TheChek News Flow

The hype around humanoid robots isn’t particularly new. Thank Tesla CEO Elon Musk and his Optimus robot, as well as the myriad videos of Boston Dynamics’ Atlas robot, for that.

Behind that hype, though, there is real progress. The physical capabilities of robots continue to improve, and researchers now believe that the AI techniques behind large language models can make complex robots capable of learning nearly any task.

Those tailwinds have encouraged a new batch of companies to jump in on the promise of profits from humanoid robots. And many of the latest entrants are Chinese automakers.

Earlier this week, Xpeng’s robotics unit raised more than $900 million at a post-money valuation of more than $6.3 billion. The round, led by IDG Capital with participation from Gaorong Ventures, Tencent, and Alibaba, was described by the company as the largest single-round private financing ever recorded in China’s “embodied AI” industry (AI systems built directly into physical machines).

This month, AiMOGA, the robotics unit of China’s Chery Automobile, reportedly began preparing for an IPO, while BYD unveiled a humanoid robot called Xiao Di. Other Chinese automakers, including Changan, GAC, Li Auto, SAIC, and Seres, are also developing humanoid robots.

Among all of them, Xpeng is the Chinese automaker that most closely watches and follows Tesla’s initiatives, according to Michael Dunne, CEO of San Diego- and Singapore-based advisory firm Dunne Insights.

“It’s the most focused on autonomy, it’s the first to commit in a big way to humanoid robots,” Dunne told TechCrunch, adding that Xpeng founder He Xiaopeng is a tech billionaire known for his agility and quick adjustments. “He sees razor-thin profit in cars on the near horizon. Robots look much more promising.”

Xiaopeng and Xpeng co-president Brian Gu are bullish enough that they’ve put their own funds behind the robotics unit. According to the Wall Street Journal, the pair invested about $100 million into the recent fundraising round.

Xpeng’s bet is on Iron, a humanoid robot with a realistic human shape that is built for commercial deployment.

Chinese automakers like Xpeng do bring a manufacturing edge.

“They have all the hardware to get the job done,” Dunne said. “Question is if they can catch Tesla on the AI side of the equation.”

There are, of course, many other companies developing humanoid robots, including Agility Robotics, Apptronik, and Figure, all chasing the same goal: commercial deployment at scale.

Hyundai-owned Boston Dynamics is getting closer to that goal. Hyundai plans to bring Boston Dynamics’ Atlas humanoid robot to its Georgia factory this year and eventually deploy the robots for tasks like parts sequencing by 2028. The Korean automaker, which partnered with Google’s AI research lab DeepMind to speed up the development of Atlas, is opening a U.S. facility this year called a Robot Metaplant Application Center, which will teach robots how to map movements like lifts and turns.

Other automotive companies are also jumping, including supplier Mobileye, which acquired humanoid robot startup Mentee Robotics earlier this year for $900 million. Even Rivian is dabbling in robots with its Mind Robotics spinout — although its robots are not expected to look quite like the humanoids in development elsewhere.

Gabriel Perez agreed to pay a penalty and give up winnings to federal regulator CFTC

WASHINGTON—The White House teleprompter operator under investigation for placing bets using proprietary information faces penalties from the Kalshi prediction market and federal regulators after reaching a settlement.

Under the terms of the settlement, Gabriel Perez, the teleprompter operator, will pay a $65,000 civil monetary penalty and give up more than $107,000 won in bets to the Commodity Futures Trading Commission. He will also be suspended from trading for three years.

Kalshi’s lawyer and head of government enforcement Robert J. DeNault said in a social-media post late Friday that Perez was also subject to penalties by the company as part of the settlement.

“It doesn’t matter who you are: violate our rules or federal law and you will face the consequences,” said DeNault, who cited the CFTC announcement.

The settlement comes after federal regulators probed Perez for making more than $100,000 in wagers based on early access to President Trump’s speeches.

“In his position, Perez had access to presidential speeches prior to those speeches being delivered and Perez misappropriated that information,” the CFTC said Friday. The White House didn’t immediately return a request for comment.

The federal agency said it appreciates Kalshi for its assistance with the probe.

Kalshi began to flag Perez’s bets earlier in the year and interviewed him, as part of an internal investigation. It also froze his account and kept more than $90,000 of his profits. The internal investigation was then referred to the CFTC.

The CFTC also disclosed that Perez made bets on at least a dozen markets related to Trump’s speeches since December, including predictions on what the president would say in an address to the nation and what places Trump would mention in his State of the Union. Trump addressed the nation in July and delivered his State of the Union in February.

“Before Perez traded each contract, he would read President Trump’s speech and place his trades based on the words he saw in the speech,” the CFTC wrote in a filing detailing Perez’s actions. “For example, he would only purchase a ‘Yes’ contract in a Trump mention market if he saw that the word on which the contract was based was contained in a copy of the speech President Trump would deliver. Similarly, Perez only placed ‘No’ trades on words he knew were not contained in President Trump’s prepared remarks.”

