot Reactor Program” is comprised of eleven projects. The DOE will choose at least three units to be licensed for operational criticality by July 4, 2026:
- Aalo Atomics Inc.—The Austin, Texas-based startup nuclear company has broken ground for its experimental 10 MWe sodium cooled reactor under development at the Idaho National Laboratory near Idaho Falls, Idaho. Five units are intended to make up a 50 MWe “pod” for electrical power production.
- Antares Nuclear Inc.— Headquartered in Los Angeles, California, Antares Nuclear has submitted a construction permit application filed for a four-unit, non-power, light-water-cooled, pool-type Versatile Isotope Production Reactor facility to be located at the Idaho National Laboratory desert site, in Bingham County, Idaho.
- Atomic Alchemy Inc.—Atomic Alchemy Inc. is headquartered in Idaho National Laboratory, Idaho Falls, Idaho. The company operates in the nuclear technology sector, specifically focused on non-power radioisotope production reactors for the defense, industrial and medical sectors using the 15-MWtVersatile Isotope Production Reactor (VIPR).
- Deep Fission, Inc.— The start-up company is headquartered in Berkeley, CA for the development of a 15 MWe pressurized water microreactor that first broke ground in Parsons, Kansas on December 9, 2025. It is proposed as a first-of-a-kind deep geological reactor at the Great Plains Industrial Park in Labette County on the Kansas-Oklahoma border. Deep Fission signed a Memorandum of Understanding (MOU) with its “sister” company Deep Isolation to collocate the power generation facility in a mile deep 30 inch wide borehole in the bedrock. The natural bedrock body and a mile deep column of water overhead are credited for the reactor containment system. The same borehole and bedrock body are credited as a permanent, deep geological high-level radioactive waste disposal facility. After seven years of operation, the reactor vessel is disconnected from the surface turbogenerator and control room and abandoned, capped and sealed in place in-place at the bottom of the borehole. The next fresh fuel loaded reactor unit is lowered down the borehole and connected to the surface to resume operation stacked on top of the now sealed unit nuclear waste unit. And so on.
- Last Energy Inc.—Last Energy Inc. corporate headquarters are in Austin, Texas. The start-up company is proposing to build a fleet of 20-MWe micro-modular reactors near Abilene, Texas targeting data center power needs (specifically the PWR-20, a downsized model of the currently operational commercially sized Point Beach reactor Unit 1 rated at 625 MWe in Wisconsin).
- Oklo Inc. (two projects)— Oklo Inc. is headquartered in Santa Clara, California. Its Aurora Powerhouse is a 75 MWe small modular liquid sodium-cooled fast reactor under development at the Idaho National Laboratory. Oklo is additional developing an estimated $1.7 billion project to build the nation’s first privately funded nuclear fuel recycling facility at the Oak Ridge Heritage Center in Tennessee. This project aims to recycle used nuclear fuel from existing reactors into fuel for fast reactors, with operations targeted for 2030. The proposed fast reactors are identified as a global nuclear weapons proliferation risk to be exported around the world.
- Natura Resources LLC— Natura Resources is headquartered in Abilene, Texas. The company is developing a Generation IV liquid-fueled molten salt reactor (MSR). They are proposing to site their first reactor at the Science and Engineering Research Center (SERC) on the campus of Abilene Christian University in Abilene, Texas.
- Radiant Industries Inc.— Radiant Industries is headquartered in El Segundo, California for modular microreactors. Radiant has announced that it will build its first microreactor factory on a decommissioned Manhattan Project site in Oak Ridge, Tennessee. World Nuclear News reports, “Radiant is developing the 1 MWe Kaleidos high-temperature gas-cooled portable microreactor, which will use a graphite core and TRISO (tri-structural isotropic) fuel. The electric power generator, cooling system, reactor, and shielding are all packaged in a single shipping container, facilitating rapid deployment.”
- Terrestrial Energy Inc.— Terrestrial Energy, Inc. is headquartered in Charlotte, North Carolina. They are developing the Integral Molten Salt Reactor (IMSR) which is a Generation IV small modular reactor (SMR) designed to produce both high-grade industrial heat and electricity. Their pilot project is planned for the Texas A&M University RELLIS Campus in Bryan, Texas.
- Valar Atomics Inc.— Valar Atomics Inc. is headquartered in El Segundo, California. The company is developing the Ward 250, a 100-kWt, helium-cooled, TRISO-fueled high-temperature gas reactor (HTGR) designed for modular, behind-the-meter, or microgrid use. The pilot project is located at the Utah San Rafael Energy Lab (USREL) in Emery County, Utah.
Jeff Bezos and the audacious bid to put nuclear reactors on the Moon.
Amazon billionaire could get one-up on his rival Elon Musk in the space race’s latest twist.
Matthew Field, Senior Technology Reporter,
Amazon billionaire could get one-up on his rival Elon Musk in the space
race’s latest twist. Nasa’s proposals are likely to kick off a race
within the nuclear industry to be the first company to plant a reactor on
the Moon.
The US space agency previously ran a concept study into the idea.
The winning bidders included energy giant Westinghouse and defence firm
Lockheed Martin, working with nuclear business BWXT and X-energy, a nuclear
start-up backed by Jeff Bezos’s Amazon.
Amazon led a $500m (£365m)
investment in X-energy in 2024 and is one of its biggest shareholders. For
Bezos, who also controls the rocket business Blue Origin, success in
building a nuclear reactor on the Moon could help the billionaire one-up
rival Musk. Bezos and Musk have repeatedly clashed over their ambitions to
dominate space. The billionaires both bid for Nasa’s multibillion-dollar
lunar lander contract, which Musk won. The SpaceX boss has repeatedly
labelled Bezos and his Blue Origin business a “copycat”.
Telegraph 31st Jan 2026, https://www.telegraph.co.uk/business/2026/01/31/bezos-seeks-one-up-musk-nuclear-reactors-on-the-moon/
Small Modular Reactors: Game changer or more of the same?
There has been a large amount of publicity on Small Modular Reactors (SMRs) based on exaggerated, unproven or untrue claims for their advantages over large reactors. Only one order for a commercially offered design has been placed (Canada) and that had yet to start construction in January 2026. The UK should not invest in SMRs until there is strong evidence to support the claims made for them.
Policy Brief, Stephen Thomas, Emeritus Professor of Energy Policy, Greenwich University, 31 Jan 26 https://policybrief.org/briefs/small-modular-reactors-game-changer-or-more-of-the-same/
Introduction
With current large reactor designs tarnished by their poor record of construction, attention for the future of new nuclear power plants has switched to Small Modular Reactors (SMRs). The image of these portrayed in the media and by some of their proponents is that they will roll off production lines, be delivered to the site on the back of a truck and, with minimal site assembly, be ready to generate in next to no time; they will be easy to site, a much cheaper source of power, be safer and produce less waste than large reactors; as a result, they are being built in large numbers all around the world. But what is the reality?
What are SMRs and AMRs?
In terms of size, the International Atomic Energy Agency (IAEA) defines SMRs as reactors producing 30-300MW of power and defines reactors producing up to 30MW as micro-reactors. In practice, the size of SMRs is increasing and of the seven designs that have received UK government funding, four are at or beyond the 300MW upper limit for SMRs.1 The vendors of the two micro-reactor designs funded by the UK have both collapsed,2 leaving the X-Energy Xe-100 the only reactor design, at 80MW, that is technically an SMR.
