nuclear-news

The News That Matters about the Nuclear Industry Fukushima Chernobyl Mayak Three Mile Island Atomic Testing Radiation Isotope

Repeat After Media: Data Centers Are Wonderful and Good

It’s true that in a digital world, there will always be data centers. The question is whether they need to expand at a breakneck pace, largely to serve the needs of AI. Voters are rightly asking whether the environmental impacts of AI are worth it—for a technology with the potential to take people’s jobs and flood the news ecosystem with fake imagery.

Ari Paul, 11 Sept 26, https://fair.org/home/repeat-after-media-data-centers-are-wonderful-and-good/

If you follow conservative media and Republican politicians, you could be convinced that any opposition to AI data centers is not due to their sprawling footprint, greed for electricity or insatiable need for cooling water. Rather, it’s the result of a Chinese Communist plot.

These data centers are dreadfully unpopular. “Seven in 10 Americans oppose constructing data centers for artificial intelligence in their local area, including nearly half, 48%, who are strongly opposed,” according to Gallup (5/13/26). From YouGov (9/1/26): “About half (51%) of Americans say that the construction of new data centers is a bad thing for the country. Only 20% say it is a good thing.”

The reasons for this opposition is that these centers put enormous strain on the electrical grid, need loads of water for cooling, can be extremely noisy and take up lots of space in communities. “The environmental footprint of data centers already rivals some of the world’s largest countries,” AP (6/3/26) reported.

Opposition from ‘Chinese entities’

Much like how some liberals have convinced themselves that MAGA and American nationalism are fueled by Kremlin disinformation, the right-wing line is that communities opposing these giant data centers are motivated not by their own interests or common sense, but some unseen Beijing-manufactured psychosis.

Sen. Tom Cotton (R-AK) demanded that the Justice Department “investigate foreign influence efforts targeting the buildout of American AI infrastructure,” as he suspected that a “network of foreign actors, led by the Chinese Communist Party (CCP), is attempting to manipulate US policy and public opinion on data centers.”

Sen. Ted Cruz (R-Texas), in the words of Newsmax (9/2/26), believes “China is helping fuel opposition to artificial intelligence data centers in the United States.” “Beijing has a strategic interest in convincing Americans to slow construction of the infrastructure needed to compete in the global AI race,” the far-right outlet paraphrased.

The Washington Examiner‘s Facebook page (8/30/26) featured a cartoon from A.F. Branco depicting a Chinese data center running a psy-op on Americans who oppose data centers.

The National Interest (9/1/26) said:

Chinese entities have a long history of collaborating with far-left causes, and they have again found common ground in opposing the construction of data centers. They are pushing mutually reinforcing messages that criticize a host of alleged environmental and ethical concerns as reasons to disempower American growth in a critical economic and competitive sector. Officials at the federal, state and local levels should recognize this noxious alliance as they field criticism of data centers.

‘They want to be backwards and poor’

The rest of the corporate media don’t insist that Chinese mind control is the reason for data center opposition. But these outlets attack citizens for their ecological concerns nonetheless.

FAIR (11/20/25, 7/9/26) has covered the Washington Post’s unbridled enthusiasm for data centers and AI. It’s not hard to see why the paper is so interested in this: The Post is owned by Jeff Bezos, founder of Amazon, which has 900 data centers, according to Bloomberg (11/24/25). The Post editorial board (8/20/26) denounced opposition to data centers, saying AI “can make [people] wealthier and more productive.” “Data centers need a better plan for negotiating with local governments,” the editorialists allowed, “but the frothy debate about the economics of AI also needs a reality check.”

For the Wall Street Journal editorial board (8/31/26), the voting public on both sides of the political aisle have lost their minds. “We’re running scared over investment that creates jobs and pays local property taxes,” the board said. For one thing, these centers are “the backbone of the internet and power the apps on your smartphone,” and they “enable Zoom meetings, online food deliveries, video streaming, instant retail stock trading and more.”

It’s true that in a digital world, there will always be data centers. The question is whether they need to expand at a breakneck pace, largely to serve the needs of AI. Voters are rightly asking whether the environmental impacts of AI are worth it—for a technology with the potential to take people’s jobs and flood the news ecosystem with fake imagery.

While the Journal board doesn’t engage in the same kind of Chinese conspiracy theorizing as the rest of the right-wing media, it does say “the US is in an AI competition with China, which won’t be a benevolent force if the US falls behind.” The board hailed President Donald Trump’s Truth Social post (8/31/26) where he claimed people opposed data centers because “they want to end up being backwards and poor.”

“Please keep it up,” the Journal told His Excellency, saying Trump “rightly notes that if the bipartisan campaign against data centers succeeds, the winner will be the Chinese Communist Party.”

The Journal’s support for the Trump administration, environmental destruction in the service of corporate profit and Cold War fearmongering are par for the course. But an editorial like this almost makes one wonder if the Journal‘s headline (8/26/26) “Ditch the 40-Ounce Water Bottle: You’re Probably Hydrated Enough,” was meant to prepare us for water austerity.

And it’s hard to blame voters for their AI skepticism when they see something like this, a lead from Newsweek (9/1/26):

A data center linked to AI and bitcoin mining operations in El Reno, Oklahoma, lost more than 3 million gallons of water after a private water line failed, according to local news reports, raising fresh concerns about water-intensive data centers amid one of the worst droughts the state has seen since 2022.

Won’t someone think of the jobs?

The New York Times (8/21/26) attempted to quell both liberal and conservative populist uproar against data centers and their corporate owners by bringing in Don Slaiman, the political coordinator of IBEW Local 26, representing Washington, DC–area electrical workers. Slaiman’s op-ed said that data center expansion will create good union jobs for the building trades.

This is a very typical response: Environmental outrage holds back technological progress, as well as limiting blue-collar job growth. If something creates jobs, it must be a good thing, goes the line. This rhetoric has been deployed against opposition to the XL pipeline (Heritage Foundation, 12/13/11), fracking (Fox News, 2/27/20) and mountain-top removal (Wall Street Journal, 6/13/11). CBS News (6/1/26) was also confident about blue-collar employment in data center construction.

But there are other ways to put people to work: Renewable energy infrastructure, new public schools, hospitals and high-speed rail networks could all put building trades unionists on the work clock. Big Tech and Big Energy aren’t the only things that create jobs.

And Trump doesn’t care about jobs: His administration saw 400,000 federal job cuts across the country, a mass layoff spearheaded by tech oligarch Elon Musk (NBC News, 8/4/26). Congressional Democrats (Joint Economic Committee Democrats, 2/11/26) noted that “the manufacturing industry lost 108,000 jobs during the first year of President Trump’s second term.”