The CFTC said that Perez once changed his trading position on a prediction market based on his insight into Trump’s nonpublic prepared remarks.

The White House had previously warned staff against using their roles to make bets on prediction markets like Kalshi, including in a March memo from the White House Management Office. Then-White House press secretary Karoline Leavitt said that the White House was unaware of Perez’s bets and that the president was furious over the idea of individuals in his administration trading on inside information. Perez was placed on unpaid administrative leave.

Federal authorities are continuing to expand their crackdown on trading conduct in prediction markets, which offer a way for users to bet on political races, sports, corporate performance and crypto and commodity prices. Earlier this year, federal prosecutors charged a soldier—who took part in the U.S. operation to capture former Venezuelan strongman Nicolás Maduro—with using classified information to place winning wagers on Polymarket that Maduro would be removed from office. The winnings exceeded $400,000, according to prosecutors. He has pleaded not guilty.

More recently, federal authorities have been preparing charges against a U.S. servicemember who is suspected of placing Polymarket bets related to military strikes in Iran and Venezuela that earned him more than $1 million, The Wall Street Journal previously reported.

Federal authorities are also pursuing another case involving an employee at the global accounting and consulting firm KPMG. The employee is under investigation for betting on whether a specific public company would beat the consensus estimate for quarterly earnings. A KPMG spokesperson declined to comment. Charges in both cases could come this fall, although a final decision on bringing charges hasn’t been made, according to people familiar with the investigations.

US campaign to isolate Iran closes in on Dubai

Gulf city has for decades served as an offshore hub for Iranian financial services, trade and access to global markets that the regime needs

Along the main waterway in Dubai’s old commercial heart, the long stretches of quay, home to its historic maritime trade with Iran, are now largely abandoned.

Shipping activity involving dhows, the wooden vessels that for decades ferried goods from Dubai Creek to Iranian ports, has visibly faded. Further downstream, fishing boats are tied to the quays, but one of a handful of crewmen still present said they’re not allowed into the Gulf any more.

A dedicated terminal for dhow traffic at the mouth of the creek was showing more signs of activity, with forklifts and cranes loading pallets and small containers on to a dozen or so ships at sunset. But a security guard said the vessels were heading to Oman and India. “No more Iran since last week,” he said.

“Before the war, many Iranian cargo ships came here,” said Amal, the fisherman. “Now it’s finished.”

For decades, Dubai has served as Iran’s offshore commercial hub, providing the financial services, trade and access to global markets that Tehran’s sanctions-hit economy badly needed.

Sharing commercial and historical ties that preceded the founding of the United Arab Emirates in 1971, Dubai is home to several hundred thousand Iranian residents and merchants, as well as banks, trading houses and the dhow trade.

That is now under threat, after the UAE and US announced plans to target Iranian trade and bring about what Washington has described as an “economic D-Day” to force the regime in Tehran to its knees.

The UAE, which bore the brunt of Iranian retaliatory attacks after the US-Israeli war started in February, last week said it was suspending all trade, commercial exchanges and financial transactions, hours after accusing Tehran of targeting maritime traffic in its waters.

A day later, US President Donald Trump declared a new campaign of “economic warfare” against Iran and Washington launched “Operation Economic Outcast”, saying it would expand secondary sanctions against those trading with the republic.

This has left Dubai’s Iranian business community once again worried about their future. While China is Iran’s largest trading partner, “the type of trade that comes through the UAE really matters”, said Esfandyar Batmanghelidj, chief executive of the UK-based Bourse & Bazaar Foundation think-tank.

“The UAE is the only avenue for Iranian imports, as Iran has access to no other major container port other than Jebel Ali in Dubai. If you are trying to keep an economy of 90mn growing, you really need that infrastructure.”

Trade between the UAE — one of the US’s closest regional allies — and Iran amounted to nearly $30bn annually before the war, much of it enabled by the wooden dhows that shaped Dubai’s old city, where its spice and gold markets are also situated.

Analysts are sceptical whether the economic pressure campaign can succeed. But the six-month conflict, which has devastated Iran’s economy, and the shuttering of the Strait of Hormuz have taken a toll on traders in the old city.

Danial, an Iranian shopkeeper who declined to give his full name, still sold saffron from Iran out of stocks that were purchased before the war began. But as shipping traffic from Iran had slowed, Iranian businessmen now brought smaller boxes of saffron in their luggage, he said, raising the luxury spice’s price by more than one-fifth.

The impact of the war on business was evident in the souk near Danial’s shop. At least half a dozen outlets, some owned by Iranians but also Afghans, had been shut, with windows carrying signs announcing they were “closed for maintenance” — echoing the luxury hotels that closed their doors for long-term renovations earlier in the conflict.

“We have no income, yet we still have to pay rent,” said Danial. In a sign Dubai officials might be tightening the regulatory screws on Iranian merchants, commerce inspectors in recent weeks fined him and several other shops for what he described as small bureaucratic infractions. “They’ve become stricter, 90 per cent of all of us got a fine,” he said.