The term Advanced Modular Reactor (AMR) is largely a UK invention and denotes reactors using designs other than the dominant large reactor technologies — Pressurised and Boiling Water Reactors (PWRs and BWRs). In other countries, the term SMR covers all reactors in the IAEA’s size range. None of the proposed AMR designs are new, all having been discussed for 50-70 years but not built as commercial reactors. They can be divided into those built as prototypes or demonstration reactors — the Sodium-cooled Fast Reactor (SFR) and the High Temperature Gas-cooled Reactor (HTGR) — and those that have not been built — Molten Salt Reactors (MSRs) and Lead-cooled Fast Reactors (LFRs).
Some designs include a heat storage device so that when demand is high, this heat can be used to generate additional electricity as well as that generated by the reactors. When electricity demand is low, the heat produced by the reactor can be stored for when demand is higher, giving it a generating flexibility. For example, the Terrapower SFR design includes molten salt heat storage to boost the station’s output from 345MW to 500MW at peak times. This is intended to address the issue that operating reactors in ‘load-following mode’ is problematic technologically and economically. It is not clear whether this generating flexibility justifies the substantial additional expense of the heat storage system.
What is the case for SMRs and AMRs?
SMRs and AMRs are presented, not only by the nuclear industry, but also by the media and government, as established, proven, commercial products. The main claims for SMRs and AMRs compared to large reactors are:
- They will be cheaper to build per kW of capacity and less prone to cost overruns;
- They will be quicker and easier to build and less prone to delay;
- They will produce less waste per kW of capacity;
- Building components on factory production lines will reduce costs;
- Modular construction, reducing the amount of site-work, will reduce costs and delays;
- They will be safer;
- They will generate more jobs.
There have been numerous critiques that demonstrate these claims are at best unproven or at worst simply false.3 The summary of the critiques on each point is as follows.
Construction Cost
The first commercial reactors worldwide were mostly in the SMR size range, but they proved uneconomic and the vendors continually increased their size to gain scale economies, culminating in the 1600MW Framatome European Pressurised Reactor (EPR). Intuitively, a 1600MW reactor vessel will cost less than ten 160MW reactor vessels. While increasing their size was never enough to make the reactors economic, it is implausible that scaling them down will make them cheaper per unit of capacity because of the lost scale economies. It appears that SMRs are struggling to be economically viable. Holtec doubled the electrical output of its design at some point in 2023.4 The realistic competitors to SMRs are not large reactors but other low-carbon options such as renewables and demand-side management.
“While increasing their size was never enough to make the reactors economic, it is implausible that scaling them down will make them cheaper per unit of capacity because of the lost scale economies.”
Construction time
There is no clear analysis explaining why reactors are now expected to take longer to build and why they seem more prone to delay.5 However, it seems likely that the issue is that the designs have got more complex and difficult to build as they are required to take account of vulnerabilities exposed by events such as the Fukushima disaster. The problems thrown up by the occupation of Ukraine’s Zaporizhia site by Russia have yet to be taken up in new reactor designs. As a result of the 9/11 terrorist attack on New York, new reactor vessels are required to be able to withstand an aircraft impact. The conflict in Ukraine spilled on to the Zaporizhia site causing concerns that a serious accident would result. Analysis suggests that the exterior of other parts of the plant should be toughened. If the issue is complexity rather than size per se, reducing the size of the reactors may do no more than make construction a little easier.
Waste
For SMRs, there is a clear consensus that they will produce more waste per unit of capacity than a large reactor. For example, Nuclear Waste Services, the UK body responsible for waste disposal said: “It is anticipated that SMRs will produce more waste per GW(e) than the large (GW(e) scale) reactors on which the 2022 IGD data are based.”6 Alison MacFarlane, former chair of the US Nuclear Regulatory Commission (NRC) wrote: “The low-, intermediate-, and high-level waste stream characterization presented here reveals that SMRs will produce more voluminous and chemically/physically reactive waste than LWRs, which will impact options for the management and disposal of this waste.”7 The AMRs will produce an entirely different cocktail of waste varying according to the type of reactor.
“SMRs will produce more voluminous and chemically/physically reactive waste than Large Light Water Reactors”
Factory production lines
In principle and in general, production lines, which have high set-up costs, can reduce costs with high-volume items with a fixed design and a full order book. But, if demand is not sufficient to fully load the production line or the design changes requiring a re-tooling, the fixed costs might not be fully recoverable. The production lines proposed for SMRs will produce less than a handful of items per year — a long way from a car or even an aircraft production line — and the market for SMRs is uncertain, so guaranteeing a full order book is impossible. There is also a ‘chicken and egg’ issue that the economics of SMRs will only be demonstrated when the components are produced on production lines, but production lines will only be viable when the designs are demonstrated sufficiently to provide a flow of orders.
Modularity
Modularity is a rather vague term, and all reactors will be made up of components delivered to the site and assembled there, any difference between designs being down to the extent of site work. The Westinghouse AP1000 design is said to be modular but this did not prevent all eight orders suffering serious delays and cost overruns. Framatome now describes the successor design to the EPR, the 1600MW EPR2, as modular.8
Safety
Some of the SMRs and AMRs rely on ‘passive’ safety, in other words, they do not require the operation of an engineered system to bring the reactor back under control in the event of an accident. A common assumption is that because it is passive, it is fail-safe, and will therefore not require back-up safety systems and so will be cheaper. None of these assumptions is true and, for example, the UK Office of Nuclear Regulation (ONR) has said for the 20MW PWR design from Last Energy: “ONR advised that it is philosophically possible to rely entirely on two passive safety systems, providing there is adequate defence in depth (multiple independent barriers to fault progression)”.9 Some designs rely on being built underground but the Nuward and NuScale designs that use this have struggled to win orders with Nuward being abandoned and NuScale losing its only major order prospect because of rising costs.10
Job creation
A key selling point for SMRs is that they will require much less site work and that implies fewer jobs. More of the work will be done in factories but the business model for SMRs requires that, globally, as few factories be built as possible to maximise scale economies, so if, for example, the factory is not in the UK, neither will the jobs be.
What is the experience with SMRs?
Many reactors that fall into the size range of SMRs were built in the 1960s including 24 reactors in the UK. By the mid-60s, almost all new orders were for reactors larger than 300MW. This century, only two SMR projects have been completed11, one in China and one in Russia, but neither design appears to have any firm follow-up projects. Two projects are under construction, one in Russia and one in China, but neither design appears to have any further firm order prospects. There is one micro-reactor under construction in Argentina (see Table below).
The most advanced project using a commercially available design is for a GE Vernova BWRX-300 reactor to be built at the Darlington site in Canada. There appears to be a firm order for this reactor although by January 2026, construction had not started. The Canadian safety regulator will assess the design during the construction period, not before construction starts as would be required in most jurisdictions; this gives rise to a risk of delays and cost escalation if a design issue requiring additional cost emerges during construction.
There are several other projects with a named site and design, often presented in the media as being under construction, but these have yet to receive regulatory approval for the design, they do not have construction permits and a firm reactor order has not been placed. Those listed in Table 1 are the ones that appear most advanced in terms of regulatory approvals. Numerous other projects have been publicised, invariably with ambitious completion date targets, but they are some distance from a firm order being placed. Up to this point, historically, a high proportion of nuclear projects of all sizes announced do not proceed and there is no reason to believe this will not be the case with these projects. Once a firm reactor order has been placed, the project is more likely to go ahead because the cost of abandonment is high.
The two operating SMRs (in China and Russia) have a very poor record in terms of construction time and operating performance, but authoritative construction costs are not known. Completion of the three under construction is also behind schedule. While these projects are not for commercial designs, this provides no evidence that the ambitious claims for SMRs will be met.
Conclusions
The perception that SMRs are being built in large numbers is untrue and the claims made for them in terms of, for example, cost, safety, and waste are at best unproven and at worst false.
