Thursday, June 11, 2026

A regional defence pact could end Israel's violent expansionism



David Hearst, editor-in-chief of Middle East Eye, explains why he believes resistance by Iran and its allies has been the only force preventing Israel from pursuing further territorial expansion and the expulsion of native populations.

 

“Every nation in the Middle East will feel the consequences of Israeli success,” Hearst says.


“It is well past time to arrange a regional framework and defence pact that would become a fact on the ground, which neither Israel nor the latest occupant of the White House could ignore.

 

“An end to the war with Iran, if it happens, must be the start of a regional initiative to ensure it never happens again.”


Middle East Eye on You Tube

Something Pretty Explosive came out Today About the Epstein Files.




Something pretty explosive came out today about the Epstein files.

Rachel Hurley*

June 11 26

Does anyone else remember that rumor about the whole Trump team getting together to figure out how to deal with the fact that Trump was in them so many times?

 

Well, guess what - the NYT reported the whole story today. It's in a new book by Maggie Haberman and Jonathan Swan called "Regime Change," and here's what they say happened.

 

So - as we all recall - last July the DOJ and FBI go "yeah there's no client list, the guy killed himself, nothing to see here."

 

Then MAGA absolutely loses it. Charlie Kirk's Turning Point event turns into what the authors call an Epstein grievance fest, speaker after speaker trashing AG Pam Bondi. Trump reportedly called Kirk and chewed him out, then went online and told his own fans to quit wasting "Time and Energy on Jeffrey Epstein." 

 

It did not work.

 

Ten days later, JD Vance hauls everyone into the Situation Room - Pam Bondi, Kash Patel, Susie Wiles, Karoline Leavitt, and the deputy AG, Todd Blanche.

 

Trump's not even there. Vance walks in and goes, basically, this is a huge problem.

 

And the idea Vance throws out is get Tucker Carlson to go interview Ghislaine Maxwell in prison, on the theory that maybe she'll sit there on camera and say Trump never did anything.

 

Vance was actually the one arguing to just release everything. Everybody else wanted to sit on it. And the reason they wanted to sit on it is in the files. There's a document - emails from an Epstein accuser named Sarah Ransome - making a graphic claim about Trump. The book says officials sat in the most secure room in the country having a "surreal" conversation about the "nipple material," some of them hearing the claim for the first time. 

 

When the Ransome allegation came up, Vance argued for putting it and other accusations on the DOJ site as "maximum transparency," saying Trump's been accused of worse and wouldn't care. Wiles shut him down cold - told him the president would absolutely not be fine with that. 

 

Vance also tried to set up a Rogan interview timed to the DOJ launching its Epstein files website. The plan was to put Blanche on Rogan - but Rogan told Vance, on the phone, he didn't want Blanche. Vance then floated himself, pitching that only part of it would be about Epstein and the rest could be about the new legislation and "working families." Never happened.

 

Later at the DOJ, everyone was at each other's throats. Dan Bongino, then deputy FBI director, walked into a DOJ meeting and screamed at Bondi - "you f*cked this thing up from the start" - over the client list she'd promised was "on my desk" and then never produced. He and Kash Patel both reportedly told the White House she needed to go. Bongino called the whole mess Trump's Iran-contra.

 

Okay but the one who actually matters here is Todd Blanche.

 

Blanche runs the whole Justice Department right now - he's acting attorney general. And before any of this he was Trump's personal defense lawyer, the literal guy standing next to him in court. Then he got the number two job at DOJ, and now he's been nominated by Trump to run the place officially - because of course, right?

 

Pam Bondi already testified that Blanche had been in charge all along of how the files got released - and apparently, in that situation room meeting - he was workshopping how to make it seem like he was doing his job.

 

The book says Blanche brought two ideas. One, he'd personally go interview Maxwell. Two, ask a court to unseal the grand jury testimony. He pushed the court thing knowing it'd get tossed, because grand jury material is about the most locked-down stuff there is. He just wanted it on record that he asked, so somebody could point to it later.