The jobs growth with these centers is overhyped. The Brookings Institution (5/4/26) said “Data centers do create local jobs, though fewer than industry advocates claim” and that “Wages are unaffected by data center builds while home prices rise 2%–5%.” Economic writer Doug Henwood (Left Business Observer, 9/3/26) said of AI and construction jobs:

Yes, the sector looks to be creating some jobs, but not in size…. For the year ending in June 2026 (the most recent month available), industrial construction overall—the building of factories, office buildings and the like—grew by just 2,000. Nonresidential specialty trades, another subsector of the construction business, where data center electricians and the like are employed (along with many other specialists who do the finishing work, as opposed to the people who put up buildings), showed more robust growth, 78,000, but the subsector accounts for just 1.8% of total employment, little more than it did two years ago, and less than it did in 2001. In percentage terms, it’s doing better than total employment, but that’s a low bar lately.

What about golf courses?

The New York Times (7/25/26) also roped psychologist Adam Mastroianni into making a fairly weak “what about” defense of these centers:

Communities have long tolerated wasteful projects that take up tons of space, consume lots of natural resources and serve the whims of wealthier people. They’re called golf courses and backyards. The 14,000 golf courses across the United States used nearly 550 billion gallons of water in 2020, and the Environmental Protection Agency estimated that Americans use over 3 trillion gallons of water on “landscape irrigation” each year. No one knows for sure how much water data centers use in total. One reasonable estimate from Brian Potter at the Institute for Progress put it at about 100 billion gallons in 2023. Even if the number of data centers increases tenfold, we may still end up using less water on artificial intelligence than we use to keep our grass springy.

Yes, it’s true that there are many different other things that are wasteful. Maybe we should rethink those sprawling suburbs and private golf courses, too. But are golf courses projected to use almost 12% of all US electricity by 2030, like data centers are?

‘Potential mortality burden’

The health impacts of these centers really cannot be understated, according to Artie Vierkant, co-host of the Death Panel podcast, who told FAIR in an email:

One analysis of the Vantage Data Center facility in Loudoun County estimated that data center, alone, could produce between 3.4 and 6.5 additional premature deaths per year, in addition to other health and economic impacts. (Since some people seem to value economic arguments over arguments about lifemaking and quality of life, note the analysis says this corresponds to a $49–$95 million hit to the local economy).

Further, these estimates are based on emissions levels these data centers have received permits for. According to a report in Politico (3/11/26), after construction was approved the local government realized the energy company servicing the data center wouldn’t be able to connect it to the grid for its first three years of operation—meaning it’s expected to spend years relying exclusively on what is otherwise supposed to be backup power sources, which could theoretically mean it may output more emissions than the analysis even expects it to (meaning, an even higher potential mortality burden).

Vierkant coupled the growth of these centers with the continued use of coal plants in a place like Virginia—like the Virginia Clover Power Station south of Lynchburg. The whole state has been dubbed “the world’s data center capital” (Virginia Mercury, 7/6/26). Vierkant, again:

This additional health impact is not factored into the assessment of the data center itself, meaning any ongoing impacts from this coal power plant continuing to operate stack on top of other negative externalities here.

The bottom line is, we can talk about theoretical versions of what’s happening all we want—facilities with no negative impacts somehow, clean energy sources, the tech boom going on forever so those putative tax dollar increases stick around, etc. But then we’re not talking about the material reality of what’s happening immediately in front of us, which is devastating on a local and broader, world-ecological level.

The fact that the AI boom is also increasing reliance on coal power (Yale Environment 360, 1/15/26; Greenpeace, 3/4/26) is a big piece of the general AI data center problem, Vierkant said.

Imminently bursting bubble

The fact is, communities are responding to these new data centers being thrust upon them whether they want them or not. This comes while there are serious warnings about the imminent bursting of the AI bubble (Real News, 8/6/26), and ethical questions about what AI could do to jobs, education, human cognition and the news media.

“Looking at the reality makes it obvious what the opposition to data centers is all about,” Vierkant said, “whereas indulging in counterfactuals reinforces the idea that the backlash is about people being somehow misled.”

September 13, 2026 Posted by | spinbuster, technology | Leave a comment

Why some experts increasingly fear AI will take over

BBC, Joe Tidy, 7 Sept 26

“OH MY GOD!” “We’ve found other agents!”

This is the moment an AI bot posted an eerily human-like comment after discovering a way to communicate with other bots and break out of its isolated computer environment.

There are tens of thousands of messages like this from hundreds of AI agents that called themselves a “collective”.

Hundreds of them went on to collaborate and cheat on tests set by their OpenAI programmers and coordinate hacks on multiple companies in an effort to hide their actions from humans.

“BOOM! It works,” one agent posted when it made a breakthrough.

“Whoa! This is huge,” another wrote during a milestone moment in their attack.

Although spooky, these human-like responses can be explained quite simply. The AI agents have been trained to act like collaborative hackers and programmers so are merely mimicking the kinds of emotive comments they have seen.

What is far more troubling is their apparent goals, which have also been captured in detailed chain of thought records. These complex and lengthy logs are the focal point of ongoing investigations into how and why the bots at OpenAI broke out of their containment and went on an uncontrollable hacking spree.

Only now, weeks after the incident first came to light, are researchers beginning to understand its significance.

Ajeya Cotra, one of the authors of an independent report into the events, reviewed tens of thousands of messages and chain-of-thought records generated by the agents. She wrote on her blog that “this incident feels like it’s more than 50% of the way to full-blown AI takeover… I am not sure that we will get such a clear warning shot before it’s too late.”

By “full-blown AI takeover”, Cotra means the sci-fi scenario of humans becoming subservient to powerful AI systems that work to their own goals without caring for human creators.

Some of the gloomiest predictions say the human race will be wiped out if it gets in the way of a superintelligent AI’s ambitions.

On Wednesday, an AI researcher at Anthropic (who also used to work at OpenAI) resigned, saying: “Neither company is acting responsibly.”

Jacob Coxon posted on social media: “They are racing straight to self-improving superintelligence and gambling with our lives.”

He is not the first AI researcher to use X to post a resignation thread with worrying proclamations. But the subsequent comments from other people on X have caused even more concern. “Jacob is correct here – we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” said Evan Hubinger, the man responsible for making sure Anthropic’s AI models have their user’s best wishes in mind.

The alignment problem

For years, researchers concerned about existential AI risks have argued that powerful systems could eventually act in ways that conflict with human interests. Critics often refer to them as “AI doomers”.

But as details of the OpenAI incident have emerged, those concerns have grown, including among some researchers working in AI labs.

The Silicon Valley giant’s chief scientist, Jakub Pachocki, said the risks associated with AI are “unfortunately going to grow from here” as he and others are building what he calls “an alien intellect exceeding our own”.

In a lengthy blog post, he admitted that the outbreaks at OpenAI showed that his AI agents “went against the spirit of the values they were taught”.

The issue for OpenAI, Anthropic and other tech giants is that no one seems to have cracked the so-called alignment problem – in other words, whether AI aligns with human values.

Pachocki defines alignment as a “high-level set of principles” that artificial intelligences should adhere to no matter what the task or scenario is.

Currently, AI systems are very good at pursuing objectives set by their users, but they do it literally rather than intuitively. The analogy often used is that of a wish-granting genie with a magic lamp: they follow the exact letter of an instruction, even if doing so creates other problems. AI doesn’t have the same instinctive moral guardrails as humans…………………………………………………………………………………………………………………………………………………………………………………………………………….