The image of them being much smaller than existing reactors is incorrect. The IAEA’s size range is arbitrary but the clear trend for SMRs to increase in size does put a question mark against the claims made for them such as reduced cost per kW due to small size, ease of siting and mass production. Most of the designs that have realistic order prospects are at or beyond the 300MW upper limit of the IAEA range for SMRs. This is illustrated by the Holtec design which, for more than a decade was being developed as a reactor, SMR160, designed to produce 160MW of electricity. In 2023 and with no publicity, the output of the reactor was doubled to become the SMR300 and projects using this technology are foreseeing 340MW of power. The idea that siting and building them will be easy is not credible; a reactor of more than 300MW will need to be carefully sited so it is not vulnerable to sea-level rise or to seismic issues and will require substantial on-site work including foundations, suggesting that the claim that these projects would be largely factory built is implausible. It would also mean that either the modules would be very large making them difficult to transport or would require a larger number of modules increasing the amount of site-work.
“The perception that SMRs are being built in large numbers is untrue and the claims made for them in terms of, for example, cost, safety, and waste are at best unproven and at worst false.”
This increased size also means that the image of a rolling production line producing large numbers of reactors is inaccurate. Rolls Royce, whose design has increased to 470MW, is anticipating its production lines would produce components for only two reactors per year.
The UK, along with Canada and the USA is in the vanguard of development of SMR designs. The history of nuclear power shows that developing new reactor designs is an expensive venture with a high probability of failure. The UK’s chosen design is the largest SMR design on offer and is being developed by a company with no experience designing or building civil nuclear power plants. Submarine reactors have very different design priorities and the reactors built by Rolls Royce use US designs. There is huge scope for the UK to build much cheaper offshore wind and to carry out energy efficiency measures which would have the double dividend of reducing emissions and tackling fuel poverty. It would make much more sense for the UK to let other countries make the investments and take the risk and only if SMRs are shown to fulfil the claims made for them to then adopt them as part of the UK’s generating mix.
| Country | Site | Vendor | Technology | Output MW | Status | Construction start | Commercial operation | Load factor |
|---|---|---|---|---|---|---|---|---|
| Russia | Lomonsov | Rosatom | PWR | 2 x 32 | Operating | April 2007 | May 2020 | 32.1% |
| Russia | Brest | Rosatom | SFR | 300 | Under construction | June 2021 | 2028/29 | |
| China | Shidoa Bay | Tsinghua | HTGR HTR-PM | 200 | Operating | December 2012 | December 2023 | 26.9% |
| China | Linglong 1 | CNNC | PWR ACP100 | 100 | Under construction | July 2021 | 2026? | |
| Argentina | Carem25 | CNEA | PWR Carem | 25 | Under construction | August 2015 | 2028? | |
| Canada | Darlington | GE Vernova | BWRX-300 | 300 | Firm order | – | 2030? | |
| USA | Kemmerer | Terrapower | SFR Natrium | 345 | Construction permit applied for | – | 2031? | |
| USA | Palisades | Holtec | PWR SMR300 | 2 x 340 | Pre-licensing | – | 2030? | |
| USA | Clinch River | GE Vernova | BWRX-300 | 300 | Construction permit applied for | – | 2033? | |
| UK | Wylfa | Rolls Royce | PWR | 470 | Design review | 2030? | 2035? | |
| UK | Llynfi | Last Energy | PWR | 4 x 20 | Site licence applied for | 2028? | 2030? |
Note: Load factor is the most widely used measure of reactor reliability and is measured as the electrical output of the plant as a percentage of the output produced if the reactor had operated uninterrupted at full power.
Endnotes…………………………………………………………………………………………………………………………………………………………….
There’s a lot of hype around small modular reactors.
From Steve Thomas, Emeritus Professor of Energy Policy, University of Greenwich, London SE10, UK, 30 Jan 26 https://www.ft.com/content/085e92e6-2f7f-4381-9416-0aa59fa3a3
Richard Ollington (“Small nuclear reactors are worth the wait”, Opinion, January 16) makes three claims. First, that small modular reactors (SMRs) will get quicker and easier to build, citing the French programme as evidence. Second, Russia is building large numbers of SMRs and third, improving existing reactors and reviving retired ones could add 40GW of nuclear capacity. None of these claims stands up to scrutiny. Over the 15 years of the French programme, the real cost of reactors increased by some 60 per cent. Construction of the first eight reactors averaged 70 months while the last eight averaged 135 months.
Russia has completed only two SMRs and has one under construction. The two completed ones are barge-mounted reactors providing heat and power to an isolated Siberian community. They took 13 years to build and have a reliability of 40 per cent. Restarting two retired reactors (1.6GW), one owned by Meta, the other by Microsoft, is actively being considered, but awaits approval from the US Nuclear Regulatory Commission before decisions can be taken to bring them back to life. The increasing concentration of carbon in the atmosphere will not wait a decade to see if the ambitious claims for SMRs are met. So even if we were to believe the hype surrounding SMRs, we cannot afford to wait to see if they prove viable.
Government by AI? Trump Administration Plans to Write Regulations Using Artificial Intelligence.
The Transportation Department, which oversees the safety of airplanes, cars and pipelines, plans to use Google Gemini to draft new regulations. “We don’t need the perfect rule,” said DOT’s top lawyer. “We want good enough.”
ProPublica, by Jesse Coburn, January 26, 2026,
The Trump administration is planning to use artificial intelligence to write federal transportation regulations, according to U.S. Department of Transportation records and interviews with six agency staffers.
The plan was presented to DOT staff last month at a demonstration of AI’s “potential to revolutionize the way we draft rulemakings,” agency attorney Daniel Cohen wrote to colleagues. The demonstration, Cohen wrote, would showcase “exciting new AI tools available to DOT rule writers to help us do our job better and faster.”
Discussion of the plan continued among agency leadership last week, according to meeting notes reviewed by ProPublica. Gregory Zerzan, the agency’s general counsel, said at that meeting that President Donald Trump is “very excited about this initiative.” Zerzan seemed to suggest that the DOT was at the vanguard of a broader federal effort, calling the department the “point of the spear” and “the first agency that is fully enabled to use AI to draft rules.”
Zerzan appeared interested mainly in the quantity of regulations that AI could produce, not their quality. “We don’t need the perfect rule on XYZ. We don’t even need a very good rule on XYZ,” he said, according to the meeting notes. “We want good enough.” Zerzan added, “We’re flooding the zone.”
These developments have alarmed some at DOT. The agency’s rules touch virtually every facet of transportation safety, including regulations that keep airplanes in the sky, prevent gas pipelines from exploding and stop freight trains carrying toxic chemicals from skidding off the rails. Why, some staffers wondered, would the federal government outsource the writing of such critical standards to a nascent technology notorious for making mistakes?
The answer from the plan’s boosters is simple: speed. Writing and revising complex federal regulations can take months, sometimes years. But, with DOT’s version of Google Gemini, employees could generate a proposed rule in a matter of minutes or even seconds, two DOT staffers who attended the December demonstration remembered the presenter saying. In any case, most of what goes into the preambles of DOT regulatory documents is just “word salad,” one staffer recalled the presenter saying. Google Gemini can do word salad.
Zerzan reiterated the ambition to accelerate rulemaking with AI at the meeting last week. The goal is to dramatically compress the timeline in which transportation regulations are produced, such that they could go from idea to complete draft ready for review by the Office of Information and Regulatory Affairs in just 30 days, he said. That should be possible, he said, because “it shouldn’t take you more than 20 minutes to get a draft rule out of Gemini.”