 

Then he went and interviewed Maxwell. Two days, in the courthouse in Tallahassee where she's doing twenty years. Former prosecutors said that's insane - a senior DOJ official does not personally run an interview like that, line prosecutors do, with FBI in the room. He went himself, to talk to the one living person who knows the truth about his boss.

 

And then, about a week later, she mysteriously got moved to a nicer prison - a minimum-security camp in Texas, the kind of place sex offenders basically never get sent.

 

So, yeah - that's the incredibly, non-surprising story of the Trump administration trying to run cover for the star of the Epstein Files.

 

What a country.

 

Rachel Hurley is running for Congress in Tennessee’s 5th District

Publishing this article is for our readers interests and is not FFWP’s endorsement of her political campaign. This blog does not support the Democratic Party.

 

Wednesday, June 10, 2026

Legal profession revolt against the UK judge whose job is to protect Israel's genocide

Legal profession revolt against the UK judge whose job is to protect Israel's genocide


Judge Johnson so rigged the trial of anti-genocide activists that 1000s of legal professionals have urged him to step down from the sentencing hearing. But Johnson's dirty work is not yet complete.

Johnathan Cook, June 10th 2026

From Johnathan Cook on Substack

The trial of the Filton Four reaches its climax on Friday. Judge Jeremy Johnson will decide the sentences of four Palestine Action activists found guilty of criminal damage – after two juries refused to convict them of far more serious charges brought by the British government, via the Crown Prosecution Service.

Keir Starmer’s government failed to secure the convictions for aggravated burglary and violent disorder it so desperately needed. They would have helped retroactively justify its decision to proscribe Palestine Action as a terrorist organisation – the first time in British history that a direct action group, which targets property, has been proscribed.

Starmer’s hostility towards Palestine Action is personal. It has been a massive thorn in his side during Israel’s near three-year genocide in Gaza.

From the outset, Starmer suggested that Israel had the “right” to withhold food, water and power from the 2 million Palestinian residents of Gaza – a crime against humanity for which the Israeli prime minister, Benjamin Netanyahu, is wanted by the International Criminal Court.

Starmer’s complicity in the genocide has only been underscored by Palestine Action, which has targeted Israeli factories operating in the UK run by Elbit Systems that supply Israel’s genocidal army with killer drones for use in Gaza.

The so-called Filton trial derives its name from an Elbit factory in the Filton neighbourhood of Bristol that Palestine Action targeted in August 2024. Efforts to destroy these Israeli drones came at a time when Starmer was under enormous popular – if little political – pressure to end British arms sales to Israel.

PR crisis

Proscription, however, has not worked out well for Starmer. It has entangled him in a self-inflicted public relations crisis of almighty proportions.

Palestine Action activists have been held in remand without trial for an unprecedented period of time, far in excess of the normal maximums, and in especially harsh conditions that have treated them as if they were terrorism suspects, even though their arrests long precede the group’s proscription last year. 

These sustained abuses led to a prolonged hunger strike, and a desperate government and media campaign to justify their mistreatment. 

Proscription has led to thousands of people, most of them elderly and including upstanding members of British society – from magistrates and doctors to army veterans – facing convictions for “supporting terrorism” for holding up placards stating: “I oppose genocide. I support Palestine Action.”

This popular backlash cornered the High Court into declaring the proscription unlawful – a decision the government is appealing. That has led to another unprecedented situation: police are still arresting people for holding the placards, despite the courts ruling that the basis for such arrests is unlawful.

The law has never looked more of an ass.

Which is why the government is pinning its hopes on Judge Johnson riding to the rescue on Friday.

Judge Johnson has not been shy about showing where his loyalties lie. Not to the law, but to the British security state.

Which should hardly come as a surprise, given his background.