Since the start of the war, the UAE has absorbed more Iranian attacks than any other Gulf state, with almost 3,000 missiles launched at the country, causing a slump in tourism and damaging key infrastructure.

This fuelled a hawkish stance in the UAE, which even launched secret air strikes against Iran earlier in the war, and Dubai early on closed an Iranian-linked club and hospital.

Some trade resumed — including the dhows and Iranian flights to Dubai — as Abu Dhabi and Tehran sought to de-escalate tensions after a shortlived US-Iran interim deal in June that was supposed to reopen the strait.

But that diplomatic effort was upended after Iran last week fired missiles towards the UAE, which it had not targeted directly since mid-May.

Still, Batmanghelidj said the UAE “has an amazing ability to pursue its interests through economic diplomacy”.

“Over the last 20 years they have modulated their level of economic engagement with Iran in different periods,” he said. “But they’ve never really allowed US sanctions to fully determine how much trade they do.”

Some Iranian nationals Danial knew had decided to leave the UAE, not only because of concerns over whether they were still welcome in the country where many were born or had lived their entire life, but “most of them are bankrupt”. He, too, was making plans to pack up and leave.

But many argue that Iran’s commercial relations with Dubai are too deep to be uprooted entirely. “Iranian traders have been a foundational component of Dubai’s economy since long before the oil era,” said Justin Alexander, director of consultancy Khalij Economics and a non-resident fellow at the Baker Institute for Public Policy.

The US Treasury department on Friday announced that it would cut the UAE branches of Banque Misr, Egypt’s second-largest bank, off from banking access to US financial institutions as part of its economic crackdown on Tehran.

Merchants in Iran sounded a defiant note and appeared prepared to face any further restrictions on their activities. The country has spent decades setting up alternative commercial routes and channelling payments through exchange houses or shell companies designed to withstand sanctions.

“We have put a lot of goods in warehouses in Turkey now in case the UAE route is closed again,” said a Tehran-based businessman. He said the Iranian dhow trade with Dubai appeared to have been curtailed, but said some shipments were rerouted through Oman. The US has maintained a naval blockade on Iranian ports since mid-July.

However, the businessman said he was able to load cargo on to ships in Dubai bound for Iran’s Bandar Abbas port by way of entering fake destinations on official documents. The cost of shipping goods had soared, but was still lower than at the height of the war, when the UAE was subject to almost daily Iranian attacks and few ships were moving through the Strait of Hormuz.

“We are bypassing the blockade in the same way we have been bypassing sanctions,” the Iranian businessman said. “Not much has changed.”

Dubai’s media office, the Emirati foreign ministry and the port authorities did not respond to requests for comment.

Elsewhere in Dubai, Iranian businesses carried on as usual. A fruit and vegetable vendor at the large indoor waterfront market sold Iranian produce — lettuce, sweetcorn, tomatoes and almonds — that, according to him, arrived via the established maritime channels. “Sometimes it’s a bit less, sometimes a bit more, but now it’s OK,” he said.

The Iranian Business Council, a non-profit promoting bilateral commerce, is still operational, according to an employee reached by phone. Travel agents in Tehran said this week that Iranian airlines were still flying to Dubai.  

And two leading Iranian lenders, Bank Melli and Bank Saderat — both under US sanctions — were still open serving clients. US Treasury secretary Scott Bessent called for every Melli branch in the world to be shuttered. The two Iranian banks combined have more than a dozen branches in the UAE.

An employee of Melli said some of their clients, whether from Iran, Russia or local, were initially worried about the fate of their deposits but there had been no noticeable impact on day-to-day financial transactions so far, including lending and securing letters of credit. The bank recently renewed its licence with the UAE Central Bank, she said.

“Business is down, but the whole world is affected,” the Bank Melli employee said. She also said she had worked for more than two decades at the bank and remembered previous rounds of sanctions. 

“We’re used to this,” she said.

Airbus seeks US space unit sale as it shifts to European-built satellites

Potential disposal comes after company announced deal with Leonardo and Thales

Airbus is exploring a sale of its US space business as the Franco-German planemaker focuses on building a pan-European satellite champion through a three-way merger.

The company is gauging interest from prospective buyers for the US business, according to people familiar with the matter. Airbus US space systems include a factory in Florida and the Arrow family of small satellites that can be mass produced for both commercial and government communications and imaging, including national security needs.

Airbus in 2024 took full control of its Airbus OneWeb Satellites (AOS) joint venture after acquiring the 50 per cent stake it did not already own. Established in 2016, the venture had built more than 600 satellites for OneWeb’s first-generation constellation, a low-orbit network that can provide space-based internet coverage, according to its website.

Airbus has been shifting satellite production back to Europe from the US for some time, including building 440 low Earth orbit satellites in its Toulouse facility, which it called a “further step for European sovereignty”. The orders have come from France’s Eutelsat, which acquired OneWeb of the UK out of bankruptcy in 2022.