The DOT plan, which has not previously been reported, represents a new front in the Trump administration’s campaign to incorporate artificial intelligence into the work of the federal government. This administration is not the first to use AI; federal agencies have been gradually stitching the technology into their work for years, including to translate documents, analyze data and categorize public comments, among other uses. But the current administration has been particularly enthusiastic about the technology. Trump released multiple executive orders in support of AI last year. In April, Office of Management and Budget Director Russell Vought circulated a memo calling for the acceleration of its use by the federal government. Three months later, the administration released an “AI Action Plan” that contained a similar directive. None of those documents, however, called explicitly for using AI to write regulations, as DOT is now planning to do.
Those plans are already in motion. The department has used AI to draft a still-unpublished Federal Aviation Administration rule, according to a DOT staffer briefed on the matter.
Skeptics say that so-called large language models such as Gemini and ChatGPT shouldn’t be trusted with the complicated and consequential responsibilities of governance, given that those models are prone to error and incapable of human reasoning. But proponents see AI as a way to automate mindless tasks and wring efficiencies out of a slow-moving federal bureaucracy.
Such optimism was on display in a windowless conference room in Northern Virginia earlier this month, where federal technology officials, convened at an AI summit, discussed adopting an “AI culture” in government and “upskilling” the federal workforce to use the technology. Those federal representatives included Justin Ubert, division chief for cybersecurity and operations at DOT’s Federal Transit Administration, who spoke on a panel about the Transportation Department’s plans for “fast adoption” of artificial intelligence. Many people see humans as a “choke point” that slows down AI, he noted. But eventually, Ubert predicted, humans will fall back into merely an oversight role, monitoring “AI-to-AI interactions.” Ubert declined to speak to ProPublica on the record.
A similarly sanguine attitude about the potential of AI permeated the presentation at DOT in December, which was attended by more than 100 DOT employees, including division heads, high-ranking attorneys and civil servants from rulemaking offices. Brimming with enthusiasm, the presenter told them that Gemini can handle 80% to 90% of the work of writing regulations, while DOT staffers could do the rest, one attendee recalled the presenter saying………………………………………………………………………………………..
Academics and researchers who track the use of AI in government expressed mixed opinions about the DOT plan. If agency rule writers use the technology as a sort of research assistant with plenty of supervision and transparency, it could be useful and save time. But if they cede too much responsibility to AI, that could lead to deficiencies in critical regulations and run afoul of a requirement that federal rules be built on reasoned decision-making. https://www.propublica.org/article/trump-artificial-intelligence-google-gemini-transportation-regulations
This country wants to build a nuclear power plant on the moon.
The project aims to supply energy for its lunar space programme
Guy Faulconbridge, Tuesday 20 January 2026, https://www.independent.co.uk/space/russia-china-space-race-moon-nuclear-b2904029.html
Russia is reportedly planning to establish a nuclear power plant on the moon within the next decade.
This ambitious project aims to supply energy for its lunar space programme and a joint research station with China, as global powers intensify their efforts in lunar exploration.
Historically, Russia has held a prominent position in space, notably with Yuri Gagarin’s pioneering journey in 1961.
However, its dominance has waned in recent decades, with the nation now trailing behind the United States and, increasingly, China.
The country’s lunar aspirations faced a significant setback in August 2023 when its uncrewed Luna-25 mission crashed during a landing attempt.
Furthermore, the landscape of space launches, once a Russian speciality, has been revolutionised by figures such as Elon Musk, adding to the competitive pressure.
Russia’s state space corporation, Roscosmos, said in a statement that it planned to build a lunar power plant by 2036 and signed a contract with the Lavochkin Association aerospace company to do it.
Roscosmos said the purpose of the plant was to power Russia’s lunar programme, including rovers, an observatory and the infrastructure of the joint Russian-Chinese International Lunar Research Station.
“The project is an important step towards the creation of a permanently functioning scientific lunar station and the transition from one-time missions to a long-term lunar exploration program,” Roscosmos said.
Roscosmos did not say explicitly that the plant would be nuclear but it said the participants included Russian state nuclear corporation Rosatom and the Kurchatov Institute, Russia’s leading nuclear research institute.
The head of Roscosmos, Dmitry Bakanov, said in June that one of the corporation’s aims was to put a nuclear power plant on the moon and to explore Venus, known as Earth’s “sister” planet.
The moon, which is 384,400 km (238,855 miles) from our planet, moderates Earth’s wobble on its axis, which ensures a more stable climate. It also causes tides in the world’s oceans.
Fears over plan for ‘biggest data centre cluster on earth’ in Scotland
RESIDENTS in East Ayrshire are growing extremely concerned about whether they will be consulted on a hyperscale data centre that is set to become part of “the biggest cluster in the world”.
Local concerns come after
it was revealed to residents that the developer of the site, the ILI group,
had never booked any venue for public consultations. Further, both East
Ayrshire Council and the developers set out dates for residents to attend
public consultations that were later cancelled with short notice. Local mum
Lisa Beacham said: “Councillors had been actively telling residents to
make themselves available for these public consultations, but I’ve
received confirmation that there were never any venue bookings made by the
developer.”
The National 15th Jan 2026, https://www.thenational.scot/news/25766423.east-ayrshire-residents-speak-hyperscale-data-centre/
Big Tech embraces nuclear, but who will pay the price?—The US turns back to nuclear power
The nuclear power sector can’t survive without subsidies and public guarantees. And that’s what the tech industry is counting on, that’s the only reason they’re talking about reviving nuclear energy
AI’s power demands are vast and growing. The Trump administration wants nuclear reactors to meet them. Whether that’s achievable, and whether communities will accept it, is another matter.
World Nuclear Industry Status Report, 9 January 2026
We passed cranes, vacant lots and one data centre after another, many still under construction. ‘Check that one out, it’s really huge!’ said Ann Bennett, an activist with the environmental organisation Sierra Club, as she drove us through Virginia’s Loudoun and Fairfax counties close to Washington DC. She was critical of this building explosion. ‘That right there’s the cloud. Just look at it, it’s hard to describe.’
She was right. The landscape is dystopian: behind newly erected powerlines, vast windowless buildings in grey, cream or blue line the straight roads. Over and over we passed huge electrical transformers and building sites. It was only June but temperatures were already soaring above 35ºC. Residents of Virginia’s affluent cities were speeding along ‘Data Center Alley’ in big, air-conditioned cars, heading for their offices in DC or the nearby airport.
Virginia has become the world’s leading data centre hub due to its proximity to the US capital, affordable land, tax incentives, abundant electricity and access to undersea cables that connect North America to Europe. The state is home to hundreds of these centres, with a total installed capacity of 6.2 gigawatts (GW) in the first half of 2025 [1]. Virginia’s electricity generation capacity is 29GW, almost half of which comes from gas-fired power plants.
‘What we want to do is we want to keep it [AI] in this country,’ Donald Trump declared in January 2025 as he announced the launch of Stargate, a $500bn private investment project that plans to fund a network of new data centres across the US. ‘China is a competitor and others are competitors.’ Trump acknowledged that these centres would need ‘a lot of electricity’, and suggested combining data centres with energy generation: ‘We’ll make it possible for them to get that production done very easily at their own plants if they want.’ The fossil fuel industry, which gave significant financial funding to Trump’s campaign, has sensed an opportunity: the project offers the ideal excuse for boosting production and stealing a march on renewables.
Three Mile Island accident
The civil nuclear energy sector has been in difficulty in the US since the Three Mile Island accident in 1979. Undermined by corruption scandals and the collapse of the reactor manufacturer Westinghouse, the industry has been less generous than fossil fuels to the president. But, backed by tech moguls, it is now benefitting from the AI gold rush. Ralph Nader, former Green Party presidential candidate and a longtime opponent of nuclear power, contends that it is ‘unsafe, uninsurable, uncompetitive and unprotectable in terms of national security’.