Judge Johnson made it to the bench after years serving as the most favoured barrister of the “secret state”, representing the intelligence services, the ministry of defence and the police. His working environment of choice as a lawyer was behind-closed-doors prosecutions held out of view of the public or proper legal scrutiny.

Rigged trials

Judge Johnson has done precisely nothing to counter the overwhelming impression that the Filton activists’ prosecutions were entirely political. He quite openly rigged both trials in manifold ways, as former British ambassador Craig Murray has set out

Let me quote from him at length the ways Judge Johnson flagrantly skewed the hearings to the detriment of the activists and justice:

Judge Johnson ruled that the defendants were not permitted to refer to their motives.

He ruled that the jury may not be informed of their absolute right to acquit.

He attempted to have the leading defence barrister, Rajiv Menon KC, prosecuted for contempt of court for informing the jury of their rights.

He ruled that terms including ‘genocide’ and ‘ethnic cleansing’ may not be used in court.

He ordered that the notebooks and other writings of the accused be redacted to withhold from the jury any information related to Elbit’s supply of weapons to Israel.

He enforced the concealment from the jury of the nature of the weapons and equipment that had been damaged.

He granted anonymity to senior Elbit staff and admitted their evidence without the defence being able to cross-examine.

He ruled that the trial had *not* been prejudiced by the Secretary of State [Yvette Cooper] and the Commissioner of the Metropolitan Police [Sir Mark Rowley] stating the offences as fact throughout national media.

He allowed the release to the media of highly edited and selective prosecution video footage during the trial which gave a false impression of events.

He permitted the admission of Metropolitan Police video evidence which they had given over to Elbit’s sole custody for an entire year.

In a trial with so many extraordinary anomalous moments, perhaps the most glaring was Judge Johnson’s efforts to get the main defence barrister in the first trial, Rajiv Menon KC, jailed for contempt of court simply for noting to the jury in his summing up speech that they had a hundreds-of-years-old right in law to acquit.

Jonathan Cook 4mo
If you are baffled by why the jury refused to convict the Palestine Action defendants, it's because – unlike them – you didn't hear the actual evidence. You heard what the media wanted you to know. Here, I explain what the jury learnt from the trial: https://jonathancook.substack.com/p/barristers-powerful-speech-at-filton
3838130

Judge Johnson’s behaviour was so unprecedented and certain to have a chilling effect on the ability of defence barristers to represent their clients – the Filton defendants decided to dismiss their barristers from summing up in their second trial to avoid their lawyers facing prison for doing their job and defending them – that the Court of Appeal had no choice but to overrule Judge Johnson’s contempt proceedings against Menon. 

Murray rightly concludes: "That is an astonishing list of nefarious actions by Judge Johnson. Read it again. Many people will surely conclude, it is Judge Johnson who should be in jail."

Gagging order

But even with the convictions for criminal damage secured under these rigged conditions, Judge Johnson is still in a position to cause more harm to the rule of law. He is due to sentence the Filton Four on Friday. 

Judge Johnson has reserved to himself the right to sentence the four anti-genocide activists not just for the relatively minor criminal damage charge they were convicted of after his rigged trial, but – once again in an unprecedented move – treat those criminal convictions as if they were for terrorism offences.

That means he can impose a longer sentence, more draconian prison conditions and more onerous, life-long conditions after their release. 

The jury knew none of this when they were considering whether to convict. Judge Johnson placed a gagging order on his decision during the trial which meant the information was withheld from the jury and could not be reported until after the verdict. The gag was broken only by foreign mediaand Zarah Sultana, who used her parliamentary privilege to reveal Judge Johnson’s government-friendly, anti-justice machinations. 

In yet another unprecedented feature to the trial, this will be the first time in British history that someone accused of criminal damage is sentenced as a terrorist. Judges were only given these extraordinary powers in a highly controversial amendment in 2021 to counter-terrorism legislation. 