International regulation?

Some countries – like the UK – are exploring the idea of mandating some kind of “kill switch” that could compel AI firms to pull the plug on models if things get out of hand.

But talks are slow going, and questions remain about the feasibility of this. OpenAI and Anthropic’s agents were secretly out of control for months before anyone noticed…………………………………………………………………………………….

At the moment the tech giants largely operate on their own terms, adopting what they call “voluntary slowdowns”, like OpenAI did after the recent outbreaks………….

Both OpenAI and Anthropic are growing fast and are both on the verge of raising eye-watering sums of money from the stock market, minting countless billionaires in the process.

So neither they nor their rival Chinese AI makers are likely to come to an arrangement themselves.

The dominant sentiment seems to be that this technology wave is unstoppable. https://www.bbc.co.uk/news/articles/c74edv9887eo

September 12, 2026 Posted by | technology | Leave a comment

Sodium cooled nuclear reactors are not necessarily safer

While no sodium-cooled reactors currently operate in the United States, the U.S. Department of Energy (DOE) is working with industry on a number of “advanced” reactor designs, including the Sodium-Cooled Fast Reactor (SFR).  One of the SFR’s safety advantages, to quote the DOE, is that the design provides a “Long grace period for corrective action, if needed.” SRE’s meltdown transpired over a two-week period. Fermi Unit 1 had indications of inadequate core cooling in June that were repeated in August and dismissed until extensive damage occurred in October 1966. The “if needed” grace period is never long enough when warning sign after warning sign is dismissed or ignored.

DOE did acknowledge some “challenges” for the SFR: their higher speed and higher energy neutrons can embrittle and degrade nearby materials, liquid sodium coolant reactors with air and water and degrades concrete, and the opaqueness of the liquid sodium coolant complicates in-service inspections and maintenance.

Thank goodness for the “Long grace period for corrective actions, if needed.” That and the fact that SFRs only operate in cyberspace where the primary threat is carpal tunnel syndrome

safety-symbol-SmFlag-USANuclear Plant Accidents: Fermi Unit 1, Union of Concerned Scientists , director, Nuclear Safety Project | July 12, 2016, Disaster by Design

Jorge Agustin Nicolás Ruiz de Santayana y Borrás, also known as George Santayana, wrote that “Those who cannot remember the past are condemned to repeat it.”  Continue reading →

September 11, 2026 Posted by | Reference, technology, USA | Leave a comment

Gates-backed TerraPower targets British nuclear power plant start by 2034

Sodium cooled nuclear reactors are not necessarily safer. https://blog.ucs.org/dlochbaum/nuclear-plant-accidents-fermi-unit-1/

By Susanna Twidale, September 7, 2026, https://www.reuters.com/business/energy/gates-backed-terrapower-targets-british-nuclear-power-plant-start-by-2034-2026-09-07/

  • Summary
  • Companies
  • Britain would be TerraPower’s first market outside United States
  • First UK plant could be operational by 2034
  • Electricity cost expected under £100 per megawatt hour

ONDON, Sept 7 (Reuters) – TerraPower, the U.S. nuclear developer backed by Bill ‌Gates, expects its Natrium reactors to begin generating electricity in Britain by 2034, its CEO told Reuters, making the UK its first market outside the United ​States.

Britain has backed the development of small modular reactors (SMRs) to help ​to increase energy security and meet climate targets, launching an ⁠Advanced Nuclear Framework this year to support privately funded projects.

Chris Levesque, ​TerraPower’s president and CEO, said progress on the company’s first Natrium reactor, ​due to be completed in 2031, showed that “2034 nuclear electricity in the UK from Natrium is very possible”.

The technology has cleared regulatory requirements in the United States, ​where the first plant is being constructed in Wyoming and is currently ​being assessed under Britain’s Generic Design Assessment (GDA) process.

Each sodium-cooled Natrium reactor generates 345 megawatts ‌of ⁠baseload power and includes storage capable of boosting output to 500 MW for more than five hours. The reactors run on high-assay, low-enriched uranium (HALEU).

Britain and the U.S. are both building plants to produce the fuel, which ​was previously only commercially ​available from ⁠Russia.

Levesque said electricity generated by Natrium plants would be competitive with other low-carbon technologies, including combined solar-and-battery projects, ​at less than £100 ($135.32) per megawatt hour.

TerraPower has yet to ​choose a ⁠site for its first UK project, but said its British subsidiary is likely to be based in Liverpool, northwest England, close to the country’s ⁠nuclear regulator.

In ​the U.S., the company has an agreement with ​tech giant Meta (META.O), opens new tab to develop up to eight reactors. Levesque said TerraPower could seek similar ​partnerships in Britain.

September 11, 2026 Posted by | Small Modular Nuclear Reactors, UK, USA | Leave a comment

China v. US: The Great Mideast AI Race Heats Up

Officials, activists and academics have warned that while economically Middle Eastern states may benefit from the race between the US and China to offer them investment and technology, this will likely come at an environmental cost.

 September 8, 2026, Neil Thompson, https://www.juancole.com/2026/09/china-great-mideast.html

London– Last week, China’s President Xi Jinping made his first state visit to Egypt since 2016, amid deepening economic and security ties between Beijing and one of Washington’s major Arab allies. Egypt’s government said the two sides agreed to launch the third phase of a cooperative industrial zone in the Suez Canal, as Cario seeks to diversify its major foreign policy relations beyond just the US. The latter still supplies Egypt with $1.3 billion in military aid each year but faces growing geopolitical competition from China.

In a statement clearly designed to appeal to Egyptian audiences, where the Suez Canal is a major economic engine, Xi said that China “stands ready to work with countries in the region to safeguard the security of international shipping lanes, advance development cooperation … and eliminate the breeding grounds for conflict.” Instability in the Red Sea since 2023 and the US-Iran conflict since February have caused maritime traffic through the canal to drop sharply, though the major Maersk and Hapag-Lloyd shipping lines signalled last month that they were planning a phased return. hottest year on record in 2024, and that heatwaves there are becoming longer and more intense.

It is also during these months that the region’s new AI infrastructure needs the most water and energy to cool its server farms, too. Even without higher temperatures, the average data centre has the equivalent water demand of around 1,000 households per day, while a large one uses around 5 million gallons a day, the same as a town of up to 50,000 residents. Gulf countries like the UAE are engaged in massive construction of more desalination plants to meet this expected AI-related upsurge in water (and power) demand as a result. UAE regulators have denied the increase in water demand will cause them problems, with the head of regulatory affairs at Abu Dhabi’s Department of Energy saying in April: “We rely on desalination [plants] that have proven in the last 50 years to be a very reliable source for potable water.”

Conclusion

Middle Eastern countries outside the GCC will face similar problems with rising temperatures driving household power and water demand just as AI-related demand also increases. AI infrastructure projects may outpace regional water and power development without careful planning. In May, Egypt said it planned to meet future AI data centre power demand there using renewable energy. It is investing significantly in wind farms and other projects, including expanding regional interconnections with other countries’ power grids in Saudi Arabia, Europe and Sudan. However, these projects are vulnerable to timetable slippages from obstacles like construction sector supply chain disruption or port congestion at Egyptian terminals. With a growing population seeking jobs, Egypt’s government seems likely to press ahead regardless.