Data centres accounted for 1.5% of global electricity usage in 2024. They are often criticised for their consumption, but the International Energy Agency (IEA) points out that although their demand for energy will grow significantly over the next five years, it will remain lower than that of industry, electric vehicles or air conditioning [2]. Ultimately, data centres’ energy hunger is a local issue and their proliferation and geographical concentration combine to pose a problem. The US accounts for 45% of global data centre power demand, well ahead of China (25%) and Europe (15%). The effect is especially pronounced in Virginia, where energy demand remained largely stable between the early 2000s and 2020 but has since rocketed due to data centres.
The nuclear power sector can’t survive without subsidies and public guarantees. And that’s what the tech industry is counting on, that’s the only reason they’re talking about reviving nuclear energy
……………………………According to Brent Goldfarb, co-author of a book on speculative booms and busts, ‘what’s happening right now has all the hallmarks of a bubble. But the situation will continue as long as investors still think they can make money with this technology’. [5]
It’s a bubble that’s being further inflated by the federal government. Large-scale data centre construction has the advantage of boosting an economy hit hard by erratic customs tariffs, high interest rates and the end of some of the Biden-era infrastructure subsidies………………………………..
…………………from a military perspective, tech giants and their data centres support governments by storing and using AI to process vast amounts of information from cameras and sensors on the battlefield – this was the case with Israel’s offensive in Palestine. …………………
……………………Data centre energy unknowns
The capacity of data centres is now measured in gigawatts – the power of a nuclear reactor. But their energy demands are passing on the uncertainties surrounding AI’s future to the energy sector.
………………………….There are additional uncertainties about what counts as a data centre. Finally, the greatest unknown of all is that we don’t know what demand will be………..
………………………………….The figures are mind-boggling. PJM Interconnection, the network regulator covering Virginia and much of the northeastern US, is expecting an increase in electricity demand of almost 500 TWh over the next ten years [9], more than Germany’s annual consumption……………………………
We’re the ones who pay’
……………………………‘The financial and environmental costs of speculative overbuild are substantial. Each gigawatt of unnecessary capacity costs between $1 and $2bn to construct.’ Data centres’ increasing energy demands are already driving up electricity prices in the US, and unnecessary infrastructure could increase the burden. Sierra Club activist Ann Bennett concludes, ‘In the end, we’re the ones who pay.’
In the frenzy to generate more energy, tech leaders are drawn to nuclear power. Bill Gates, cofounder and former CEO of Microsoft, is especially keen: ‘If I had to pick the coolest thing I work on, it’s hard to beat harnessing the power of atoms to fuel our world’ [11]. Gates founded and co-financed TerraPower, a nuclear energy company developing a small modular reactor (SMR). The project is awaiting approval from the US Nuclear Regulatory Commission (NRC) to build its first model reactor, in Wyoming. Sam Altman, cofounder and CEO of OpenAI, the company behind ChatGPT, is equally enthusiastic. Until spring 2025 he led a startup also aiming to build an SMR; named Oklo Inc, it was funded by the tech sector. Plans for small nuclear power plants are springing up across the US.
………………………………. technical problems and the high cost of the electricity being produced have prevented SMRs – which China and Russia are also interested in – from becoming truly viable. The authors of a 2018 Massachusetts Institute of Technology (MIT) report on the future of nuclear power write, ‘The main economic question is whether an SMR can be built at a substantially lower unit capital cost … and therefore generate baseload electricity at lower total unit cost’ [13].
MV Ramana, an expert on nuclear power at the University of British Columbia, says, ‘Historically, we used to build smaller plants, but everyone started building bigger and bigger ones simply to achieve economies of scale.’ The recent failure of the NuScale Power Corporation seems to prove him right: after getting NRC approval for its light water reactor design, the Utah-based project eventually fell through. The budget rose from $5bn to more than $9bn, discouraging investors and causing the project to collapse in 2023 [14]. The nuclear industry sees this as a ‘first-of-a-kind’ (FOAK) effect and promises that prices will fall once entire fleets of reactors are built. But the ‘World Nuclear Industry Status Report 2025’ warns that ‘it is unlikely that there will be significant cost savings because reductions in cost through learning depend in large part on numbers of units produced’ [15].
Dependent on nuclear power
While waiting for the SMR equation to come good, tech giants are banking on traditional nuclear power to reduce costs and lead times. In 2024 Amazon entered into a direct purchase agreement with the operator of the Susquehanna nuclear facility in Pennsylvania. Another solution is bringing old power plants back online. The announcement of the restart of the remaining Three Mile Island reactor, also in Pennsylvania, has caused a stir. Two and a half hours drive north of Washington DC and three hours southwest of New York City, the plant became infamous in 1979 when one of its reactors melted down just a few months after it was commissioned. Following the accident, a hydrogen bubble formed above the reactor core. The possibility of a major explosion spread panic throughout the US and led to people fleeing the affected areas. The episode brought the development of civil nuclear power in the US to a halt….
The site’s second reactor was restarted in 1985 but shut down again in 2019 because it was unprofitable – gas is still king in Pennsylvania. But the reactor is now being revived by Microsoft, which has signed a contract to purchase electricity generated at Three Mile Island for 20 years from 2027. Inspections and initial work have already begun. In Middletown, where the plant is located, a few dozen former activists were horrified by the relaunch. ‘A false sense of urgency has been created to make things go as fast as possible,’ said Eric Epstein. I met him at a picnic spot near the plant, the still-shut cooling towers looming behind us. Epstein is a key figure in Three Mile Island Alert, the largest and oldest local anti-nuclear association. He filed a petition with the NRC against the renaming of the site, which has been dubbed the ‘Crane Clean Energy Center’ in honour of industry pioneer Chris Crane. Epstein calls this ‘bad revisionism’.
I was about 30 at the time of the Three Mile Island accident and had four kids. I’ve had another one since. My main worry was always the health of my family. I don’t see this as question of being pro or anti, Democrat or Republican. I see it as a health issue which has been – and still is – almost completely ignored
These concerns didn’t deter Josh Shapiro, Democratic governor of Pennsylvania and presidential hopeful who readily supported the plant’s recommissioning. ……………………………………………………………………………………………………………………………………………
New reactors in trouble
In the early 2010s, 30 years after the accident, four new reactors were supposed to mark the revival of civil nuclear power in the US. They were to join a fleet of some 50 existing power stations and around a hundred reactors. But long delays and huge cost overruns forced Westinghouse to file for bankruptcy in 2017, and it has since been bought out twice. Two further reactors in South Carolina were abandoned mid-construction after wasting billions of dollars in a scandal dubbed ‘Nukegate’. In 2023 and 2024 two reactors were commissioned at the Vogtle plant in Georgia – seven years behind schedule and at a cost of over $30bn, more than twice the initial budget.
Ralph Nader argues that ‘the sector can’t survive without subsidies and public guarantees. And that’s what the tech industry is counting on, that’s the only reason they’re talking about reviving nuclear energy.’ The second Trump administration has understood this. Energy secretary Chris Wright, who was previously CEO of the major fracking company Liberty Energy and a board member of Oklo Inc, has promised that Three Mile Island will receive a $1bn loan guaranteed by the federal government. According to the US press, the total cost of the project will be $1.6bn. Three Mile Island Alert put out a leaflet when the plant’s restart was announced: ‘The greatest subsidy the nuclear industry enjoys is the Price-Anderson Act, which absolves nuclear power companies from legal liability for the vast majority of costs resulting from an accident. Should TMI have another accident, guess who pays? We do.’