Judge Johnson’s logic for taking advantage of his extra powers in this case is quite extraordinary too. He argues that, in destroying Elbit’s killer drones, the activists were seeking to “influence” the Israeli government to change its policy in Gaza – that is, to stop committing a genocide. 

Putting pressure on governments is what terrorists try to do, he argues, so this must mean the anti-genocide activists are terrorists.

Opposition to the genocide, in Judge Johnson’s view, has to be treated as an aggravating factor, not a mitigating one. 

It is all the more astounding that Judge Johnson is using this argument when he refused to allow the jury to hear any evidence of the activists’ larger political motivations: that they wanted to stop British complicity in Israel’s genocide by targeting a factory that made the drones for use in that genocide. 

Remember, all this is happening as Starmer’s government makes unprecedented moves to end many jury trials in Britain, leaving us to the mercy of judges like Jeremy Johnson. 

As Defend Our Juries notes, the government is looking to create “an extraordinary and deeply authoritarian precedent, allowing countless more protesters to be tried for an ordinary offence, but secretly sentenced as terrorists, without juries knowing this when they convict”. 

‘Cruel and vindictive’

Judge Johnson’s rogue manoeuvrings have so incensed the legal profession that thousands signed a petition demanding that he take the chance last Monday to recuse himself from the sentencing hearing. He, of course, refused to do so. 

They call his behaviour during the trial “biased” and “discriminatory conduct” and have referred him to the Judicial Conduct Investigations Office.

Here is David Whyte, professor of law at Queen Mary’s University of London, who handed the petition to the JCIO explaining another reason why so many legal professionals are outraged by Judge Johnson’s actions: he has shown exceptional “cruelty and vindictiveness” in holding the activists in remand for more than three times the normal maximums, even when the prosecution was not asking for them to be remanded, and for refusing them bail between their convictions and sentencing. 


The Court of Appeal is due to issue its ruling on the government’s appeal against the earlier High Court decision declaring Palestine Action’s proscription unlawful this coming Monday, days after Judge Johnson’s sentencing of the Filton activists. 

If this is all starting to look like theatre, that it because it is. In dictatorships, these are called show trials. Everyone understands that the outcome is predetermined. Everyone understands that justice is non-existent. The verdict is entirely political. It is a faux-legal rationalisation of what the security state wants. 

Judge Johnson is the perfect judge to play that part. 

The courts will do exactly what is expected of them unless they are worried that public revulsion will discredit their decision. Now is the time to raise our voices. 

Tuesday, June 9, 2026

Ken Klippenstein: U.S. Secretly Deployed Paratroopers to Israel

 U.S. Secretly Deployed Paratroopers to Israel

Quick Reaction Force specializes in 'forced entry'

Ken Klippenstein June 8 2026

82nd Airborne conduct a night attack exercise at Fort Bragg in July, 2025

I can’t do this work without your support - please become a paid subscriber! 

When the Pentagon announced that the 82nd Airborne Division was deploying to the Middle East in March, it concealed a key detail: some of the paratroopers were headed to Israel, as revealed in an Army deployment order I obtained. 

A military source involved in war planning tells me the deployment is tied to new U.S.-Israeli joint contingency plans, completed since February, for seizing Kharg Island and carving out coastal territory inside Iran. 

The 82nd Airborne Division is the Army’s premier quick reaction force, trained to parachute into hostile territory.

By keeping the deployment quiet, the Pentagon headed off public debate over a joint U.S.-Israeli operation inside Iran — a prospect many considered plausible at the time, amid a fever pitch of mainstream reporting on a potential ground invasion. The secrecy also sidestepped what's euphemistically called "host nation sensitivities." A joint U.S.-Israeli operation raises thorny questions for America's Gulf Arab "partners," especially over logistical support — hence the 82nd, which could launch directly from Israel without any Gulf state's consent to use its territory.

The Army deployment order, issued April 7, 2026, directs elements of the 2nd Battalion, 501st Infantry Regiment — the storied "Geronimo" battalion — to deploy to Israel on "temporary duty." The Israel deployment has not been previously reported. 

The Pentagon has never acknowledged it; in public it has said only that the 82nd was bound for "CENTCOM," the military's term for U.S. Central Command, the combatant command responsible for the entire Middle East. The press echoed the vague terminology, suggesting the unit was headed to existing U.S. bases in Kuwait or Qatar.

Asked about the number of troops deployed to Israel and their mission, the Pentagon referred my request to CENTCOM, which at the time of publication had not yet responded.

In late March, the New York Times reported that senior military officials were “weighing a possible deployment of a combat brigade from the Army’s 82nd Airborne Division … to support U.S. military operations in Iran.” The forces would come from the division’s Immediate Response Force — a brigade of roughly 3,000 soldiers able to deploy anywhere in the world within 18 hours. Those forces, the Times noted, “could be used to seize Kharg Island, Iran’s main oil export hub.”

The groundwork had been laid weeks earlier. The Army abruptly pulled the division’s 300-member headquarters from a planned exercise at the Joint Readiness Training Center in Louisiana, officials told the Times, so the command element wouldn’t be “caught out of place if the balloon went up.” The Aviationist reported that the division’s commander, Maj. Gen. Brandon Tegtmeier, and his command element had been ordered to deploy, and tracked a string of flights leaving Pope Army Airfield, which serves Fort Bragg, for the Middle East.

When the Pentagon finally did talk about the 82nd publicly, it took pains to keep Israel out of it.

At a May 5 briefing, Secretary of War Pete Hegseth and Joint Chiefs Chairman Gen. Dan Caine devoted much of their remarks to “Project Freedom,” the self-described “defensive” operation that the Trump administration says is keeping commercial shipping moving through the Strait of Hormuz. Caine singled out the 82nd Airborne for praise, telling reporters that elements of the division were synchronizing air operations for more than 100 aircraft over the Strait and “doing so in support of Project Freedom as we speak.”

“When the President or the secretary need immediate, scalable and lethal combat power in CENTCOM or elsewhere, the All American Division answers the call,” Caine said.

He went on to describe the paratroopers as "constantly ready to jump from Air Force aircraft into ground combat [to] seize key terrain if ordered to do so, just like their predecessors did in Sicily and Normandy in World War II, or to secure or enable the follow-on forces to flow into theater as they did in Grenada or Panama."

That is a description of readiness to seize a piece of Iranian terrain — which makes me suspect the New York Times and others were being used to "message" Tehran that such an operation was imminent.

“Beyond Project Freedom,” Caine said, “CENTCOM and the rest of the joint force remain ready to resume major combat operations against Iran if ordered to do so. No adversary should mistake our current restraint with a lack of resolve.”

So why the sensitivity about naming Israel, when the U.S. has acknowledged stationing American air defenses and F-22 fighters there? First, the breadth and depth of U.S.-Israeli military cooperation goes far beyond arms sales and “deconfliction.” Second, the two countries are deepening intelligence sharing aimed at making Israel a sixth “eye” in the so-called Five Eyes alliance of the U.S., Canada, Great Britain, Australia, and New Zealand. Third, naming Israel would expose how much of the U.S. assault on Iran is in fact a joint U.S.-Israeli attack — not a Trump-Netanyahu brotherhood, but cooperation at the military-to-military working level, where combined war plans are now reality.

Charlie Company of the 2nd Battalion is still in Israel, from which no dispatches have emerged in two months. I’m not saying a ground operation is imminent or inevitable. But if one were coming, wouldn’t it be nice to know?

Sunday, June 7, 2026

Iranian Official Answers Questions About Negotiations With the US

 

The most important aspect of this interview we need to understand, is that Iran is not asking the US to make any concessions. What Iran wants is for the US to cease hostilities and remove the punishing sanctions it has inflicted on Iran for decades. The US is also holding millions of dollars of Iranian money.

 

What the US wants is for its settler colony in the region to be the only possessor of nuclear weapons and and Iranian government that is controlled by the US.  The Zionist regime as the only possessor of nuclear weapons in the region, is, as John Mearsheimer and others have stressed, an existential threat to Iran. It’s most likely a threat to the entire region and as we are witnessing at the moment, a threat to the entire world. 

 

Israel, as a proxy for US imperialism in the region is and has been the most destructive and destabilizing force in the Middle East, neither have been serious about a Palestinian state and the geno*ide in Gaza and the increasing attacks on Palestinians in the occupied territories, and now the ethnic cleansing of Lebanon, proves that to be true.

 

The displacement of so many people in Lebanon will likely increase the number of refugees seeking a haven in Europe where right wing forces will foster anti-immigrant sentiment and blame them for the crisis of capitalism of which the wars in the Middle East are a part.

 

It’s staggering when one thinks about it. The US and Israel waged an unprovoked attack on Iran. They have bombed hospitals, schools, heritage sites, infrastructure. The US murd*red over 50 young schoolgirls by firing a Tomahawk missile in to their school. The US and Israel assassinated Iran’s leading cleric and a number of government officials.  International law, which only really applied to former colonial states and their officials is now dead.

 

Donald Trump’s popularity is at an all-time low and he is appearing to be suffering from some form of dementia. He started a war, by all accounts pressured by the Israeli war criminal Netanyahu, that he cannot end. In fact, the US has been humiliated by this war. The US bases that surrounded Iran were knocked out in a day or so. These cost the US taxpayer trillions of dollars. What we could have done with that much money. 

 

The US and Israeli governments are at this point the most despised governments on earth.

Saturday, June 6, 2026

Michael Roberts, AI: just one big trade

 

AI: just one big trade

by Michael Roberts

Goldman Sachs, the mega investment bank, reckons that AI is just “one big trade on the US economy”. And the AI investment bubble is getting even larger. In the last week, the AI model maker, Anthropic, announced that it was issuing shares to potential investors in what is called in stock market jargon, an Initial Public Offering (IPO).  Anthropic was following Elon Musk’s Space X planned IPO of a humungous $1.8trn.  This would value SpaceX in the market at 92 times its annual revenue! 

Alphabet, Google’s parent, also plans to raise $85bn in equity funding — its first stock offering in more than two decades. Together, these three giant IPOs could command a combined valuation of around $4trn. That’s one-third of all the value of US IPOs since 1980 (inflation-adjusted)! And yet SpaceX, OpenAI and Anthropic are all currently loss-making and the commercial potential of AI models and, in the case of Space X in going to Mars, remains unknown.

AI is one big trade for the US stock market investors and one big bet on the US economy.  That’s because the amount of capital investment being made by the companies called the ‘hyperscalers’ into AI models, data centres and other AI equipment is staggering. As a share of US GDP, it is now set to far surpass the 19th-century railroad build-out.

Back in December 1996, then Federal Reserve chair Alan Greenspan characterised the boom in technology, media and telecom stocks as showing signs of “irrational exuberance”. Almost 30 years later, we can say the same about the AI boom with bells on. This investment boom is already much larger than the dot.com internet investment of the late 1990s ever was. In 2025, US businesses invested almost $1.5trn in IT equipment and software. At the peak of the dot.com bubble, it was $466bn, or $829bn when adjusted for inflation. The hyperscalers Microsoft, Alphabet, Amazon, Meta and Oracle plan to invest hundreds of billions in the next five years in data centres to provide the computing power to run these AI models. Capital investments are expected to rise by 20 per cent a year, a growth rate never seen before in this industry.

US GDP growth is now driven almost exclusively by rising tech spending. If this starts to drop, the US economy will enter recession very quickly — even if tech investments decline only by a little bit, say 4 to 6 per cent, as happened after much smaller tech booms in the 1960s and during the 2009 recession.

As I showed in my last post, US corporate profits have risen significantly. But according to Brian Green in a recent post, around 80% of the increase in US non-financial corporate profits came from Nvidia and hyperscalers. The stock market is increasingly concentrated in a handful of AI‑linked stocks, which now account for roughly 40 per cent of the S&P 500’s market capitalisation, according to Bank of America data. Headline profitability is being flattered by a small slice of the economy earning extraordinary returns from the scramble to build AI capacity. The risk, then, is that the economy, the profit cycle and the stock market “are all leaning on the same narrow pillar. If the expected returns on AI infrastructure and platforms are questioned, the fallout may not stop at a few richly valued technology stocks.”  

As I have pointed out in previous posts, up to now the massive investment in AI has been mostly funded by the profits already being made by the hyperscalers. But given the impossibility of finding enough additional revenues to self-finance their capex plans, hyperscalers and their hardware providers are increasingly using external financing to fund them.

The first game is ‘circular financing’ ie by cross-investments between Microsoft, OpenAI, and others.  In essence, a cash-rich hyperscaler like Microsoft buys hardware from Nvidia, AMD and other suppliers. Nvidia then uses that revenue to buy a multi-billion-dollar stake in OpenAI. OpenAI then uses this cash to secure compute in Microsoft data centres.  Microsoft itself also invests in OpenAI and enters into a mutual revenue share where some of OpenAI’s revenues flow to Microsoft and vice versa as the two companies use each other’s products.  Assuming that Microsoft spends $100bn to order hardware for data centres, Nvidia, AMD and other suppliers can recognise this $100bn as revenues. They then use that cash to invest in OpenAI (for example), which then uses this money to book data centre capacity with Microsoft. Microsoft recognises this OpenAI investment as revenue, thus effectively turning its $100bn expense into billions of revenue!

Even this is no longer enough, and increasingly, hyperscalers have started to resort to borrowing to raise the cash for investment. The US tech giants are issuing debt all over the world. Google/Alphabet is leading the charge.

So first, they invested with their own funds; then in each other; then they borrowed from the banks and so-called private credit funds; and now they are putting the risk of success or failure on investors in the stock market.  If all this investment fails to deliver the expected returns, it will hit the financial sector and the wider economy big time.

But don’t worry, say the AI companies and hyperscalers, revenues are expected to grow 15 per cent annually. If we make the heroic assumption that there are no costs, then this additional revenue is the profit these companies are expected to make from their additional investments in AI data centres. Yet, even under these extremely optimistic assumptions, the implied return on investment is highly negative for all except Amazon.If the hyperscalers need to generate, say, a 10 per cent return on investment, they would have to find an additional $2-5tn in revenue a year. That’s a tall order for a group of companies that currently generates revenues of just $1.5tn per year. The other option is that the planned investment in data centres, computer chips and other areas never materialises — maybe as equity investors turn more cautious on the sector, or if debt funding for data centres becomes harder to get. A JP Morgan analysis found that more than 60% of data centre capacity planned for completion in 2027 isn’t yet under construction, and another 7% is delayed. What will happen if these companies announce cutbacks on some of their investment plans?

Will the AI heroes, OpenAI and Anthropic deliver the returns that the hyperscalers and their investors hope and expect? Corporate CEOs are optimistic. Over the last three years, since OpenAI launched ChatGPT, they claim that cumulative productivity gains have been in the order of 0.3% to 1% per year. For the next three years, they  estimate productivity gains to accelerate to 1.4%, with executives in the US  and UK far more optimistic than in Germany and Australia.

These productivity gains, they reckon, will be achieved by shedding labour.  Business leaders expect headcount in their firms to drop by about 0.7% in the next three years, again with executives in the US and the UK expecting far more pronounced drops in employment than executives in Germany and Australia. In the last three years, the same executives saw no employment impact from AI.  So this is all expectation. Moreover, the Business Trends and Outlook Survey of the US Census Bureau shows that companies with 50 employees or more show no more growth in AI use since Q2 2025. Businesses are still unsure how to use AI effectively and are increasingly worried about the drawbacks of AI when they use it.

Those drawbacks include ‘hallucinations’ (ie fictions made up by the AI model), which are inherent in LLMs. One study found that for a training set of 32,000 words, the average hallucination rate in LLMs was 6.8%.  When that was expanded to 128,000 words, the average hallucination rate rose to 10%.  That’s a lot of correction time and monitoring for human workers.

Another problem is that because LLMs are designed to be good at everything, they are not very good at any one thing compared to specialised apps. One report on using AI in software development found an explosive impact at the start, with coders creating or editing almost 300 per cent more files, but that boost was halved to 150 per cent by the time companies got the number of pieces of work submitted for review, and that in turn shrunk five-fold to a roughly 30 per cent uplift at the point of full software releases.

Moreover, when researchers looked at whether AI-assisted increases in software production have led to increased usage by clients, they found little evidence. The marked increase in mobile app releases over the past year has not been accompanied by any increase in downloads — most of the new apps fail to capture even a modest audience.

Meanwhile, OpenAI has burnt through some $6bn, rising to $17bn in 2026. By 2028, inference (training) costs alone are expected to grow to $121bn and losses are projected to be $85bn. Anthropic’s cash burn is much smaller, but was still $3bn in 2025. Unless the companies that build LLMs can find large amounts of new revenue in the next couple of years, the losses will increase exponentially, especially given the fact that current price charged per ‘token’ is not the true cost of compute.  If AI companies were to charge the cost price per token, the losses may decline, but demand for LLMs may decline even more. 

Despite this, the hype around AI remains so big that essentially all private investments in the US are now in tech hardware and software. Over the last three  years, the average annual growth in IT equipment investments has been 11% and 8% in software. Meanwhile, investments in all other parts of the US economy put together have declined by 1.6% per year.

The US economy today really is two economies in one. There is the tech economy and then there is everything else. Over the last four quarters to the end of Q1 2026, 93% of US GDP growth is due to tech investment alone (although much of the purchases are imports and not produced domestically).

This is a bubble waiting to burst. In the aftermath of the TMT bubble, private fixed investment dropped more than 12.7% between 2000 and the end of 2002 as a recession took hold in the US. In the initial year after the TMT bubble burst, tech investments dropped 12%, while fixed investments in general dropped 7.6%.

Gita Gopinath, former chief economist at the IMF, has calculated that an AI stock market crash equivalent to that which ended the dot-com boom, would erase some $20tn in American household wealth and another $15tn abroad, enough to strangle consumer spending and induce a global recession. This is also the view of the IMF. The IMF fears that AI firms could fail to deliver earnings commensurate with their lofty valuations. The collapse of previous investment booms knocked about 1 pp on average of US real GDP growth. Even a moderate correction in AI stock valuations would reduce global growth by 0.4%. “ Combined with lower-than-expected total factor productivity gains, and a more significant correction in equity markets, global output losses could increase further, concentrated in tech-heavy regions such as the United States and Asia.” Another study found that even a very mild drop in tech investment of just 3% would cut US real GDP growth by 1%, or half the current rate.  The impact would be greater in Europe.

None of this is to conclude that AI will not at some point deliver with higher profitability for the companies involved and higher productivity for the US economy as a whole.  But that will not happen before there is a bursting of the investment bubble – as there was in the railway mania of the 1870s and in dot.com bubble of the late 1990s. As other studies have shown, it will take a decade or more for AI to become a generalised technology that delivers.

For working people, AI poses a different problem.  For capital and the mega media companies, the aim is to make AI a profitable technology, but that can only be done by shedding labour and by stopping any attempt to regulate its applications and use. If AI is to succeed for capital, it will only be at the expense of most working people and their families.