While the company does not publish revenue figures for the US space business, two people familiar with the unit estimated that it had several hundred million dollars in sales. The space business employed more than 200 people at its Merritt Island facility on Florida’s Space Coast near the Kennedy Space Center and close to Jeff Bezos’s rocket business Blue Origin. That compares with about 11,000 employees worldwide supporting space activities at Airbus.

The potential sale of the US assets underscores Airbus’s efforts to turn around its struggling global space business, where it booked €989mn in charges in 2024. The company in October struck a deal with Leonardo of Italy and France’s Thales to combine activities spanning satellite manufacturing to space systems and services to better compete with Elon Musk’s SpaceX and Chinese rivals.

Codenamed Bromo, the three European parties aim to submit the plan to antitrust authorities in Brussels later this year with a view to making the new company operational in 2027. The firm has been reviewing its US space footprint for some time and the divestment plans are not related to Bromo’s talks with the European Commission around potential remedies to secure competition approval, some of the people said.

While the planned three-way combination has drawn criticism from other European players, including Germany’s OHB and Spain’s Indra Space, the three companies are betting on growing calls in Brussels for European champions.

The proposed tie-up comes as Europe’s space industry faces growing pressure from US and Chinese competitors. Brussels is putting greater focus on increased European sovereignty in the space sector, as the bloc’s members seek to end their dependence on the US by building satellite constellations for reconnaissance, intelligence and communications.

European satellite makers have struggled to adapt to the revolution in satellite demand, starkly demonstrated by Ukraine’s reliance on SpaceX’s Starlink in its war against Russia.

Airbus declined to comment on “market rumours or speculation”.

Nasa to launch $4.3bn space telescope to scan universe

Roman Space Telescope is expected to help explain dark matter while discovering thousands of planets beyond our solar system

Nasa is launching a new observatory designed to scan vast swaths of sky, potentially discovering tens of thousands of distant planets and shedding light on the mysteries of dark matter and dark energy.

The Roman Space Telescope, named after the late Nancy Grace Roman, Nasa’s first chief astronomer and a leading architect of the agency’s space observation programme, is scheduled to launch on Sunday. The $4.3bn mission will complement the Hubble and James Webb space telescopes, launched in 1990 and 2021.

Roman stands out from other space observatories for its very wide field of view which will enable it to scan large areas of sky in infrared light. “Where Webb goes deep, Roman will go big,” said Julie McEnery, Roman project scientist with Nasa.

At any moment Roman will see 200 times more of the sky than Hubble and it can move far faster between images. “A survey that would take Roman one month would take a century with Hubble,” McEnery said. “Roman doesn’t just do Hubble science faster. It can address entirely new, ambitious science questions.”

One of the mission’s main purposes will be to assess the effects of dark energy and dark matter on the shape and distribution of billions of stars and galaxies. The nature of both is unknown. Cosmologists believe dark energy — a force driving space apart — makes up about 68 per cent of the universe, while invisible dark matter accounts for 27 per cent and all the ordinary matter that telescopes can image directly is just 5 per cent.

Another task for Roman will be to search for planets beyond our solar system known as exoplanets. Nasa expects the telescope to discover as many as 100,000 exoplanets to add to the 6,300 currently known, using several different techniques.

Beth Biller, an astronomy professor at University of Edinburgh, will be working on observations with Roman’s coronagraph, an instrument that blocks the glare from a star and makes it possible to image its individual planets and their atmospheres directly from the reflected starlight.

“I’m excited that in the next year we may have our first reflected-light images of exoplanets — and I think that’s how we will eventually find planets like our own Earth,” she said.

The coronagraph will be developed further for the Habitable Worlds Observatory, Nasa’s next flagship telescope mission, which is expected to launch in the 2040s to look for life on Earth-like planets.

Roman is scheduled to launch on Sunday at 7.26am local time from Kennedy Space Center in Florida on a Falcon Heavy rocket made by Elon Musk’s SpaceX.

Its destination 1.5mn km away is the “second Lagrange point” L2, where it orbits the Sun at exactly the same speed as Earth and can always point out into deep space without being blinded by solar or terrestrial light. The James Webb Space Telescope is also located there, in contrast to Hubble, which orbits just 540km above Earth.

“L2 is a very dark part of the solar system,” Biller said. “You’re always pointing away from the Sun and the temperature is stable, so it is an excellent observing point.”

The project has an unusual history, starting with the transfer to Nasa of a surplus telescope by the US National Reconnaissance Office, which makes spy satellites. The space agency was able to repurpose it for the Roman mission, to look out into the sky rather than down at the Earth.

Unlike most big science projects, Roman has come in within its planned $4.3bn budget and ahead of schedule. The original target launch date was May 2027.

Scientists around the world are waiting eagerly for Roman’s observations to emerge over the next few years. “I suspect its greatest legacy may come from the way it surveys the sky,” said Noelia Noël, an astrophysicist at the University of Surrey in the UK.