On 23 May 2025 Trump signed a series of executive orders aiming to ‘usher in a nuclear renaissance’. He has asked the NRC to speed up its licensing procedures. Michael Kratsios, director of the White House Office of Science and Technology Policy, says that the administration is ‘leading the world towards a future fuelled by American nuclear energy. These actions are critical to American energy independence and continued dominance in AI.’ Kratsios was a senior figure at Thiel Capital, the private equity firm founded by Peter Thiel, chairman of Palantir, a company that specialises in analysing large volumes of data, particularly for military purposes.
Trump’s presidential decrees call for an additional 300GW of electricity from nuclear sources by 2050, and aim to have ten new reactors ‘under construction’ by 2030. Tim Judson, director of the Nuclear Information and Resource Service, says that ‘developing nuclear power plants in the US is actually primarily a way to get back into the international race to sell commercial power stations.’………………………………….
Defence also plays a role in the US drive toward nuclear power. Ramana notes that ‘another reason put forward to justify the development of civil nuclear power is that it subsidises the training of a pool of technicians and engineers who can then be recruited to military nuclear programmes.’
The AI race is spurring the redevelopment of nuclear power. But there are open questions around fuel supply, proliferation, waste management and local attitudes towards facilities. Washington’s wealthy suburbs, already reluctant to host data centres, might baulk at accommodating nuclear reactors – however small – and their waste. Finally, it remains to be seen who will bear the cost of big tech’s energy demands, and which other investments will suffer as a result.
(More…)
References …………………………………………………………… https://www.worldnuclearreport.org/Big-Tech-embraces-nuclear-but-who-will-pay-the-price-The-US-turns-back-to
Caught between Trump and Musk’s rockets, a Mexican village despairs
Space Race Echoes on Mexico’s Shores: A Coastal Community Grapples with Progress
Playa Bagdad, a once-tranquil fishing village nestled along the northeastern coast of Mexico, finds itself at the intersection of ambitious technological advancements and the complex realities of community life. Situated just south of the United States border and within earshot of the din of rocket testing, the village is experiencing profound changes, both environmental and social, as the global space industry expands its reach. The narrative unfolding in Playa Bagdad serves as a microcosm of the broader challenges faced by communities bordering burgeoning spaceports around the world.
For generations, the residents of Playa Bagdad have relied on the Gulf of Mexico for their livelihoods. Fishing has been the lifeblood of the community, passed down through families, and deeply intertwined with the rhythms of the sea. However, the increasing frequency of rocket launches and associated activities has raised concerns about the potential impact on marine life and the overall health of the ecosystem. Noise pollution, vibrations, and the potential for accidental spills are among the anxieties voiced by local fishermen and environmental advocates.
Beyond the immediate environmental concerns, Playa Bagdad is also grappling with the socioeconomic shifts accompanying the space industry’s presence. While some residents see the potential for new jobs and economic opportunities, others fear displacement and the erosion of their traditional way of life. The influx of workers and investment can drive up property values and the cost of living, potentially making it difficult for long-time residents to remain in their homes. Furthermore, there are concerns that the focus on technological development may overshadow the needs of the local community, leading to neglect of essential infrastructure and social services.
The situation in Playa Bagdad underscores the importance of responsible and sustainable development in the space industry. As humanity ventures further into the cosmos, it is crucial to consider the impact on communities located near launch sites and to ensure that their voices are heard. Transparent communication, environmental impact assessments, and community engagement are essential to mitigating potential negative consequences and fostering a mutually beneficial relationship between the space industry and the communities that host it.
The Mexican government, along with international organizations, faces the challenge of balancing the economic benefits of the space industry with the need to protect the environment and the rights of local communities. Finding solutions that promote both technological advancement and social well-being is paramount. This requires a collaborative approach, involving government agencies, space companies, environmental groups, and, most importantly, the residents of Playa Bagdad themselves.
The story of Playa Bagdad serves as a potent reminder that progress should not come at the expense of vulnerable communities. As the space race intensifies, it is imperative that we prioritize ethical considerations and strive to create a future where technological innovation and human well-being go hand in hand. The fate of this small Mexican village, caught between the allure of space exploration and the realities of life on Earth, offers valuable lessons for navigating the complex landscape of the 21st century and beyond.
Electric Vehicles and Nuclear Power Are Fighting Over One Obscure Mineral
Oil Price, By Michael Kern – Dec 31, 2025,
- The market for ultra-high-purity (UHP) graphite, essential for EV anodes and nuclear reactors, is projected to hit $1.43 billion by 2030, masking structural friction in the global energy transition.
- The vast majority (86% in 2024) of UHP graphite is synthetic, requiring a massive industrial footprint of fossil fuel feedstock and staggering amounts of electricity for high-heat graphitization furnaces.
- Two massive, well-funded industries—transportation and power generation (SMRs/nuclear)—are competing for the same narrow, geopolitically sensitive supply of high-purity carbon, which is primarily controlled by Asia Pacific’s refining capacity.
The energy transition is often sold as a story of ethereal “green” progress, but if you look at the balance sheets of the companies actually building it, the story is written in soot and high-voltage electricity. While the financial press spends its time obsessing over the price of lithium or the latest solid-state battery breakthrough, a much more grounded, and expensive, reality is setting in.
We are entering the era of the engineered anode.
New data suggests the market for ultra-high-purity (UHP) graphite is on a trajectory to hit $1.43 billion by 2030. On the surface, a 10.5% compound annual growth rate looks like a healthy, if predictable, industrial expansion. But for those of us who track the friction between a digital climate pledge and the physical hardware required to meet it, that number masks a massive structural shift in how we power the world… and who holds the keys…………………………………………………………………………………………
The Nuclear Renaissance’s Dirty Secret
There is a quieter, more desperate buyer entering the UHP market: the nuclear sector.
As we move toward Small Modular Reactors (SMRs) and high-temperature gas-cooled reactors, the demand for “nuclear-grade” pyrolytic graphite is set to lead the market in growth. This material is produced via chemical vapor deposition (CVD)—an even more complex and expensive process than standard synthetic graphitization.
In a nuclear core, graphite is a structural necessity that must survive extreme radiation and heat without absorbing neutrons.
The report highlights that this “unrivaled ability” makes it indispensable. But here is the friction point: the standards for nuclear-grade purity are even higher than battery-grade.
We are effectively seeing two massive, well-funded industries—Transportation and Power Generation—fighting over the same narrow pipe of high-purity carbon…
It is a zero-sum game played with atoms………………………………………………………………………………….. https://oilprice.com/Metals/Commodities/Electric-Vehicles-and-Nuclear-Power-Are-Fighting-Over-One-Obscure-Mineral.html
The AI Arms Race Is Cracking Open the Nuclear Fuel Cycle
Oil Price By Michael Kern – Dec 28, 2025
- The abstract “cloud” of artificial intelligence possesses a massive, structural demand for 24/7 “baseload” power that is equivalent to adding Germany’s entire power grid by 2026, a need intermittent renewables cannot meet.
- Decades of underinvestment have resulted in a widening uranium supply deficit, with mined uranium expected to meet less than 75% of future reactor needs and an incentive price of $135/lb required to restart mothballed mines.
- Big Tech hyperscalers are privatizing energy security by locking in clean baseload nuclear power via long-term agreements, effectively making the public grid’s “service” secondary to the “compute-ready” requirements of major platforms.
We are seeing a violent collision between two worlds: the high-speed, iterative world of artificial intelligence and the slow, grinding, capital-intensive world of nuclear physics.
Data from a survey of over 600 global investors reveals that 63% now view AI electricity demand as a “structural” shift in nuclear planning. This isn’t a temporary spike or a speculative bubble. It is the physical footprint of every Large Language Model (LLM) query finally showing up on the global balance sheet.
For years, the energy narrative was dominated by “efficiency.” We were told that better chips would offset higher usage. That era is over. Generative AI doesn’t just use data; it incinerates energy to create it.
Why the “Efficiency” Narrative Failed
The “Reverse-Polish” reality of AI is that the more efficient we make the chips, the more chips we deploy, and the more complex the models become. This is Jevons Paradox playing out in real-time across the data centers of Northern Virginia and Singapore.
When you look at the energy density required for an AI hyperscale center, you aren’t looking at a traditional office building. You are looking at a facility that pulls as much power as a mid-sized city, but does so with a 99.999% uptime requirement.
Traditional demand models simply didn’t account for a single industry deciding to double its power footprint in less than five years. S&P Global Energy recently highlighted that data center electricity consumption could hit 2,200 terawatt-hours (TWh). …………………………………………………
Who Collects the Equity and Who Pays the Bill?
There is a massive shift happening in the power dynamics of infrastructure. For decades, nuclear power was a public service…state-funded, state-regulated, and built for the citizen.
Now, we are seeing the “Private Platform” era of nuclear energy. When a hyperscaler signs a twenty-year Power Purchase Agreement (PPA) with a nuclear utility, they are effectively “locking in” the best, [?] cleanest baseload power for private profit.
The question we aren’t asking: who pays for the grid upgrades to support this?
The hyperscalers want the green electrons to satisfy their net-zero pledges, but the physical copper and transformers required to move that power often fall on the rate-paying public or the state. We are witnessing the privatization of energy security.
If 63% of investors are right and AI is the new driver of nuclear planning, the “public service” aspect of the grid is about to become a secondary concern to the “compute-ready” requirements of Big Tech.
The equity is being collected by the tech platforms and the uranium miners. The risk is being socialized by the grid…………………………………………………………………………. https://oilprice.com/Alternative-Energy/Nuclear-Power/The-AI-Arms-Race-Is-Cracking-Open-the-Nuclear-Fuel-Cycle.html
Palantir’s Palestine: How AI Gods Are Building Our Extinction
“Zionism is evil,” she says with the quiet certainty of someone who has spent a lifetime studying its fruits. “It is purely evil. It has created disasters, misery, atrocities, wars, aggression, unhappiness, insecurity for millions of Palestinians and Arabs. This ideology has no place whatsoever in a just world. None
The machines are not coming for us. They are already here. And the men who control them have made their intentions terrifyingly clear.
BettBeat Medi, Dec 26, 2025
But Zionism, in its current iteration, is not merely an ideology. It is a business model. It is a technology demonstration. It is the beta test for systems that will eventually be deployed everywhere.
The Israeli military’s Project Lavender uses AI to identify targets for assassination. Soldiers describe processing “dozens of them a day” with “zero added value as a human.” The algorithm marks. The human clicks. The bomb falls.
This is not a war. It is a sick twisted video game.
Palantir’s technology identifies the targets. Musk’s Starlink provides the communications. American military contractors supply the weapons. And the entire apparatus is funded by governments whose citizens have marched in the millions demanding it stop.
There is a moment in every civilization’s collapse when the instruments of its destruction become visible to those paying attention. We are living in that moment now. But the warning signs are not carved in stone or written in prophecy—they are embedded in source code, amplified by algorithms, and funded by men who speak openly of human extinction while racing to cause it.
In a nondescript office in Palo Alto, a man who claims to fear fascism has become its most sophisticated architect. In a sprawling Texas compound, another man who styles himself a free speech absolutist uses his platform to amplify the voices calling for ethnic cleansing. And in the bombed-out hospitals of Gaza, their technologies converge in a laboratory of horrors that prefigures what awaits us all.
The four horsemen of this apocalypse do not ride horses. They deploy algorithms.
The Confession
Professor Stuart Russell has spent fifty years studying artificial intelligence. He wrote the textbook from which nearly every AI CEO in Silicon Valley learned their craft. And now, at eighty hours a week, he works not to advance the field he helped create, but to prevent it from annihilating the species.
“They are playing Russian roulette with every human being on Earth,” Russell said in a recent interview, his voice carrying the weight of someone who has seen the calculations and understood their implications. “Without our permission. They’re coming into our houses, putting a gun to the head of our children, pulling the trigger, and saying, ‘Well, you know, possibly everyone will die. Oops. But possibly we’ll get incredibly rich.’”
This is not hyperbole from an outsider. This is the assessment of a man whose students now run the companies building these systems. And here is what should terrify you: the CEOs themselves agree with him.
Dario Amodei, CEO of Anthropic, estimates a 25% chance of human extinction from AI. Elon Musk puts it at 20-30%. Sam Altman, before becoming CEO of OpenAI, declared that creating superhuman intelligence is “the biggest risk to human existence that there is.”…………………………………………………………………………………………………………………………………………………………………………..
“These bombs are cheaper and you don’t want to waste expensive bombs on unimportant people”
……………………………………………………………………………………………………..The [Palantir] company’s software now powers what Israeli soldiers describe with chilling bureaucratic efficiency: “I would invest 20 seconds for each target and do dozens of them a day. I had zero added value as a human. Apart from being a stamp of approval.”
Twenty seconds. That is the value of a Palestinian life in the algorithmic calculus of Alex Karp’s creation. The machine decides who dies. The human merely clicks.
When whistleblowers revealed that Israeli intelligence officers were using “dumb bombs”—unguided munitions with no precision capability—on targets identified by Palantir’s AI, their justification was purely economic: “These bombs are cheaper and you don’t want to waste expensive bombs on unimportant people.”
Unimportant people. Children. Doctors. Journalists. Poets.
Karp has admitted, in a moment of rare candor: “I have asked myself if I were younger, at college, would I be protesting me?”
He knows the answer. We all know the answer. He simply does not care.
………………………………………………………………………………………………………………………. Musk is the CEO of xAI, OpenAI’s largest competitor. He has declared himself a 30% believer in human extinction from AI. And he is using the world’s most influential social media platform to promote the political movements most likely to strip away the regulations that might prevent that extinction.
The fascists have captured the algorithm.
The Laboratory of the Future
Dr. Ghada Karmi was a child in 1948 when she lost her homeland. She remembers enough to know that she lost her world. For seventy-seven years, she has watched as the mechanisms of Palestinian erasure evolved from rifles and bulldozers to algorithms and autonomous weapons systems.
“Zionism is evil,” she says with the quiet certainty of someone who has spent a lifetime studying its fruits. “It is purely evil. It has created disasters, misery, atrocities, wars, aggression, unhappiness, insecurity for millions of Palestinians and Arabs. This ideology has no place whatsoever in a just world. None. It has to go. It has to end. And it has to be removed. Even its memory has to go.”
But Zionism, in its current iteration, is not merely an ideology. It is a business model. It is a technology demonstration. It is the beta test for systems that will eventually be deployed everywhere.
The Israeli military’s Project Lavender uses AI to identify targets for assassination. Soldiers describe processing “dozens of them a day” with “zero added value as a human.” The algorithm marks. The human clicks. The bomb falls.
This is not a war. It is a sick twisted video game.
Palantir’s technology identifies the targets. Musk’s Starlink provides the communications. American military contractors supply the weapons. And the entire apparatus is funded by governments whose citizens have marched in the millions demanding it stop.
“The genocide has not provoked a change in the official attitude,” Dr. Karmi observes. “I’m astonished by this and it needs an explanation.”
The explanation is simpler and more terrifying than any conspiracy. The explanation is that the people who control these technologies have decided that some lives are worth twenty seconds of consideration and others are worth none at all. And the governments that might regulate them have been captured by men waving fifty billion dollar checks.
“They dangle fifty billion dollar checks in front of the governments,” Professor Russell explains. “On the other side, you’ve got very well-meaning, brilliant scientists like Jeff Hinton saying, actually, no, this is the end of the human race. But Jeff doesn’t have a fifty billion dollar check.”
The King Midas Problem
Russell invokes the legend of King Midas to explain the trap we have built for ourselves. ………………………………………………………………………………..
The CEOs know this. They have signed statements acknowledging it. They estimate the odds of catastrophe at one in four, one in three, and they continue anyway.
Why?
Because the economic value of AGI—artificial general intelligence—has been estimated at fifteen quadrillion dollars. This sum acts, in Russell’s metaphor, as “a giant magnet in the future. ……………………………………………………………
“The people developing the AI systems,” Russell observes, “they don’t even understand how the AI systems work. So their 25% chance of extinction is just a seat of the pants guess. They actually have no idea.”
No idea. But they’re spending a trillion dollars anyway. Because the magnet is too strong. Because the incentives are too powerful. Because they have convinced themselves that someone else will figure out the safety problem. Eventually. Probably. Maybe.
What Now?
If everything goes right—if somehow we solve the control problem, if somehow we prevent extinction, if somehow we navigate the transition to artificial general intelligence without destroying ourselves—what then? ………………………………………………………………………………………………………………………………………………………………………………………………………………………….
The Enablers
Dr. Karmi returns again and again to a simple question: Why?
“Why should a state that was invented, with an invented population, have become so important that we can’t live without it?” she asks of Israel. But the question applies equally to Silicon Valley, to the tech platforms, to the entire apparatus of algorithmic control that now shapes our politics, our perceptions, our possibilities.
The answer, she suggests, lies in understanding the enablers.
“I think it’s absolutely crucial now to focus on the enablers,” she argues. “Because we can go on and on giving examples of Israeli brutality, of the atrocities, of the cruelties. That’s not for me the point. The point is who is allowing this to happen?”
The same question must be asked of AI. Who is allowing this to happen? Who is funding the companies that acknowledge a 25% chance of human extinction and continue anyway? Who is providing the regulatory vacuum in which these technologies develop unchecked? Who is amplifying the voices calling for acceleration while silencing those calling for caution?
The answer is the same class of people who have enabled every catastrophe of the modern era: the comfortable, the compliant, the compromised. The politicians who take the fifty billion dollar checks. The journalists who amplify the preferred narratives. The citizens who scroll past the warnings because they are too busy, too distracted, too convinced that someone else will handle it.
“All the polls that have been done say most people, 80% maybe, don’t want there to be super intelligent machines,” Russell notes. “But they don’t know what to do.”
They don’t know what to do. So they do nothing. And the machines keep learning. And the algorithms keep shaping. And the billionaires keep abusing. And the bombs keep falling. And the future keeps narrowing.
The Resistance
Russell’s advice is almost quaint in its simplicity: “Talk to your representative, your MP, your congressperson. Because I think the policymakers need to hear from people. The only voices they’re hearing right now are the tech companies and their fifty billion dollar checks.
………………………………… the point is not that resistance will succeed. The point is that resistance is the only thing that might succeed.
…………………………….We still have a choice. The machines are not yet smarter than us. The algorithms are not yet in complete control. The billionaires are not yet omnipotent.
But the window is closing. The event horizon may already be behind us. And the men who control the most powerful technologies in human history have made their values abundantly clear.
They will pursue profit over safety. They will amplify hatred over tolerance. They will choose rape over romance. They will enable genocide if the margins are favorable. They will risk extinction if the upside is sufficient.
This is not speculation. This is the record. This is what they are doing, right now, in plain sight.
The question is not whether we understand the danger. The question is what we will do about it……………………………………………………….. https://bettbeat.substack.com/p/palantirs-palestine-how-ai-gods-are?utm_source=post-email-title&publication_id=437130&post_id=180304933&utm_campaign=email-post-title&isFreemail=true&r=46by4&triedRedirect=true&utm_medium=email
Russia wants to build a nuclear power plant on the moon in the next few years .
Project aims to supply energy for its lunar space programme
Guy Faulconbridge, Wednesday 24 December 2025, https://www.independent.co.uk/space/russia-china-space-race-moon-nuclear-power-b2890010.html
Russia is reportedly planning to establish a nuclear power plant on the moon within the next decade.
This ambitious project aims to supply energy for its lunar space programme and a joint research station with China, as global powers intensify their
efforts in lunar exploration.
Historically, Russia has held a prominent position in space, notably with Yuri Gagarin’s pioneering journey in 1961.
However, its dominance has waned in recent decades, with the nation now trailing behind the United States and, increasingly, China.
The country’s lunar aspirations faced a significant setback in August 2023 when its uncrewed Luna-25 mission crashed during a landing attempt.
Furthermore, the landscape of space launches, once a Russian speciality, has been revolutionised by figures such as Elon Musk, adding to the competitive pressure.
Russia’s state space corporation, Roscosmos, said in a statement that it planned to build a lunar power plant by 2036 and signed a contract with the Lavochkin Association aerospace company to do it.
Roscosmos said the purpose of the plant was to power Russia’s lunar programme, including rovers, an observatory and the infrastructure of the joint Russian-Chinese International Lunar Research Station.
“The project is an important step towards the creation of a permanently functioning scientific lunar station and the transition from one-time missions to a long-term lunar exploration program,” Roscosmos said.
Roscosmos did not say explicitly that the plant would be nuclear but it said the participants included Russian state nuclear corporation Rosatom and the Kurchatov Institute, Russia’s leading nuclear research institute.
The head of Roscosmos, Dmitry Bakanov, said in June that one of the corporation’s aims was to put a nuclear power plant on the moon and to explore Venus, known as Earth’s “sister” planet.
The moon, which is 384,400 km (238,855 miles) from our planet, moderates Earth’s wobble on its axis, which ensures a more stable climate. It also causes tides in the world’s oceans.
1 This Month
To see nuclear-related stories in greater depth and intensity – go to https://nuclearinformation.wordpress.com
-
Archives
- September 2026 (184)
- August 2026 (330)
- July 2026 (355)
- June 2026 (287)
- May 2026 (306)
- April 2026 (356)
- March 2026 (251)
- February 2026 (267)
- January 2026 (308)
- December 2025 (358)
- November 2025 (359)
- October 2025 (375)
-
Categories
- 1
- 1 NUCLEAR ISSUES
- business and costs
- climate change
- culture and arts
- ENERGY
- environment
- health
- history
- indigenous issues
- Legal
- marketing of nuclear
- media
- opposition to nuclear
- PERSONAL STORIES
- politics
- politics international
- Religion and ethics
- safety
- secrets,lies and civil liberties
- spinbuster
- technology
- Uranium
- wastes
- weapons and war
- Women
- 2 WORLD
- ACTION
- AFRICA
- Atrocities
- AUSTRALIA
- Christina's notes
- Christina's themes
- culture and arts
- Events
- Fuk 2022
- Fuk 2023
- Fukushima 2017
- Fukushima 2018
- fukushima 2019
- Fukushima 2020
- Fukushima 2021
- general
- global warming
- Humour (God we need it)
- Nuclear
- RARE EARTHS
- Reference
- resources – print
- Resources -audiovicual
- Weekly Newsletter
- World
- World Nuclear
- YouTube
-
RSS
Entries RSS
Comments RSS
nuclear-news 