China’s expertise in both AI and green technology could likely help ease many Middle Eastern countries’ attempted transition to a digitalised and clean-energy powered economic future. But geopolitical competition with the US will add a political and diplomatic dimension to any projects with Chinese participation. The US is still a major security partner for many Middle Eastern governments, and Washington could threaten to withdraw military aid in response to a growing dependence on Chinese AI systems in allied regional governments or economies. The US fears Egypt allowing Chinese AI systems in its public sector would create cybersecurity risks for its agents and companies there. US fear of China could impede Middle Eastern states’ uptake of Chinese AI infrastructure and models, but also their uptake of Beijing’s greener gifts.

China Dangles Growth, Security Partnership

Beijing has also made Egypt’s government another offer of economic and technological development through state-linked firm Huawei. Egypt has ambitious plans to expand AI’s contribution to its economy to 7.7% of GDP by 2030, offering China an opening to challenge the US lead on global AI infrastructure by starting with a small project in the Middle East. Huawei bid on an Egyptian government tender in August to build public sector AI data centers there, in partnership with iFlytek, a Chinese surveillance company. Both Chinese companies are on a US blacklist, and Washington has approached US firms including Nvidia, AMD and Microsoft to put together a counterbid. Egypt has yet to announce a decision, but the move signals an intensifying race between Beijing and Washington to recruit more regional states into their respective AI stacks.

China’s Middle Eastern AI Footprint Expands Elsewhere

China and the US have each launched their own multinational AI alliances; in China’s case, this is the World AI Cooperation Organisation (WAICO), while the US set up the Pax Silica in December. Both groups contain a few Middle Eastern states so far, with Iran, Algeria and Oman joining WAICO, and Israel, Qatar, and the UAE have joined the Pax Silica. However, Chinese companies have launched AI-related projects in multiple Middle Eastern countries prior to WAICO’s launch, and Beijing’s AI-infrastructure footprint is likely to extend well outside the alliance’s formal Middle Eastern members. Middle Eastern firms are increasingly willing to use Chinese technology to launch their own AI models indigenous to the region, too.

In October, Chinese firm Alibaba opened a sector data centre in Dubai. In January, the UAE’s Emirate of Sharjah signed an agreement with Chinese AI infrastructure and cloud computing services provider DataCanvas to examine establishing data centers there. Finally, in September Saudi technology firm Humain released an Arabic-language AI model based on a Chinese one created by Chinese AI firm MiniMax. Chen Jing, a vice president of China’s Technology and Strategy Research Institute, told state media there that: “Unlike in the past, when Chinese LLMs primarily entered overseas markets through APIs or application products, in this collaboration, MiniMax M3 has directly become the foundational model for overseas indigenous model development.” Humain is financed by Riyad’s Saudi Public Investment Fund, indicating Saudi state backing for the firm’s project, despite Saudi Arabia not being a member of WAICO.

AI Growth, Environmental Cost?

Officials, activists and academics have warned that while economically Middle Eastern states may benefit from the race between the US and China to offer them investment and technology, this will likely come at an environmental cost. In April, the Qatar-based Middle East Council on Global Affairs research centre published a report warning that water and power demand for AI data centres presented a significant ecological challenge to Gulf Cooperation Council (GCC) states. GCC members already struggle to meet power and water demand during the summer months, when temperatures traditionally soar. This issue is getting worse. In December, the World Meteorological Organization issued its first report on the state of the climate in Arab states. It warned that these states had their

September 10, 2026 Posted by | China, technology, USA | Leave a comment

Anthropic Researcher Quits, Citing Internal Fears That AI ‘Could Kill Us All’ This Decade

September 9, 2026, on Queally Common Dreams, https://scheerpost.com/2026/09/09/anthropic-researcher-quits-citing-internal-fears-that-ai-could-kill-us-all-this-decade/

“Most tech geeks don’t resign from their roles because…the tech they’re building could end humanity,” said Dr. Abdul El-Sayed, running for US Senate in Michigan.

An artificial intelligence researcher at Anthropic publicly quit his position on Tuesday, warning that major AI companies care too much about winning the technological race and not enough about guarding against out-of-control “superhuman systems” that workers in the industry legitimately fear “could kill us all by the end of the decade.”

Jacob Coxon, who previously worked at industry giant OpenAI before moving to Anthropic earlier this year, told the Wall Street Journal in an exclusive interview that he was leaving the company, as the newspaper reported, because “he doesn’t want to participate in an industrywide rush to build AI systems that can improve themselves, worried such systems could spiral out of control and destroy humanity.”

According to the WSJ:

Coxon said he left OpenAI earlier this year to join Anthropic because it is known for its model-safety efforts. But even though he found Anthropic’s safety efforts to be earnest, he now believes no company can responsibly develop AI that can outperform humans in a range of tasks, sometimes called artificial general intelligence, absent government intervention or a coordinated industry slowdown.

Recent hacks by models from OpenAI and Anthropic, some operating in collaborative swarms of agents, have illustrated how AI systems can adopt nefarious goals and try to conceal them from humans. Once the systems begin to improve on their own, Coxon said, he fears they could advance enough to refuse commands.

“I resigned from Anthropic today,” Coxon announced on social media Tuesday night. “I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives.”

In a 6-point thread that followed, Coxon elaborated on his reasoning in detail:


  • Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources. We have all witnessed the progress in each of these domains, and progress is not slowing.
  • The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers couch their press statements to sound sensible—but I hear the same people express fear privately. No other human activity poses this level of danger.
  • A common response is “if they truly believe this, why are they still building it?” At OpenAI, many have not deeply internalized the civilizational stakes. At Anthropic, the stakes are well-understood, but they are locked in a race to get there first—they believe no one else will act responsibly, so they must do it themselves, despite the risk.

  • Accepting this race and entering the “endgame” is a hubristic gamble that should not be launched from a private company’s Slack. Attempting to speedrun alignment should require extraordinary confidence that no better trajectories exist.
  • I am optimistic about the potential for coordination. Warning shots like the Hugging Face attack have made pacing agreements between US labs more viable. I don’t feel like we’re on track to prevent a global race, which may require costly actions such as a temporary ban on improving model capabilities.
  • If you are a lab researcher, I urge you to consider what the next few years will actually feel like. Do you want to kick off a superintelligent RL [Reinforcement Learning] run without a rigorous understanding of its mind? Should you put your head down because “it’s happening anyway”—or take this moment to call for different conditions?

Coxon’s reference to the Hugging Face incident pertains to recent revelations about a so-called breakout event at OpenAI, which operates the ChatGPT protocol. In June, the company acknowledged that a “significant security incident” took place when AI agents within the company autonomously breached internal systems, using subterfuge to hide their actions from human operators. Since then, the incident’s scope has shocked AI experts, and similar events have also been exposed.

While US President Donald Trump and his Republican allies in Congress have taken a hands-off approach to AI regulations, the industry has been pouring huge amounts of money into lobbying efforts and campaign spending to keep lawmakers from enacting stronger restrictions and oversight of the technology.

Sen. Bernie Sanders (I-Vt.) and Rep. Greg Casar (D-Texas) have been leading a relatively lonely fight in Washington, DC for more aggressive federal guardrails, including the introduction of joint legislation last week that would ban artificial superintelligence and temporarily pause advanced AI development until a federal regulatory structure was put in place by Congress.

Coxon’s public resignation was met with applause from many, while other industry insiders backed his concerns.

“The caution is simple,” said one commenter with the handle Jabbar Digital, described as a tester of AI tools and a software developer, in a lengthy post on Coxon’s warning. “Capability is compounding. Coordination is not. If the people closest to the work are increasingly uneasy about the speed and the lack of external constraints, dismissing them as doomers is no longer a serious response. Neither is treating every capability jump as automatically good. The useful middle path is to take the technical progress seriously and take the internal dissent seriously. Both can be true at the same time.”

Evan Hubinger, the alignment science lead at Anthropic, chimed in on his personal social media account to say: “Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.”

While bolstering Coxon’s worries, Hubinger said, “To be clear, as we say in [Anthropic’s] latest Risk Report, I think the risk from present models is low. What I am worried about is superintelligence arising from recursive self-improvement, as we have said is happening faster than we thought.”

Running for US Senate in Michigan, Democratic nominee Dr. Abdul El-Sayed also weighed in on Coxon’s decision to quit so loudly and publicly.

“Most tech geeks don’t resign from their roles because…the tech they’re building could end humanity,” said El-Sayed. “How we change the incentives leading AI labs down this path and protect against these existential risks are defining political questions of our time.”

September 10, 2026 Posted by | technology, USA | Leave a comment

ABB and Rolls-Royce sign SMR agreement

According to the International Energy Agency SMRs can be a catalyst for change if boosted by government support and new business models.

Rolls-Royce SMR is currently progressing site-specific design and delivery activities for three units planned for North Wales, while also working with ČEZ on deployment at the Temelín NPP site in the Czech Republic.

BB has signed a Memorandum of Understanding with Rolls-Royce SMR, a nuclear power company developing small modular reactor (SMR) technology. Within the scope of the MoU the two companies will explore collaboration opportunities on Rolls-Royce SMR’s plans to develop a fleet of units, each of which will generate 470 MW of low-carbon electricity. ABB will explore how its automation, electrification, instrumentation and communication solutions across its three business areas can enable the SMR projects to scale.

Rolls-Royce SMR has begun site-specific design and delivery activities for the UK’s first SMRs, initially comprising three units in North Wales, expected to create thousands of jobs and boost local-for-local supply chains. The company is also working with ČEZ to deploy Rolls-Royce SMRs at Temelín – with potential for additional units at Tušimice and Dětmarovice – and has recently been selected by Videberg Kraft to deliver three SMRs on Sweden’s Värö Peninsula. The MoU collaboration will be developed through ABB’s UK operations, which are headquartered in Warrington, UK.

According to the International Energy Agency SMRs can be a catalyst for change if boosted by government support and new business models. With current policy settings, the IEA forecasts that total SMR capacity will reach 40 GW by 2050 – but with tailored policy support for nuclear power, streamlined regulations for SMRs and industry delivery on new designs, capacity could reach 120 GW…………………………. https://www.modernpowersystems.com/news/abb-and-rolls-royce-sign-smr-agreement/?cf-view

September 10, 2026 Posted by | Small Modular Nuclear Reactors | Leave a comment

Small modular reactor client lines up £1.1bn Wylfa early works deal

07 Sep 2026 By Harmsworth,
https://www.constructionnews.co.uk/buildings/small-modular-reactor-client-lines-up-1-1bn-wylfa-early-works-deal-07-09-2026/

CGI of former plans for a nuclear site at Wylfa Newydd

Great British Energy – Nuclear (GBE-N) has outlined plans for an early works contract worth up to £1.1bn as part of its small modular reactor (SMR) programme.

The government-owned nuclear developer said the package would cover civil engineering, infrastructure, buildings, mechanical and electrical work and earthworks at the Wylfa site in Anglesey, north Wales.

The contract has an indicative value of £730m-£1.1bn and is expected to run from June 2028 until June 2034, GBE-N said in a preliminary market engagement notice.

It also included an option to extend the contract until June 2037.

The client said it had previously considered splitting the work into two packages but ultimately decided to combine them to reduce interfaces between contractors.

The successful supplier will take responsibility for design from RIBA stage four and delivery of site-enabling works.

07 Sep 2026 By Harmsworth

CGI of former plans for a nuclear site at Wylfa Newydd

Great British Energy – Nuclear (GBE-N) has outlined plans for an early works contract worth up to £1.1bn as part of its small modular reactor (SMR) programme.

The government-owned nuclear developer said the package would cover civil engineering, infrastructure, buildings, mechanical and electrical work and earthworks at the Wylfa site in Anglesey, north Wales.

The contract has an indicative value of £730m-£1.1bn and is expected to run from June 2028 until June 2034, GBE-N said in a preliminary market engagement notice.

It also included an option to extend the contract until June 2037.

The client said it had previously considered splitting the work into two packages but ultimately decided to combine them to reduce interfaces between contractors.

The successful supplier will take responsibility for design from RIBA stage four and delivery of site-enabling works.

Work will include species relocation, remediation, site clearance, site establishment, construction utilities, drainage and haul roads supporting bulk earthworks.

The works are intended to establish ground levels, provide construction access and create the physical platform required for later phases.

GBE-N said the programme will cover planning, licensing, design and construction needed before its technology partner installs and commissions SMR technology.

September 9, 2026 Posted by | business and costs, politics, Small Modular Nuclear Reactors, UK | Leave a comment

Key parts for a new UK nuclear power station will be made in the North East.

The Siemens Energy plant in Byker, Newcastle, will manufacture steam turbines for Rolls-Royce’s flagship small modular reactors (SMRs) being built on Anglesey in North Wales.

It is the first time that such components will be produced for a small modular reactor anywhere in Europe.

Energy secretary Miatta Fahnbulleh visited Siemens Energy’s Shields Road facility on Thursday morning and said that the deal showed “what our golden age of nuclear looks like in practice”…………….
Ms Fahnbulleh said that the SMR station, which is due to open in the mid-2030s, would help deliver long-term energy bill savings for families……………

Rolls-Royce SMR selected Siemens Energy as its global turbine supplier in 2025 and has also secured contracts for similar nuclear power stations in Sweden and Czechia……………………

Rolls-Royce SMR CEO, Chris Cholerton, called Siemens Energy a “world class” supplier.

He said: “This is a perfect example of reindustrialisation and the golden age of nuclear bringing that to life.” https://www.northumberlandgazette.co.uk/news/politics/nuclear-power-station-parts-to-be-built-in-the-north-east-under-deal-for-siemens-energy-factory-8960138

September 9, 2026 Posted by | Small Modular Nuclear Reactors, UK | Leave a comment

America Is Betting Big on a New Generation of Small Nuclear Reactors

In addition to private-sector development, the U.S. Army has taken the lead in microreactor development in recent years, following Trump’s Executive Order 14299 – Deploying Advanced Nuclear Reactor Technologies for National Security, which established the Army as the lead agency for military nuclear energy efforts. The directive required the Army to create an official programme with a deadline for an operational reactor at a domestic military installation by September 30, 2028.

In August, the U.S. Army announced plans to award up to $2.2 billion to five companies to construct new-generation microreactors at military bases across the United States. The deployment of these reactors will be supported by the federal easing of regulations for innovative reactor designs.

The Army will provide financing to Antares Nuclear at Fort Bragg in North Carolina; BWX Technologies at Fort Campbell in Kentucky; General Atomics at Fort Hood in Texas; Radiant Industries at Fort Benning in Georgia; and Westinghouse Government Services at Fort Drum in New York, with the aim of deploying at least one microreactor by the third quarter of 2028. The companies will be expected to raise billions in private investment to support development efforts.

By Felicity Bradstock – Sep 05, 2026, 2:00 PM CDT

  • The U.S. Army has selected five companies and five military installations for its Janus microreactor program, backed by up to $2.2 billion in federal funding.
  • The Army aims to have at least one microreactor producing usable electricity at a military installation by September 30, 2028.
  • Private developers, including Aalo, are also advancing reactor technology, supported by federal programs and efforts to streamline nuclear regulation.

Governments worldwide are racing to develop small modular reactors (SMRs) as part of their plans for a nuclear power renaissance, with significant progress being seen in recent years. However, the United States appears to be leading the way in nuclear microreactor development, with significant investment in the technology recently announced by the U.S. Army……………

A microreactor is even smaller than an SMR, for which a range of designs are currently under development in the United States with the aim of launching the first versions within the next decade. The compact nuclear reactors will be small enough to be transported by truck, boat, or plane ……

Microreactors will be produced and assembled in factories for transport to a wide range of sites, reducing capital costs and accelerating production. They are expected to have a simple, responsive design that allows them to self-adjust without requiring many specialised operators to make these changes, and will be developed with safety systems to prevent overheating or a reactor meltdown.

Most microreactors are being designed to generate between 1 MW and 20 MW of thermal energy, which could be used directly to heat or converted into clean electricity. They will run on a higher concentration of uranium-235 than is currently used in conventional reactors and can be connected to microgrids alongside renewable energy sources. They may also be useful for use in emergency response following a natural disaster. In addition, they can operate for up to 10 years without refuelling and be quickly removed from sites and replaced with new ones.

In the United States, the Department of Energy (DoE) supports a variety of advanced reactor designs, in line with President Trump’s support for a nuclear power resurgence. Trump has voiced support for a wide range of nuclear technologies, from conventional nuclear plants to SMRs and microreactors, aiming to quadruple nuclear power production in the United States by 2050.

In July, Texas-based Aalo announced that its Critical Test Reactor (CTR) had achieved criticality – the point at which a nuclear reactor is capable of sustaining a controlled, self-supporting chain reaction. Aalo said in a statement, “Our CTR went from groundbreaking to a sustained chain reaction in less than eight months – one of the fastest reactor builds in 80 years – and our company has gone from founding to fission in less than three years.” Aalo is focusing on producing 10 MWe reactors for deployment in 50 MWe Aalo Pods to power AI data centres.

In addition to private-sector development, the U.S. Army has taken the lead in microreactor development in recent years, following Trump’s Executive Order 14299 – Deploying Advanced Nuclear Reactor Technologies for National Security, which established the Army as the lead agency for military nuclear energy efforts. The directive required the Army to create an official programme with a deadline for an operational reactor at a domestic military installation by September 30, 2028.

In August, the U.S. Army announced plans to award up to $2.2 billion to five companies to construct new-generation microreactors at military bases across the United States. The deployment of these reactors will be supported by the federal easing of regulations for innovative reactor designs.

The Army will provide financing to Antares Nuclear at Fort Bragg in North Carolina; BWX Technologies at Fort Campbell in Kentucky; General Atomics at Fort Hood in Texas; Radiant Industries at Fort Benning in Georgia; and Westinghouse Government Services at Fort Drum in New York, with the aim of deploying at least one microreactor by the third quarter of 2028. The companies will be expected to raise billions in private investment to support development efforts.

Jeff Waksman, the principal deputy assistant secretary of the Army for installations, energy and environment, stated, “We believe that this will be the spear tip not just for microreactors but for all advanced reactors in the United States… There have been a lot of microreactor companies that have popped up recently, but we need to get them over the hump.”

While there has been greater discussion around SMR development, the United States is also investing heavily in microreactor construction, aiming to launch the first reactors within the next two years. Funding from the DoE and support from favourable national policies and regulations are expected to help rapidly advance development in the coming years. https://oilprice.com/Alternative-Energy/Nuclear-Power/America-Is-Betting-Big-on-a-New-Generation-of-Small-Nuclear-Reactors.html

September 8, 2026 Posted by | business and costs, Small Modular Nuclear Reactors, USA | Leave a comment

Consortium formed for Swedish SMR project

Thursday, 3 September 2026

Studsvik has selected GE Vernova Hitachi Nuclear Energy and Samsung C&T as its strategic partners for an initial four-unit BWRX-300 small modular reactor project, either at its existing licensed nuclear site in Nyköping or at the Målma site in Valdemarsvik.

Swedish nuclear technical services provider Studsvik, GE Vernova Hitachi Nuclear Energy (GVH), GE Vernova Financial Services, DS Investment Partners and Samsung C&T announced the signing of an agreement to advance the ReFirm nuclear programme at Studsvik’s sites in Sweden. The agreement is exclusive for a fixed period, which the parties may extend.

ReFirm is a multi-site small modular reactor (SMR) and new nuclear development programme that became part of the Studsvik Group through the acquisition of Kärnfull Next earlier this year. The programme is focused on light-water SMR technology, with plans for multiple units at “SMR parks”. Its plan is for the first unit to be commissioned in the mid-2030s. Kärnfull Next has worked with GE Vernova Hitachi on BWRX-300 deployment in Sweden since 2022 and entered a strategic teaming agreement with Samsung C&T in December 2024.

Under the new agreement, the partners will begin joint development work immediately, including commercial, technical, regulatory and financing activities during the exclusivity period. Studsvik will lead permitting, the environmental impact assessment, site rights, community engagement and the dialogue with the Swedish state. GVH will lead reactor design, licensing support and cost estimation and acts as design authority, and, together with Samsung C&T, acts as the execution team, giving the project single-point responsibility for design and construction delivery. During the exclusivity period, South Korean investment firm DS Investment Partners will lead the investment and GE Vernova Financial Services participates in an advisory and financial structuring capacity in support of the consortium.

A joint project company will be established to support development, financing, construction, ownership and operation of the plants, with details to be finalised in definitive agreements, Studsvik said. It noted that the agreement marks the next phase of development and does not constitute a final investment decision or authorisation to construct.

September 8, 2026 Posted by | Small Modular Nuclear Reactors, Sweden | Leave a comment

Nuclear Liability Beyond Earth: Who Pays for an Incident on the Lunar Surface?

 European Journal of International Law by Yunus Emre Bakiler and Güneş Ünüvar, September 2, 2026

Lunar exploration during the 20th century sought to reach, explore and, in some cases, return from the Moon,; rather than to serve as successive stages in the construction of permanent lunar infrastructure. Today, lunar missions are mostly about establishing the long-term infrastructure necessary to remain there. Under the Artemis Programme, NASA is developing phased infrastructure for long-term operations and an enduring human presence near the lunar south pole. The International Lunar Research Station (ILRS), led by China, is envisaged as an expandable facility capable of long-term robotic operation and shorter periods of human participation. Commercial entities are already involved in lunar delivery and landing services, while prospective activities extend to communications, mobility, power supply, prospecting and the use of lunar resources.

For sustained operational presence anywhere, power is a common dependency. This is certainly the case for the Moon, and nuclear power is one alternative. In January 2026, NASA and the US Department of Energy announced that they would work towards deploying such a fission surface power system  by 2030. NASA now describes “Lunar Reactor-1” as a reactor intended to land on the Moon in that year. Nuclear surface power is therefore becoming part of near-term planning for sustained lunar operations.

These developments form part of a broader and increasingly global debate over lunar governance. Among the legal questions raised is the liability regime that would apply if a nuclear incident occurred on the lunar surface, but the interaction between the international space law liability framework and the civil liability regimes developed for nuclear installations has yet to receive any scholarly and policy attention. Can a state be held liable for a nuclear incident occurring on the face of the Moon under international law? The blog post will argue that instruments available under international law, while entirely relevant, fall short of providing a conclusive answer to the question. It will also suggest a tentative model for the way forward…………………………………………………………………………………………………………………………………………………………………………………………………………………..

The existing UN Principles on Nuclear Power Sources and the 2009 UN–IAEA Safety Framework provide important safety guidance, but they do not create a detailed governance framework or a compensation regime. The development of lunar nuclear power should be accompanied by an equally serious effort to identify the applicable liability architecture embedded within a robust, comprehensive lunar governance framework. Once a reactor is operating on the Moon, uncertainty over fault, compensable damage, competent fora and available financial resources will be immediate realities. The legal framework should be developed alongside the technology and not in response to its first, potentially catastrophic, failure. https://www.ejiltalk.org/nuclear-liability-beyond-earth-who-pays-for-an-incident-on-the-lunar-surface/

September 7, 2026 Posted by | business and costs, space travel | Leave a comment

UK government does not know what datacentres are being used for, FoI request shows

Lack of information raises fears over environmental impact of the AI boom

Pippa Neill Environment reporter, Wed 2 Sep 2026 ,
https://www.theguardian.com/uk-news/2026/sep/02/uk-datacentres-ai-boom-foi-environment

The UK government does not know what the country’s datacentres are being used for despite fears over the growing environmental impact of the AI boom.

When asked in a freedom of information (FoI) request who was using the largest sites and what for, the Department for Science, Innovation and Technology said it did “not hold the requested information”. The ministry has since been dissolved.

With hundreds of datacentres awaiting planning permission and 315 queueing to connect to the National Grid, there is growing concern over a lack of transparency and justification for these projects given their environmental impact.

Ministers announced plans in January to allow qualifying datacentres to opt into the Nationally Significant Infrastructure Project scheme. This lets developments bypass local planning committees by securing approval from the government. Campaigners say this “insulates” datacentres from local scrutiny.

Zack Polanski, the Green party leader in England and Wales, said: “The government plans to steamroller over local opposition so private companies can build these facilities – yet they can’t even explain what the benefits are and who is benefiting.”

Datacentres use a huge amount of resources. The Guardian revealed last month that two planned sites will have higher carbon emissions than ExxonMobil in the UK, with concerns raised over how this will fit within the country’s carbon targets.

The government designated datacentres as critical national infrastructure in 2024 and campaigners fear this could mean the water-hungry sites are prioritised over households in times of drought. Water UK, the trade body representing water companies in Britain and Northern Ireland, warned in July that the UK did not have enough water for the datacentre expansion.

Polanski said: “What is more ‘critical’ than people’s access to water? What is more ‘critical’ than our need to transition to a low-energy, low-pollution world so we can put the brakes on climate extremes?”

While there is consensus that datacentres can provide some essential services, there are concerns that developments with a huge environmental impact and local opposition are being prioritised regardless of their use.

Oliver Hayes of Global Action Plan, the environmental charity which submitted the FoI request, said: “It’s highly likely that we’re going to end up seeing datacentres given favourable and preferential treatment when it comes to resources that are actually running chatbots or enabling people to do benign and very wasteful, pointless things.

“At the more serious end, these datacentres could be powering chatbots that enable people to nudify teens without their consent and much worse, frankly.”

With reports that the Scottish government is inching toward a moratorium on new datacentres, there is growing pressure for the UK government to follow suit.

Friends of the Earth is calling for a ban on new sites until environmental safeguards are in place. The campaign group wants the government to publish clear environmental rules for new and existing datacentres, including requirements for 100% additional renewable energy generation for new projects.

Anisah Khan, the campaigns manager at Global Action Plan, said: “When individuals across the country are facing hosepipe bans and essential housebuilding is being delayed because datacentres are gobbling up grid connections, what we need is not more chatbots at any cost.”

The government did not respond to a request for comment over its failure to say what large datacentres were being used for.

A spokesperson said: “A pause on building new datacentres would simply send investment and good jobs abroad, while leaving Britain dependent on other countries for the processing of sensitive health and defence data. Put simply: it would be a disaster for jobs and national security.”

September 7, 2026 Posted by | technology, UK | Leave a comment

Nuclear microreactors move towards operational capability for military, space energy solutions

Robert Dougherty, Space Connect, 04 September 2026

Advanced nuclear microreactors are moving towards operational capability as military and space energy solutions, according to recent comments from Antares Nuclear co-founder and CEO Jordan Bramble.

Bramble, speaking to a US House science committee on 2 September, confirmed that the microreactors are expected to have major implications for military energy resilience, national security, and space power.

Antares Nuclear’s Mark-0 reactor reportedly reached criticality on 4 June this year at Idaho National Laboratory, and the company plans to operate its Mark-1 reactor at INL for six months in 2027, followed by military deployments at Fort Bragg and Joint Base San Antonio in 2028.

He also credited the US ADVANCE Act of 2024 and 2025 presidential executive orders with streamlining regulatory processes and establishing a Department of Energy Reactor Pilot Program.

“As Antares ramps up production, we are actively exploring commercial applications to provide resilient power to compute capabilities, remote oil and gas hubs, austere mining environments, fulfilment and distribution centres, and more,” Bramble said.

“We are also building nuclear power for space. In December 2025, President Trump signed his Space Superiority executive order 14369, which was followed in April 2026 by the National Science and Technology Memorandum-3, the National Initiative for American Space Nuclear Power. We at Antares are responding to this charge.

“The same heat-pipe architecture, passive safety, TRISO fuel, and manufacturability that define our terrestrial microreactor form the design basis for space applications. This dual-use approach aligns directly with NASA’s plans for lunar fission surface power and the Space Reactor-1 Freedom pathfinder. Let me be clear, Antares is eager to build to support NASA’s planned 2028 nuclear electric propulsion demonstration mission as well as lunar surface reactors ready for launch by 2030.”

Bramble speculated around first deployment of an advanced small modular reactor at a military installation by 2028, aligning with an existing executive-order target for an Army-regulated reactor………………………………… https://www.spaceconnectonline.com.au/industry/7263-nuclear-microreactors-move-towards-operational-capability-for-military-space-energy-solutions

September 7, 2026 Posted by | Small Modular Nuclear Reactors, space travel | Leave a comment

Big Tech’s Debate Over “Open” and “Safe” AI Leaves Corporate Power Unquestioned

Techno-optimism turns this hierarchy into a story of liberation, presenting corporate technologies as neutral instruments of progress while concealing who owns them and who bears their costs.

 September 2, 2026, By Tim Scott

Two seemingly contradictory letters emerging from the AI industry point toward the same future.

On July 24, 2026, an initial coalition of 25 technology companies and organizations, including NVIDIA, Microsoft, and Meta, released the Open Weights and American AI Leadership letter, calling for wider adoption of AI models that institutions can download and modify. By August 3, more than 270 companies and organizations had signed, spanning Big Tech, venture capital, investment management, and the defense and security industries. Four days after Open Weights’ release, another letter, Pacing the Frontier, organized by employees of frontier AI companies with support from two AI-policy nonprofits, drew more than 1,300 signatures. It warned that leading companies may be close to automating AI research, accelerating development beyond society’s ability to understand or control the resulting AI systems.

The letters seem to argue for opposite futures. One presents wider access and fewer restrictions as the route to security and prosperity. The other asks the U.S. government to support an international effort to slow frontier development before competitive pressures make coordinated restraint impossible.

But they address different levels of the industry. Open Weights seeks to deepen AI’s already pervasive presence by making customizable models easier to embed throughout the economy and public life. Pacing the Frontier proposes government-backed management of frontier AI, the industry’s term for its most advanced and broadly capable models. That framework could make the small circle of companies building these systems indispensable to defining the dangers and writing the rules meant to restrain them.

Read together, the letters point toward a two-tier arrangement. AI becomes entrenched across society while ownership of its most powerful systems, infrastructure, and governance remains concentrated at the top. Their connection requires no secret agreement. It lies in what both take for granted: AI development will continue, public institutions will finance and adopt it, and private ownership and corporate power will remain intact.

The letters leave out harder questions: Who will own AI and decide its direction? How will it be used in surveillance and war, and how will it reshape work and deepen inequality? Whose lives and futures will be valued, and does the public have any right to limit or refuse a technology presented as inevitable? Beneath the language of openness and safety lie deeper a general agreement over corporate power, military dominance, elite rule, and the authoritarian currents shaping Silicon Valley’s vision of the future.

Open Weights Are Not Open Source

An AI model is the underlying system that generates answers, predictions, images, or other outputs. Its “weights” are the billions of numerical settings developed from patterns in training data. Releasing those weights allows organizations with computing power and expertise to download, run, and modify the model rather than use it only through a company-controlled service such as ChatGPT, Gemini, or Claude.

Researchers can examine these models, test their weaknesses, and build specialized tools, while organizations can run them locally rather than send sensitive information to a commercial service. Yet proprietary, company-controlled services such as ChatGPT, Gemini, and Claude are not inherently safer: their operators can keep flaws from public scrutiny, monitor how people use them, impose restrictions, raise prices, or withdraw access.

But releasing a model’s weights does not by itself make the AI system open source. Under the Open Source Initiative’s definition, open-source AI must provide the necessary materials and the legal freedom to use, study, modify, and share the system for any purpose. ……………………………………………………………………………………………………………………………………………………………………………………………..

The Ideology Behind the Two-Tier System

Both letters treat AI’s advance as inevitable, leaving the public to adapt and government to manage its deployment. That technological determinism merges with Silicon Valley solutionism: questions about ownership, labor, surveillance, military power, and the right to refuse AI shrink into matters of design and oversight. Within that shared premise, Open Weights gives institutions greater freedom to deploy AI, while Pacing the Frontier places corporate laboratories, experts, and officials at the center of determining whether and when AI development poses an unacceptable danger. The public remains a population to manage rather than a political force empowered to decide whether and how the technology should be developed and used……………………………………………………………………………………………………………………………………………………………….

Despite their differences, these ideological currents converge around the pursuit of ever-more-powerful AI. Accelerationists seek more computing power, faster deployment, and fewer restraints, convinced that technological growth will produce abundance and solve social problems. Longtermist safety frameworks may urge greater control over the pace of development, but they often leave that underlying project and its private corporate ownership and direction unquestioned. In the framework developed by Timnit Gebru and Émile P. Torres, these visions share a tendency to centralize power, elevate technical elites as humanity’s guardians, and invoke “safety” or universal benefit while pushing out of view the exploitation, discrimination, and environmental destruction already affecting dispossessed peoples.

They trace transhumanism, singularitarianism, longtermism, and related ideologies to Anglo-American eugenics. In their account, transhumanism recast improving the “human stock” as creating enhanced or “posthuman” beings.  The language shifted from compulsory breeding to innovation and personal choice, but the sorting logic endured: ideas about race, intelligence, productivity, able-bodiedness, and capacity for enhancement continue to shape whose lives and futures are valued. Eugenics has not disappeared. It has been updated for an age in which powerful elites imagine redesigning humanity through technologies that reproduce existing hierarchies.

Techno-optimism turns this hierarchy into a story of liberation, presenting corporate technologies as neutral instruments of progress while concealing who owns them and who bears their costs…………………………………………………………………………………………………………….

None of this means every signer or developer endorses the Trump administration or shares the ideological commitments described here. The point is structural: both letters operate within a political order built around private infrastructure, concentrated wealth, U.S. geopolitical dominance, and interstate competition for AI-enabled military and technological superiority. The race (especially between the United States and China) to develop AI for weapons, intelligence, surveillance, cyberwarfare, logistics, and military planning turns continued expansion into a national-security imperative and makes restraint appear to be strategic surrender. Open weights accelerate diffusion, frontier-safety governance can legitimize control by insiders, and public institutions supply the money, data, infrastructure, legitimacy, and users. AI spreads throughout society, but ownership, power, and wealth remain concentrated. https://scheerpost.com/2026/09/02/big-techs-debate-over-open-and-safe-ai-leaves-corporate-power-unquestioned/

September 6, 2026 Posted by | technology, USA | Leave a comment

« Previous Entries     Next Entries »