“When you observe an enormous area of the sky repeatedly with this level of depth and resolution, you create conditions for discovering phenomena that no one designed the telescope to find,” she added. “What excites me most is the prospect of unexpected discoveries coming from questions we haven’t thought of asking yet.”

Glencore threatened with $1.4bn lawsuit from embattled trader Radiant

Singapore-based iron ore company Radiant has been battling liquidity squeeze

Embattled iron ore trader Radiant World is threatening legal action against Glencore, alleging that its larger rival is responsible for over $1.4bn of losses as part of a wider dispute between the two groups.

Singapore-based Radiant, which grew rapidly from relative obscurity to become one of the largest iron ore traders in the world, has been battling a liquidity squeeze in recent weeks.

Banks and trading firms including Glencore cut the group off over alleged inconsistencies in the documentation underpinning its trades. Japanese lender Mizuho — which was one of dozens of global banks that lent money to the business over the years — on Friday brought a legal challenge to Radiant World in Singapore’s Supreme Court.

Singapore’s police also confirmed to the FT that they were carrying out an investigation into Radiant — which is headquartered in the city state — after receiving complaints about the business.

Radiant, which has denied any wrongdoing, sent a legal letter to Glencore on Thursday seeking to recover more than $800mn that Radiant and another entity, Sapphire Minmetals, claim they had already paid to the Anglo-Swiss metals and mining group in recent years, as well as losses from “the destruction of the value” of its business. 

“These claims are meritless and Glencore will vigorously contest them,” Glencore told the FT, adding that it “has incurred losses and been exposed to risks by Radiant’s actions and will take appropriate action”.

Glencore said this month that it had stopped doing business with Radiant and had taken a provision relating to its contracts with the trader.

Radiant’s letter describes a complex trading and financing relationship with the London-listed commodities trading giant stretching back years that came to a head in 2021, when Glencore allegedly decided to close out derivatives positions and demanded repayment of almost $1.2bn in exposure it said Radiant and Sapphire had racked up under the derivatives contract.

Radiant previously owned a stake in Sapphire but disposed of its shares in the group in 2015, according to Hong Kong corporate records. Radiant described Sapphire as a “separate legal entity” in its letter to Glencore, while Sapphire stated in response to a Bloomberg News article earlier this month that a “historical shareholding does not imply operational or financial co-relation today”. 

Radiant has alleged that the parties reached an agreement over the derivatives exposure, which allowed them to gradually pay it down while Glencore assisted the Singapore-based trading firm in raising outside financing and investment. 

In its letter, Radiant cited a WhatsApp message allegedly sent by senior Glencore executive Peter Hill in 2024 outlining the support it had offered the iron ore trader. 

“Remember — we are the guys that you were allowed to build up $1.2bio [sic] exposure to and basically bank rolled your entire existence for the last few years,” Hill’s alleged WhatsApp message states. “So give me a break and relax, we will always support you as we have done and continue to do so.. we will do so to the extent it makes sense for us.”

Hill declined to comment.

Radiant also alleged that its relationship with Glencore went beyond that of a counterparty and described the commodities group as “a senior partner in the relationship”. It claims that Glencore reviewed and approved the qualifications of potential hires, advised the company on fundraising, and suggested the price at which it should enter into trades.

The letter also alleged that Glencore struck a deal under which it subscribed to a 4.995 per cent equity warrant in Radiant and subsequently communicated that it would “assist in bringing hedge funds and other investors” to finance the group. Glencore allegedly forecast that Radiant’s valuation could reach around $5bn through this additional investment.

Radiant claimed that Glencore recently served a termination notice to the Singapore firm and Sapphire, alleging that they owed the London-listed group $951mn, set off against $471mn that Glencore owed back to the two firms.

“The exposure on our books [to Radiant] is not material and well below our threshold of $500mn,” Glencore said earlier this month. That figure takes into account the provisions Glencore has already booked.

A person familiar with the matter said that Glencore’s net exposure to Radiant had always been under $1bn, and on average over the past decade Radiant had accounted for less than 10 per cent of Glencore’s iron ore business in volume terms. 

The spending on artificial intelligence will end, but probably not when investors suspect it will.

When it comes to a capital spending bubble, the trick isn’t dancing when the music is playing, but knowing when it’s time to sit down. As summer turns to fall, investors should keep on dancing to the AI beat.

With just four more months left in 2026, we understand the desire to step away. The S&P 500 has gained around 12% this year, but only slightly higher than it was at the beginning of June. The longer it remains stuck, the easier it is to focus on what could go wrong. And no issue looms larger than the massive amount of money companies are spending on artificial intelligence right now.

The scale of AI capital expenditures has been astounding. Since the start of 2024, roughly $500 billion has been spent on chips, $350 billion on power infrastructure, $200 billion on construction, and $100 billion on networking. That dwarfs the roughly $575 billion of capex from the entire S&P 500 in 2021, the year before OpenAI’s ChatGPT launched.

As long as companies were paying for AI out of free cash flow, no one seemed to mind. But now, the cash-flow machines of old— Amazon.com, Alphabet, Meta Platforms, and Microsoft, among others—are going into hock to pay for their AI data centers, and the worries are growing. And for good reason. The amount they aim to spend is staggering —some $2 trillion over the next two years—and concerns are starting to show up in debt markets, where the cost to insure against a credit default has skyrocketed.

Wall Street is also fretting about the stocks. Shares of AI companies have dropped 20% from a June 52-week high. It’s enough to make an investor ask if the end of the AI trade, if not the AI buildout, is upon us.

History suggests otherwise.

Capital spending bubbles almost always pop, but rarely when investors suspect they will. That has been the case for much of U.S. history, starting with the railroad buildout, through the dot-com boom, and beyond. While the AI capex explosion is almost certainly going to end badly, the chances of it ending badly now are low. Over America’s 250-year history, the economy has absorbed spending on transformation technology equivalent to about 25% of economic output before things get really hairy, according to a Barron’s analysis. Artificial intelligence isn’t there yet, and won’t be for years, suggesting that the AI trade—and the stock market rally—has more room to run.

“I think we’re going to just keep going,” says Melius Research’s head of technology research, Ben Reitzes. “This will be bigger than anybody thinks, for longer than anybody thinks.”

America has a long history of capex booms and busts. From the railroad expansion of the 1860s, through the electrification boom of the 1920s, to the dot-com and housing bubbles of the 1990s and 2000s, they have followed a similar pattern, according to Raymond James strategist Tavis McCourt. A new technology arrives, one that is met with enthusiasm by investors, pushing up valuations, and ultimately leading to equity and debt raises and market concentration issues before turning to bust. The booms don’t just end on their own, however. McCourt notes that they require an outside shock to spur the decline, one that can be policy-driven, like a Fed interest-rate hike, or an event like the San Francisco earthquake of 1906.

Historically, these cycles have followed enough of a pattern to consider where artificial intelligence falls in the cycle. One could be dubbed the rule of 25, for the amount of total spending the U.S. has been able to digest during transformational booms as a percentage of the overall economy. At the start of the railroad boom in the 1860s, for instance, U.S. gross domestic product was about $10 billion a year, according to the National Bureau of Economic Research, while rail spending eventually totaled $2.5 billion before the 1873 panic arrived. Similarly, about $1.5 trillion was spent building internet infrastructure in the late 1990s, while the U.S. economy was only $6 trillion at the time. The same holds true for the industrial and electrification buildout of the 1920s.

Using that standard, it’s possible to generate a top-down estimate of how much can be spent on AI before the economy is ready to collapse under its own weight. With U.S. GDP at roughly $30 trillion, the danger zone sits at about $7.5 trillion, or an additional $5 trillion to $6 trillion in domestic AI spending. Hyperscalers are projected to spend $3.7 trillion globally through 2029, and they aren’t the only ones, with Oracle, SpaceX, Anthropic, and OpenAI, among others, building, too. At this rate, AI spending won’t trip the rule of 25 until the early 2030s, six or seven years into the boom.

Even when it does, not every capex bubble ends up popping. For a capital spending boom to succeed, the outlays need to slow while profitability ramps up, Raymond James’ McCourt notes, allowing the businesses to become self-financing. McCourt acknowledges the potential for the AI buildout to become sustainable, but notes that it’s far more difficult without government backing or a relatively slow pace. “Historically, this smooth ramp to profitability without any hiccups in capital spending has never happened before at this scale, except for major government-funded capital spending booms,” McCourt explains.

And that’s where things get tricky. The hyperscalers—Meta Platforms, Microsoft, Amazon.com, and Alphabet—spent roughly $125 billion on new plants and equipment in 2021. In 2026, that number will be closer to $700 billion. In 2027, it will be $1 trillion, and about half of total capital spending for the S&P 500. That money is the flywheel driving the stock market, becoming revenue for just about everyone—from construction companies that build the data centers; utilities that provide the electricity to power them; electronics manufacturers who build the cables, connectors, and coolers to keep them running. And then there are the chip makers, whose products drive the computing power that makes AI work. Case in point: Nvidia, which Wednesday reported better than expected second-quarter earnings—and forecast a whopping 70% increase in revenue for fiscal 2028.

There are signs that the spending is becoming more profitable. Alphabet’s cloud business grew 82% year over year in the second quarter. Anthropic’s current annual revenue run-rate—Wall Street’s favorite metric for AI start-ups—is currently estimated to be north of $65 billion, up from $9 billion at the end of 2025. That means Anthropic has added business roughly the size of 50-year-old Oracle in eight months. SpaceX, which subsumed Elon Musk’s xAI in February, generated $3.2 billion in 2025 AI revenue, and will likely surpass $60 billion in 2027. Much of that revenue is coming from third parties, including Anthropic and Alphabet, which rent SpaceX’s terrestrial data-center capacity for billions a month. It might also be the right kind of profitability: While the S&P 500 needs $2 to $3 in assets to generate $1 in sales, a gigawatt of computing power costing $40 billion to build can produce $40 billion in sales. That is an epically good business.

But funding $6 trillion for AI capex in a handful of years is asking a lot of U.S. capital markets. Yes, they handled Alphabet’s $85 billion equity raise in June, and SpaceX set a record for a new public-company offering by raising $86 billion in its IPO and another $25 billion from debt markets just 10 days later. Others have had to get more creative. Nvidia will provide up to $100 billion in financing to OpenAI so it can use Nvidia graphics processing units, or GPUs, in up to 10 gigawatts of AI computing power. OpenAI, Oracle, and SoftBank launched the Stargate Project, an AI infrastructure funding project with commitments of some $500 billion.

Even-more opaque financing methods are popping up, as well. Meta Platforms created a special purpose vehicle called Beignet Investor to build five gigawatts of AI computing capacity in Richland Parish, La., with about $27 billion in debt and $3 billion in equity. Meta will own 20%, with Blue Owl Capital and other private-equity players owning 80%, but Meta will be the one renting the data center and paying the operating expenses, and it will be on the hook for the debt if it decides not to renew the lease. No wonder Beignet’s debt is rated A+.

All told, funding “extraordinary growth” will require everything from public equity markets to private credit, says J.P. Morgan Securities analyst Tarek Hamid. He projects more than $5 trillion in AI data-center spending from 2026 to 2030, with $1.5 trillion funded from cash flow, $500 billion from new equity, $2.5 trillion from credit markets, and $1.4 trillion from alternative sources, like Meta’s Louisiana deal, and governments.

With so much depending on financing—and untraditional financing at that—the bigger risk might be Fed rate hikes. Yields on long-term Treasuries are already near their highest levels since 2007, and the worry is that rate hikes could send yields higher across the board. That would make financing the AI buildout more expensive. “It stands to reason that sensitivity of AI buildout names to 10-year yields will increase meaningfully from here,” writes 22V strategist Dennis DeBusschere. “We should expect AI buildout names to suffer, all things equal, if 10-year yields keep grinding higher.”

Debt investors have taken notice. While bond spreads—the difference between the interest a company pays on its debt versus an equivalent Treasury—remain tight, credit default swaps, which act as a form of insurance, have surged higher. It now costs about twice as much to insure against an Alphabet default than it did at the beginning of the year, and it’s a similar story for Meta and Amazon. Though the rates are well below concerning thresholds for investment-grade companies, the moves signal that markets recognize that Big Tech companies won’t be the debt-free juggernauts they had been.

The balance sheets are still sound. The four hyperscalers still have essentially no net debt and are expected to generate roughly $1 trillion in earnings before interest, taxes, depreciation, and amortization, or Ebitda, in 2027.

Debt levels aren’t a problem for 2028. For now, the focus should be on finding the companies that have the best chance of benefiting from the continued buildout. Choosing an outright winner is next to impossible. But there will be outright AI winners. They could be SpaceX, Tesla (or a SpaceX/Tesla combination), Alphabet, Anthropic, and OpenAI. The safest of the AI plays is clearly Alphabet. It has an existing business that is pretty good, with operating income in 2026 projected to be north of $170 billion. Shares trade for a not unreasonable 20 times earnings expected over the coming 12 months, about in line with the S&P 500.

But infrastructure plays might be even safer, especially as demand shifts. If the first iteration of the AI trade was simply about building the best chips and sticking them in CPUs, the second will focus on building “AI factories,” according to Evercore analyst Amit Daryanani. Nvidia’s Rubin is designed to do just that by working across compute, networking, and storage, while companies like Meta Platforms and Microsoft are creating tools to minimize the cost of using AI. “We think this broadens the AI infrastructure opportunity…toward memory, networking, power, cooling, and software,” he writes.

Daryanani cites Amphenol and TE Connectivity as two stocks that could benefit from the trend toward more complex connectivity needs. Other popular stocks include Eaton, which provides power distribution and cooling solutions for data centers, among other products, and Vertiv Holdings, a leader in data-center hardware and services. Both are more expensive than they used to be—Eaton trades for about 28 times earnings, up from 25 times two years ago, while Vertiv changes hands at 32 times earnings, up from 27 times—though both are still well liked by analysts: 78% of analysts rate Eaton a Buy, while Vertiv is 85% Buy-rated.

The data-center backlash is another risk—but also an opportunity. While demand for AI is rising, the pushback is likely to slow the buildout of new data centers. That should boost shares of companies that already have projects approved, including SpaceX, CoreWeave, Nebius, and Fermi, according to 22V’s Dauvin Peterson. CoreWeave, in particular, looks interesting. While the market remains focused on its debt, higher AI computing prices would boost earnings—and the stock.

The good news is that when the bubble pops—and it will pop—what was built will remain. As with the railroad, the internet, and electrification booms, it will take time for financial markets to recover. But the foundation will already be in place for the next expansion—and for the dance to begin yet again.

Institution will ‘always remain open for dialogue’ around contested artworks, says Nicholas Cullinan

The deal that brought the Bayeux Tapestry to the UK could pave the way for a loan of the Parthenon Sculptures to Greece, the director of the British Museum has said, reviving the prospect that one of the world’s thorniest arts disputes could finally be resolved.

Nicholas Cullinan told the FT that the return of the near-1,000-year-old embroidery, which will go on display on September 10 after a deal between Britain and France, was “a very compelling example of international partnership” that pointed to the success of “reciprocal exchange”.

“Every example of a relationship or cultural history is different, so there isn’t a cookie-cutter template,” Cullinan said in an interview.

“That said, you would hope the fact that it took several goes to pull off the Bayeux Tapestry loan might provide a glimmer of possibility to move forward [on the sculptures].”

“I would hope that some tangible progress can happen,” he added, of the 2,500-year-old carvings. “It would be a wonderful thing.”

The sculptures, often known as the Elgin Marbles after the Scottish aristocrat who brought them to Britain from the Acropolis in Athens in the early 19th century, have long been contested.

Greece made its first formal request for the permanent return of the sculptures in 1983, and talks picked up after former Conservative chancellor George Osborne became chair of the British Museum in 2021.

As prime minister, Sir Keir Starmer made clear he would not block a loan accord, under which parts of the frieze would be shipped to Athens for specific periods while significant Greek antiquities were shown in London, if the two sides could reach agreement. But Greece has insisted on ownership.

His successor Andy Burnham told the FT this week that the future of the sculptures was a matter for the British Museum, despite previously calling for the artefacts to be returned to Greece permanently.

Approximately 1,630 other objects from the museum’s 8mn-strong collection were on long-term loan nationally last year, and Cullinan said the museum would “always remain open for dialogue” around the sculptures, acquired in 1816.

But the former head of the National Portrait Gallery and Tate curator warned of “unintended consequences” if the British Museum Act 1963 were amended to allow artefacts in its possession to be removed.

“As a museum person, I think you’re always thinking about being a temporary custodian of these objects . . . you never know how future generations will perceive what you’ve done,” he said.

“I’m glad that, when the pandemic began in 2020, museums couldn’t begin a fire sale of their collections, because that definitely would have happened if the law had not prevented de-accession.”

Three years ago, revelations of the theft of about 2,000 items including ancient semi-precious stones rocked the museum’s reputation and prompted the exit of its previous chief.

Now Cullinan, 48, and his team see the Bayeux Tapestry as a coup of cultural diplomacy that should draw positive international attention.

The display of the 70-metre fabric depiction of the Norman Conquest, thought to have been made by nuns in England in the 11th century, will be the biggest and most expensive exhibition in the institution’s 273-year history.

Running until next July, the show will coincide with the transfer of treasures from the Sutton Hoo burial ship and the Lewis Chessmen to France. 

Before his death in June, the painter David Hockney accused the museum of “vanity” in moving the tapestry from Bayeux, echoing concerns voiced by some in the French art establishment about its fragility. 

Cullinan said the tapestry was in “fantastic condition” and would not have travelled — via the Channel Tunnel last month, under police escort — if “people with intimate knowledge and expertise” had “raised a flag”. 

Other critics have hit out at entry prices to the exhibition, which is being sponsored by hedge fund billionaire Igor Tulchinsky in a deal valued at £5mn.

A standard adult ticket costs £33, while under-16s go free. The first allocation sold out in just over a day last month, yielding more than £2.5mn.

Cullinan said the museum had “thought long and hard” about prices and hoped to turn a profit on the exhibition.

“We could probably have tried to accommodate more visitors, but quality has been as important as quantity,” he said.

The institution, which this month celebrated the 20th anniversary of its international training programme for cultural heritage professionals, welcomed 6.5mn visitors last year, most from overseas. It hopes to draw more with a new entrance pavilion and garden from next spring. 

Asked about ministers’ proposal to charge overseas tourists to access venues such as his, Cullinan said ending the policy of free admission was “not something to be done lightly”. 

The museum raised £2.5mn through its inaugural ball last October. It is now stepping up efforts to attract donors to its £1bn redevelopment of the Western Range galleries, which is set to last about a decade.

During the renovation, some artefacts housed in the complex network of 10 separate buildings, which include the Parthenon Sculptures, “may be loaned as part of our commitment to increase national and international loans”, the museum has said. 

Detailed designs for the project, led by Beirut-born architect Lina Ghotmeh, would become public by 2028, Cullinan said, with a bigger “master plan” aiming to create space for “entire sections of the collection, like south-east Asia and the Caribbean” never exhibited before.

A major gift towards the renovation would be announced in the autumn, and “conversations are happening” about donations in the region of £150mn, which would match record-breaking sums secured by the National Gallery in 2025, he added. 

“The transformation of this building is a generational change, and it brings huge potential to get more of the collection out in the world and rethink the role of a universal museum,” Cullinan said.

“It’s a big, collective endeavour, and it’s not an overnight thing. We’re in the limbering-up phase.”

Gapping down
In reaction to earnings/guidance:

Other news: