TheChek News Flow

Startups that enable people to bet on events such as the New York mayoral race, NFL games and the Federal Reserve rate decisions are in talks to raise more money at multi-billion-dollar valuations, as trading volume and rivalry between the two apps intensifies.

Polymarket has considered an offer that would value the company as much as $9 billion, according to people who have spoken to the company’s leaders. That would be a huge jump in price for the startup, which raised money at a $1 billion valuation earlier this summer and is on the cusp of re-entering the U.S. market, but doesn’t yet generate revenue.

Meanwhile, older rival Kalshi is close to raising money at a $5 billion valuation including the new money, which would more than double the $2 billion valuation the startup raised at just three months ago, according to two people who have spoken to its leaders.

The fundraising discussions follow growth in trading volumes for event contracts—derivatives that pay out to investors who guess the event outcomes correctly— after the markets gained attention ahead of the 2024 U.S. presidential election. Both startups have recently moved to expand the ways customers can buy their contracts.

Polymarket earlier this month said the Commodity Futures Trading Commission had given it the go-ahead to serve U.S.-based users. The agency had barred it from accepting U.S.-based trades three years ago, saying it was running an unregistered derivatives exchange. In July, the five-year-old startup bought a U.S.-registered exchange and clearinghouse, which helped pave the way for the CFTC approval.

Trading volume on Polymarket hit $1 billion in August, more than twice its volume a year ago, according to Dune Analytics, which tracks activity on blockchains. Polymarket clears trades using cryptocurrency, saving the platform the trouble of coordinating with banks and payments providers in each country. Customers use Circle’s USDC stablecoin to buy contracts predicting outcomes of political, sporting or global events. They receive a payout if they predict correctly.

Polymarket doesn’t generate revenue but could at some point by charging fees on trades. It has also discussed launching a token, though there’s no timeline on that yet, The Information reported last year.

Kalshi has tried to distinguish itself by touting the fact that it already operates with U.S. regulatory oversight. Trading volume on Kalshi hit $875 million in August, according to an analysis of Kalsh’s public data compiled by a Dune Analytics user.

Kalshi generates revenue from charging a transaction fee on the expected earnings of a contract. In August, it announced a partnership with trading app Robinhood to offer contracts predicting the outcome of NFL and college football games on the app.

Investors have chased this growth. In June, crypto VC firm Paradigm led a $185 million investment into Kalshi at a $2 billion valuation. The financing brought its total funding to $415 million. Multicoin Capital, Sequoia Capital and Citadel Securities CEO Peng Zhao have previously invested.

Around the same time, Polymarket raised more than $200 million at a $1 billion valuation in an investment led by Founders Fund. Polymarket said in late August that Donald Trump, Jr. had joined its advisory board and that 1789 Capital, a VC firm at which Trump, Jr. is a partner, had made a strategic investment. Other investors include General Catalyst and Polychain Capital.

Eliezer Yudkowsky has spent the past 20 years warning A.I. insiders of danger. Now, he’s making his case to the public.

The first time I met Eliezer Yudkowsky, he said there was a 99.5 percent chance that A.I. was going to kill me.

I didn’t take it personally. Mr. Yudkowsky, 46, is the founder of the Machine Intelligence Research Institute, a Berkeley-based nonprofit that studies risks from advanced artificial intelligence.

For the last two decades, he has been Silicon Valley’s version of a doomsday preacher — telling anyone who will listen that building powerful A.I. systems is a terrible idea, one that will end in disaster.

That is also the message of Mr. Yudkowsky’s new book, “If Anyone Builds It, Everyone Dies.” The book, co-written with MIRI’s president, Nate Soares, is a distilled, mass-market version of the case they have been making to A.I. insiders for years.

Their goal is to stop the development of A.I. — and the stakes, they say, are existential.

“If any company or group, anywhere on the planet, builds an artificial superintelligence using anything remotely like current techniques, based on anything remotely like the present understanding of A.I., then everyone, everywhere on Earth, will die,” they write.

This kind of blunt doomsaying has gotten Mr. Yudkowsky dismissed by some as an extremist or a crank. But he is a central figure in modern A.I. history, and his influence on the industry is undeniable.

Google, too, owes some of its A.I. success to Mr. Yudkowsky. In 2010, he introduced the founders of DeepMind — a London start-up that was trying to build advanced A.I. systems — to Peter Thiel, the venture capitalist. Mr. Thiel became DeepMind’s first major investor, before Google acquired the company in 2014. Today, DeepMind’s co-founder Demis Hassabis oversees Google’s A.I. efforts.

In addition to his work on A.I. safety — a field he more or less invented — Mr. Yudkowsky is the intellectual force behind Rationalism, a loosely organized movement (or a religion, depending on whom you ask) that pursues self-improvement through rigorous reasoning. Today, Silicon Valley tech companies are full of young Rationalists, many of whom grew up reading Mr. Yudkowsky’s writing online.

I’m not a Rationalist, and my view of A.I. is considerably more moderate than Mr. Yudkowsky’s. (I don’t, for instance, think we should bomb data centers if rogue nations threaten to develop superhuman A.I. in violation of international agreements, a view he has espoused.) But in recent months, I’ve sat down with him several times to better understand his views.

At first, he resisted being profiled. (Ideas, not personalities, are what he thinks rational people should care about.) Eventually, he agreed, in part because he hopes that by sharing his fears about A.I., he might persuade others to join the cause of saving humanity.

“To have the world turn back from superintelligent A.I., and we get to not die in the immediate future,” he told me. “That’s all I presently want out of life.”

Mr. Yudkowsky grew up in an Orthodox Jewish family in Chicago. He dropped out of school after eighth grade because of chronic health issues, and never returned. Instead, he devoured science fiction books, taught himself computer science and started hanging out online with a group of far-out futurists known as the Extropians.

He was enchanted by the idea of the singularity — a hypothetical future point when A.I. would surpass human intelligence. And he wanted to build an artificial general intelligence, or A.G.I., an A.I. system capable of doing everything the human brain can.

“He seemed to think A.G.I. was coming soon,” said Ben Goertzel, an A.I. researcher who met Mr. Yudkowsky as a teenager. “He also seemed to think he was the only person on the planet smart enough to create A.G.I.”

He moved to the Bay Area in 2005 to pursue what he called “friendly A.I.” — A.I. that would be aligned with human values, and would care about human well-being.

But the more Mr. Yudkowsky learned, the more he came to believe that building friendly A.I. would be difficult, if not impossible.

One reason is what he calls “orthogonality” — the notion that intelligence and benevolence are separate traits, and that an A.I. system would not automatically get friendlier as it got smarter.

Another is what he calls “instrumental convergence” — the idea that a powerful, goal-directed A.I. system could adopt strategies that end up harming humans. (A well-known example is the “paper clip maximizer,” a thought experiment popularized by the philosopher Nick Bostrom, based on what Mr. Yudkowsky claims is a misunderstanding of an idea of his, in which an A.I. is told to maximize paper clip production and destroys humanity to gather more raw materials.)

He also worried about what he called an “intelligence explosion” — a sudden, drastic spike in A.I. capabilities that could lead to the rapid emergence of superintelligence.

At the time, these were abstract, theoretical arguments hashed out among internet futurists. Nothing remotely like today’s A.I. systems existed, and the idea of a rogue, superintelligent A.I. was too far-fetched for serious scientists to worry about.

But over time, as A.I. capabilities improved, Mr. Yudkowsky’s ideas found a wider audience.

In 2010, he started writing “Harry Potter and the Methods of Rationality,” a serialized work of Harry Potter fan fiction that he hoped would introduce more people to the core concepts of Rationalism. The book eventually sprawled to more than 600,000 words — longer than “War and Peace.” (Brevity is not Mr. Yudkowsky’s strong suit — another of his works, a B.D.S.M.-themed Dungeons & Dragons fan fiction that contains his views of decision theory, clocks in at 1.8 million words.)

Despite its length, “Harry Potter and the Methods of Rationality” was a cult hit, and introduced legions of young people to Mr. Yudkowsky’s worldview. Even today, I routinely meet employees of top A.I. companies who tell me, somewhat sheepishly, that reading the book inspired their career choice.

Some young Rationalists went to work for MIRI, Mr. Yudkowsky’s organization. Others fanned out across the tech industry, taking jobs at companies like OpenAI and Google. But nothing they did slowed the pace of A.I. progress, or allayed any of Mr. Yudkowsky’s fears about how powerful A.I. would turn out.

In 2022, Mr. Yudkowsky announced — in what some interpreted as an April Fools joke — that he and MIRI were pivoting to a new strategy he called “death with dignity.” Humanity was doomed to die, he said, and instead of continuing to fight a losing battle to align A.I. with human values, he was shifting his focus to helping people accept their fate.

“It’s obvious at this point that humanity isn’t going to solve the alignment problem, or even try very hard, or even go out with much of a fight,” he wrote.

These are, it should be said, extreme views even by the standards of A.I. pessimists. And during our most recent conversation, I raised some objections to Mr. Yudkowsky’s claims.

Haven’t researchers made strides in areas like mechanistic interpretability — the field that studies the inner workings of A.I. models — that may give us better ways of controlling powerful A.I. systems?

“The course of events over the last 25 years has not been such as to invalidate any of these underlying theories,” he replied. “Imagine going up to a physicist and saying, ‘Have any of the recent discoveries in physics changed your mind about rocks falling off cliffs?’”

What about the more immediate harms that A.I. poses — such as job loss, and people falling into delusional spirals while talking to chatbots? Shouldn’t he be at least as focused on those as on doomsday scenarios?

Mr. Yudkowsky acknowledged that some of these harms were real, but scoffed at the idea that he should focus more on them.

“It’s like saying that Leo Szilard, the person who first conceived of the nuclear chain reactions behind nuclear weapons, ought to have spent all of his time and energy worrying about the current harms of the Radium Girls,” a group of young women who developed radiation poisoning while painting watch dials in factories in the 1920s.

Are there any A.I. companies he’s rooting for? Any approaches less likely to lead us to doom?

“Among the crazed mad scientists driving headlong toward disaster, every last one of which should be shut down, OpenAI’s management is noticeably worse than the pack, and some of Anthropic’s employees are noticeably better than the pack,” he said. “None of this makes a difference, and all of them should be treated the same way by the law.”

And what about the good things that A.I. can do? Wouldn’t shutting down A.I. development also mean delaying cures for diseases, A.I. tutors for students and other benefits?

“We totally acknowledge the good effects,” he replied. “Yep, these things could be great tutors. Yep, these things sure could be useful in drug discovery. Is that worth exterminating all life on Earth? No.”

Does he worry about his followers committing acts of violence, going on hunger strikes or carrying out other extreme acts in order to stop A.I.?

“Humanity seems more likely to perish of doing too little here than too much,” he said. “Our society refusing to have a conversation about a threat to all life on Earth, because somebody else might possibly take similar words and mash them together and then say or do something stupid, would be a foolish way for humanity to go extinct.”

Even among his fans, Mr. Yudkowsky is a divisive figure. He can be arrogant and abrasive, and some of his followers wish he were a more polished spokesman. He has also had to adjust to writing for a mainstream audience, rather than for Rationalists who will wade through thousands of dense, jargon-packed pages.

“He wrote 300 percent of the book,” his co-author, Mr. Soares, quipped. “I wrote another negative 200 percent.”

He has adjusted his image in preparation for his book tour. He shaved his beard down from its former, rabbinical length and replaced his signature golden top hat with a muted newsboy cap. (The new hat, he said dryly, was “a result of observer feedback.”)

Several months before his new book’s release, a fan suggested he take one of his self-published works, an erotic fantasy novel, off Amazon to avoid alienating potential readers. (He did, but grumbled to me that “it’s not actually even all that horny, by my standards.”)

In 2023, he started dating Gretta Duleba, a relationship therapist, and moved to Washington State, far from the Bay Area tech bubble. To his friends, he seems happier now, and less inclined to throw in the towel on humanity’s existence.

Even the way he talks about doom has changed. He once confidently predicted, with mathematical precision, how long it would take for superhuman A.I. to be developed. But he now balks at those exercises.

“What is this obsession with timelines?” he asked. “People used to trade timelines the way that they traded astrological signs, and now they’re trading probabilities of everybody dying the way they used to trade timelines. If the probability is quite large and you don’t know when it’s going to happen, deal with it. Stop making up these numbers.”

I’m not persuaded by the more extreme parts of Mr. Yudkowsky’s arguments. I don’t think A.I. alignment is a lost cause, and I worry less about “Terminator”-style takeovers than about the more mundane ways A.I. could steer us toward disaster. My p(doom) — a rough, vibes-based measure of how probable I feel A.I. catastrophe is — hovers somewhere between 5 and 10 percent, making me a tame moderate by comparison.

I also think there is essentially no chance of stopping A.I. development through international treaties and nuclear-weapon-level controls on A.I. chips, as Mr. Yudkowsky and Mr. Soares argue is necessary. The Trump administration is set on accelerating A.I. progress, not slowing it down, and “doomer” has become a pejorative in Washington.

Even under a different White House administration, hundreds of millions of people would be using A.I. products like ChatGPT every day, with no clear signs of impending doom. And absent some obvious catastrophe, A.I.’s benefits would seem too obvious, and the risks too abstract, to hit the kill switch now.

But I also know that Mr. Yudkowsky and Mr. Soares have been thinking about A.I. risks far longer than most, and that there are still many reasons to worry about A.I. For starters, A.I. companies still don’t really understand how large language models work, or how to control their behavior.

Their brand of doomsaying isn’t popular these days. But in a world of mealy-mouthed pablum about “maximizing the benefits and minimizing the risks” of A.I., maybe they deserve some credit for putting their cards on the table.

“If we get an effective international treaty shutting A.I. down, and the book had something to do with it, I’ll call the book a success,” Mr. Yudkowsky told me. “Anything other than that is a sad little consolation prize on the way to death.”

The artificial intelligence giant is closing in on a deal with Microsoft regarding its future governance, but other questions stand over its huge costs.

OpenAI appears to be moving closer to resolving an existential crisis, reaching a tentative agreement to rework its partnership with Microsoft.

A finalized deal would help the artificial intelligence start-up in its quest to convert into a for-profit company, a crucial step to unlocking billions more in funding. But even that doesn’t lift all the questions that are hanging over the ChatGPT maker, as investors continue to show concerns about the A.I. boom.

What we know about the OpenAI-Microsoft talks so far:

• The Times reports that the two have agreed to rework a clause in their original agreement that would rescind Microsoft’s access to OpenAI’s most advanced technology once the start-up had achieved so-called artificial general intelligence. (Read: A.I. as capable as a human brain, though the definition is hazy.)
• The Financial Times reports that Microsoft is also set to take a roughly 30 percent stake in a reorganized OpenAI once it becomes a for-profit company, worth perhaps $170 billion. The nonprofit organization that currently controls OpenAI will get a stake worth at least $100 billion, or more than 20 percent of the for-profit entity.

The stakes are huge. Establishing a for-profit structure (via a public benefit corporation) would let OpenAI eventually go public and raise money from outside investors. Nearly $20 billion worth of funds it has raised during the past year is contingent on the conversion.

For Microsoft, maintaining its uniquely close ties to a leading A.I. developer — even if it’s reportedly deepening its relationship with rivals like Anthropic — will help it maintain a technological lead over competitors including Google. Shares in Microsoft are up 1.2 percent in premarket trading on Friday.

OpenAI still faces questions about its future. Company executives are worried about the possibility state regulators in California and Delaware could block the conversion. (Elon Musk is also fighting the move.)

Then there’s the sheer cost of researching and running A.I. Consider:

All that rests on the assumption that OpenAI’s finances will keep up with the rocket-like growth of ChatGPT use. But The Wall Street Journal points to recent research that calls into question whether customers will open their wallets as wide as expected.

With international expansion at a steady pace and a roster of popular artist IPs beyond Labubu, the Chinese blind box and plush toy maker is looking at $4.18 billion in 2025 revenue with ease.

LONDON — Global interest in Labubu is cooling, but that might be good news for its maker, the Hong Kong-listed Pop Mart.

According to Google Trends, searches in the fluffy figurine with a grin of spiky teeth peaked in the week of July 13 to 19, and has since seen a gradual decline. That said, the volume of searches for Labubu on Google at the end of August was still 12 times higher than a year prior.

Thanks to endless celebrity endorsements — most recently Naomi Osaka’s U.S. Open run, and Lady Gaga and Lisa’s custom Labubu by Brett Alan Nelson — and more than 2.8 million user-generated contents on TikTok, Labubu has dominated the conversation in 2025.

Some call it a perfect match with the world of luxury, while some dismiss it as a symbol of brain rot and equate collecting Labubu to borderline gambling addiction due to the main range being sold in the form of a blind box, meaning the customer receives a random product from a themed series. There is a small chance that one might get a highly regarded secret version.

Investors, in the meantime, are hailing Pop Mart, the company behind the global success of Labubu, as an economic slowdown-defying wonder.

The share price of Pop Mart more than tripled year-over-year to 320 Hong Kong dollars at the end of August.

Goldman Sachs last month lifted its target price to 350 Hong Kong dollars per share as Pop Mart’s revenue jumped 204.4 percent to $1.95 billion, and net profit surged 385.6 percent to $659 million in the first half of 2025. The shares have since slipped back to close at 276.80 on Friday. The latest “Pin for Love” Series featuring mini Labubus did not spark the same level of hype as anticipated.

The lion’s share of the growth came from the U.S. and Europe, up 1,142.3 percent and 729.2 percent, respectively, in the period.

Its Asia-Pacific business, which includes Singapore, Malaysia and Thailand, grew 257.8 percent year-over-year, while its core mainland China operation also logged a substantial 135.2 percent increase.

Pop Mart kick-started its global expansion following a 676 million listing in Hong Kong in 2020. As of June 30, Pop Mart operated 571 stores and 2,597 roboshops in 18 countries globally.

Retail expansion in these regions has been steady in 2025.

Pop Mart in early August unveiled its largest store yet, an expansive two-floor unit clocking in at 8,200 square feet that includes a café, in Bangkok’s Iconsiam luxury mall. The retailer also entered Germany in July by opening its first store in the Alexa shopping center in Berlin. Both openings made local news headlines as thousands queued for hours in a bid to get regional exclusives and rare Labubu drops.

The Chinese company also snapped up a prime location on London’s Oxford Street in July for $86 million. Previously occupied by Matalan, the mixed-use building with 19,000 square feet of retail space is believed to be the home of Pop Mart’s new U.K. flagship.

Online channels are flourishing as well. According to data from Charm.io, Pop Mart’s sales on TikTok Shop U.S. saw a 1,828 percent increase from June 2024 to June 2025. Revenue growth had notable surges from October 2024 onward, with the most significant jumps in April and May 2025. By April 2025, Pop Mart accounted for 0.55 percent of all TikTok Shop U.S. sales.

During an earnings call, Pop Mart founder Wang Ning said the company is on track to meet its targeted 2025 revenue goal of 20 billion yuan, or $2.78 billion, and that 30 billion yuan, or $4.18 billion, this year “should also be quite easy.”

He added that sales from North America and Asia-Pacific this year would together equal China sales in 2024, and Pop Mart will begin a phase of “relatively rapid store openings” in the near future, with 10 more U.S. shops expected to open by the end of this year. Pop Mart currently has around 40 locations in the U.S.

Wang also disclosed that the company is exploring expansion in emerging markets in the Middle East, Central Europe, and Central and South America.

Another key indicator of Pop Mart’s longevity is a growing roster of popular artist IPs beyond Labubu.

First designed in 2015 by artist Kasing Lung, Labubu is part of the Monsters series, which also includes characters Zimomo, Spooky, Tycoco and Pato. Lung signed a licensing agreement with Pop Mart in 2019.

In the first half of 2025, The Monsters range raked in 4.81 billion renminbi, or $673 million, up 668 percent year-over-year, representing 34.7 percent of total revenue.

The big jump can be largely attributed to the release of the vinyl plush “Big Into Energy” series, as well as the “Wacky Mart” blind box figure toys range.

In a sit-down interview with China’s state-owned CCTV, Wang said the company sells around 10 million Labubu per month, adding that “Our sewing machines are running hot, and we’ve been ramping up production every month.”

He also revealed that renowned film studios worldwide, including many Hollywood companies, have approached Pop Mart about coproducing Labubu films.

Fashion and luxury brands are eager to collaborate with Labubu, too. But industry sources claimed that Pop Mart is keen to protect the IP’s long-term value, and is only looking to work with top-level brands in the luxury space.

So far, the only brands that have collaborated with Labubu via Pop Mart include Vans for a big plush doll, Chinese emerging designer brand Pronounce for a runway capsule and three plush dolls, Sacai for a limited run of 14 dolls in collaboration with K-pop group Seventeen, sold at a Joopiter auction, and, most recently, Uniqlo for a range of T-shirts and sweatshirts featuring Lung’s artwork.

At Lung’s own capacity, the LVMH-owned Moynat on Tuesday unveiled a collaboration featuring his artworks depicting Labubu, Zimomo, and King Mon to appear on handbags and small leather goods, including two-dimensional bag charms.

Pop Mart, according to sources, is instead encouraging brands to work with other popular IPs such as Molly, Skullpanda, Crybabay, and Dimoo under its umbrella, a move aimed at ensuring balanced development across its portfolio for long-term gain.

While these IPs are nowhere near the global popularity Labubu is enjoying at the moment, they have each cultivated their very own identity and niche, dedicated followings, just as Labubu has done so with the affluent, Birkin-collecting Tai-tais across South East Asia before it evolved into a global phenomenon.

All of these IPs generated over 100 million renminbi, or $14 million, each in revenue in the first half of 2025 for Pop Mart.

Swiss jeweler and watchmaker Chopard, for instance, teamed up with Dimoo, a starry-eyed character imagined by Chinese artist Ayan Deng, to release two China-exclusive limited-edition figurines sporting the brand’s Ice Cube designs for Qixi, considered China’s Valentine’s Day.

The Richemont-owned Cartier, meanwhile, has worked with Crybaby’s creator, Thai artist Nisa Srikamdee, to launch an exclusive set of stickers for the messaging app Line to celebrate the opening of Cartier’s new flagship at Siam Paragon in Bangkok.

At the same time, Pop Mart is doubling down on new IP incubation.

Twinkle Twinkle, created by the Chinese artist Da Xin and signed by Pop Mart a year ago, is one of the retailer’s emerging IPs with the most rapid growth, the company said.

The company this year also signed on KeyA, a futuristic, mechanical, and cool girl IP by the 20-year-old Chinese artist Chen Yanran. The first rendition of KeyA was teased at the 2025 Pop Mart Pop Toy Show held in Beijing at the beginning of August.

Category expansion should also help Pop Mart diversify its revenue stream.

For the first time, plush products outperformed figure toys in the first half of 2025, achieving a revenue of 6.13 billion renminbi, or $823 million, representing a year-on-year increase of 1,276.2 percent and accounting for 44.2 percent of total revenue.

In June, it also expanded into jewelry with a stand-alone retail concept, Popop, which opened its first stores in Shanghai and Beijing in June.

Offering seen as signal of whether large private equity groups can begin exiting largest deals

Medline Industries, a medical supply group backed by Blackstone, Hellman & Friedman and Carlyle, could go public as soon as this year in a bellwether test of investor appetite for large companies owned by private capital.

The biggest privately owned US distributor of medical products, Medline confidentially refiled its draft prospectus for an initial public offering with the US Securities and Exchange Commission last week, putting it on pace to go public in the coming months, according to three people familiar with the matter. The listing could value the company at more than $50bn and raise between $4bn-$5bn.

Medline’s board is set to meet next week to make a final decision about whether to push forward with an IPO in the remaining months of 2025 or pursue the offering in early 2026, said the people. Blackstone, Hellman & Friedman and Carlyle did not immediately respond to requests for comment.

The listing will be closely watched by investors, as Medline delayed prior IPO plans earlier this year over concerns about tariff impacts. Many of Medline’s medical products are sourced or manufactured in China and other tariff-affected countries, including Vietnam, Japan and Mexico.

After years of postponements because of high interest rates and lower valuations, private capital groups have endured additional delay due to uncertainty surrounding the economic impact of President Donald Trump’s trade war. Medline was the largest among a handful of prominent companies that delayed their IPOs after his “liberation day”.

The offering will also be a closely followed deal for the $4tn private equity industry, which will signal whether large private equity groups can begin exiting their largest takeover deals through IPOs. The trio of private equity investors took over Medline, a family-run, Chicago-based company, for $34bn in 2021 in one of the largest leveraged buyouts of all time.

Medline has benefited from a jump in its profits since the takeover, according to Fitch, which upgraded its credit rating in January. The group’s private equity owners have been waiting to realise billions of dollars in paper gains, as a multiyear downturn in new listing activity is beginning to reverse.

Medline’s plans emerged at the end of a week in which seven large-cap companies, including buy now, pay later firm Klarna, the Winklevoss twins’ crypto exchange Gemini and Blackstone-backed data centre engineer Legence went public in the US.

There were no guarantees that Medline would decide to go public this year, the people stressed, adding that the IPO could lapse into next year if there is further market uncertainty. The group’s owners, which also include sovereign wealth funds in Singapore and Abu Dhabi and its founding family, have not set firm goals on the size or valuation of the offering, meaning figures could change, the people added.

Medline first confidentially filed IPO plans at the end of last year, to list in 2025. Medline’s updated S1 filing included figures from the second quarter in which it generated $935mn in adjusted earnings, two people said. The business was on track to generate roughly $3.8bn in earnings across 2025, they added. 

GeneDx CEO Katherine Stueland explains how her company screens babies’ DNA for mutations that cause thousands of treatable disorders.

KJ Muldoon wasn’t supposed to make it to his first birthday. He was born in August 2024 with carbamoyl phosphate synthetase 1 deficiency, a genetic disorder that afflicts about 1 in 1.3 million newborns. Patients with the disorder don’t produce an enzyme needed to remove ammonia—a byproduct of protein metabolism—from their blood. The usual result: neurological damage and death.

The standard treatment is a liver transplant, but this entails a high risk of complications in infants. Thanks to advances in gene sequencing and editing, however, doctors at the Children’s Hospital of Philadelphia and University of Pennsylvania’s medical school were able to custom-design and manufacture a substitute gene that was delivered to KJ’s liver via a lipid nanoparticle—a microscopic fat-based shipping container. Physicians announced in May that the treatment was successful, and KJ is thriving.

The episode highlights how rapid genetic sequencing and cutting-edge therapies could enable tens of thousands of babies born every year with rare and debilitating diseases to grow up normally. “The earlier we can diagnose, the earlier we give these families a chance for them to live a better life,” Katherine Stueland, CEO of GeneDx, says in an interview at the Journal’s offices.

GeneDx was founded in 2000 by two former National Institutes of Health scientists, Sherri Bale and John Compton. Many academic labs and genetic startups founded around that time sought to decode genes behind more common diseases. But rare diseases were neglected because there wasn’t believed to be much commercial upside in diagnosing them.

Sequencing a genome two decades ago cost more than a Beverly Hills mansion, and how much demand could there be to diagnose diseases that by definition are rare? Turns out, a lot. As sequencing costs have plunged, GeneDx has the potential to revolutionize the field of rare diseases—which, collectively, are more common than most people realize.

The Food and Drug Administration estimates that some 30 million Americans suffer from rare diseases. Many are never diagnosed, and GeneDx aims to change that. It sequences DNA for mutations that are responsible for thousands of disorders.

A single deletion or duplication in the genetic code can result in misshapen or missing proteins that are vital to a variety of physiological processes. A case in point: Duchenne muscular dystrophy can result from a mutation or deletion in any of the 79 protein-encoding genes that together produce the dystrophin protein, which stabilizes muscles.

Standard newborn screenings can identify a few dozen of the most common congenital disorders, such as cystic fibrosis and sickle cell disease. But “about 95% of the conditions that we’re diagnosing would be completely missed based on newborn screening standards today,” Ms. Stueland says.

GeneDx can diagnose more than 4,800 disorders caused by genetic mutations that either are inherited from parents or arise spontaneously. The company has sequenced the DNA of more than 850,000 individuals, which has enabled it to discover more links between gene variants and conditions. It now uses AI to connect the dots between diseases and genes.

“We’re making new gene disease discoveries now on a weekly basis,” Ms. Stueland, 49, says. “So we thought there were 7,000 rare diseases, but it turns out there are 10,000.” Most are caused by a single gene mutation, which means they could potentially be treated with gene-editing technologies, as KJ Muldoon was. Hundreds of gene therapies for rare disorders are undergoing clinical trials. Several await FDA approval, including treatments for Hunter and Wiskott-Aldrich syndromes. They cause enlarged organs and immune dysfunction, among other debilitating symptoms.

While not a medical doctor or geneticist, Ms. Stueland developed an interest in rare diseases when her cousins were diagnosed with cystic fibrosis. She recalls organizing a neighborhood raffle to raise money for the Cystic Fibrosis Foundation. The grand prize? Free baby-sitting. She graduated with a bachelor’s degree in English literature from Miami University where she performed in dance theater and later worked at small genetic diagnostic and biotech companies before being named CEO of GeneDx in 2021. This background enables her to blend the human element with the science.

While scientists know that some gene mutations cause certain disorders, there are also “variants of unknown significance,” Ms. Stueland says. That means scientists don’t know whether they are connected to a condition. But the more genomes GeneDx sequences, the more it can “upgrade or downgrade” a variant’s significance.

There are also gene variants that heighten the risk for disorders but not to 100%. “There are 800 genes associated with autism,” Ms. Stueland says. It isn’t well-understood how autism is affected by each of these genes, but sequencing more individual genomes can yield clues. Gene therapies might even be able to help some people with autism.

“I just heard from a family who had a young man in his early 20s that suspected that he had some form of autism his entire life,” Ms. Stueland says. “He actually had been adopted. And because we were able to diagnose him in his early 20s, there’s actually a company based in Chicago that has a gene therapy that is in clinical development that he would be eligible for.”

The company is Jaguar Gene Therapy, which is conducting a trial for autistic patients with a mutation or deletion in the SHANK3 gene. By enabling the rapid diagnoses of more rare diseases, Ms. Stueland hopes to help more patients access treatments and clinical trials.

About one-third of babies in neonatal intensive-care units have a genetic disorder, but very few of them receive a genetic test. Many of them have mutations that could be identified by GeneDx’s screening and benefit from prompt treatment. Its genome-sequencing is being used in a large-scale study backed by the New York State Health Department, in which every baby born in the New York Presbyterian hospital system is eligible to participate. Babies are screened for more than 450 “clinically actionable conditions”—those that could benefit from immediate interventions. That could mean a ketogenic diet, a beta blocker or another FDA-approved therapy, among other things.

Ms. Stueland’s hope is that as drug discovery progresses, more conditions will become “clinically actionable.” But she says GeneDx doesn’t want to give parents information about potential health risks that can’t be remedied: “No parent needs to hear that their happy healthy baby may have prostate cancer in the future.”

Parents who participate in the New York study are contacted by doctors with the results, usually within three to six weeks. The study has found a 3.7% positive rate in the first 4,000 babies. That means more than 3 in 100 have a genetic disorder that could be amenable to treatment: “We’re diagnosing disease at the earliest moment—sometimes before symptoms are fully manifesting.”

Consider rare inherited heart arrhythmias, which are responsible for nearly 20% of sudden infant death syndrome cases. Arrhythmias can be managed with beta blockers and heart monitoring. About a quarter of childhood seizures are caused by faulty genes. Identifying the responsible gene variant can help doctors determine an appropriate treatment.

Lili Hasse’s infant daughter, Margot, was diagnosed as part of the New York study with CDKL5 deficiency disorder, which causes seizures and delayed neuro-development. Ms. Hasse says Margot’s rapid diagnosis enabled her to schedule an appointment immediately with a pediatric neurologist who specializes in the condition and to connect with a support group on Facebook.

She considers herself lucky. Other parents in the group have run a yearslong medical gauntlet before their children were ultimately diagnosed. Thanks to medications, supplements, a ketogenic diet and physical therapy, “Margot has near complete seizure control,” Ms. Hasse says. Gene therapies in the pharmaceutical pipeline may also someday help Margot.

Parents of children with rare diseases are often extremely entrepreneurial and engaged, Ms. Stueland says. Some are “searching for an investigational compound that might help their child, or if not for their child, somebody else’s child in the future.”

The journal Nature this summer profiled parents of children with rare diseases who launched therapeutic startups. Some had business backgrounds, but few had any formal scientific training. Rare-disease startups like Ultragenyx and Alexion sponsor a Rare Bootcamp for parents seeking to develop rare-disease drugs for their children. Expanded genetic sequencing could enable such startups to recruit more patients for clinical trials.

Ms. Stueland says that a “chicken and egg” problem has long hindered development of treatments for rare diseases: “You need to have a diagnosis in order to motivate somebody to go find a treatment.” She hopes more states will expand access to genetic sequencing, as Florida has recently done.

Gov. Ron DeSantis in June signed the Sunshine Genetics Act, which passed the Legislature unanimously. It will allow parents to have their baby’s genetic code screened at no cost. Adam Anderson, the legislation’s sponsor, had a son who died at 4 from Tay-Sachs disease, a rare inherited syndrome that results in the death of nerve cells in the brain and spinal cord.

Ms. Stueland argues that expanding access to newborn gene screening could save the health system money by reducing the amount that is currently spent on other tests, scans and specialists to diagnose a disorder. Treating patients with rare diseases early could also forestall disease progression and prevent expensive hospital visits.

How much does a test cost? She says GeneDx gets about $3,700 for each test and its gross margins are around 80%, suggesting it costs the company around $750 to run the test. Its overall margins, however, are small—about $7 million on $302 million in revenue last year—because it pumps a lot of money into improving its technology and scale.

She’s confident the cost of testing will continue to decline while benefits from new discoveries grow. That has been the history. The first human genome took 13 years to sequence, wasn’t complete until 2003, and cost about $3 billion. “We can now interpret a whole genome in 48 hours,” she says. She credits the progress in part to a 2013 Supreme Court decision, Association for Molecular Pathology v. Myriad Genetics, which held that natural fragments of DNA couldn’t be patented. “That opened up the entire industry.”

The most prominent gene-sequencing company was 23andMe, which pitched DNA tests directly to consumers. Using only a small amount of saliva, the company claimed it could tell you about your ancestry and health proclivities, such as natural wake-up time or whether you’re more likely to excel at endurance sports or activities that require short bursts of power.

After a meteoric rise, the company flamed out and filed for bankruptcy in January. What are the lessons of its failure? “Where they really provided a change for the better was getting people curious about their genes,” Ms. Stueland says. But after shelling out hundreds of dollars for the kits, most of 23andMe’s 15 million customers weren’t willing to pay more for additional health insights of marginal value. The business model failed, she thinks, because it didn’t give healthy adults a way to use their genetic information proactively.

Ms. Stueland says she ultimately hopes to furnish information from GeneDx’s newborn screenings to the patients when they come of age, empowering them to make decisions: “Ultimately, your DNA is your DNA, and so if you want that information, we want to be able to provide it to you.”

But the central focus is helping children like KJ Muldoon and their parents. “Its examples like that,” she says, that “should give us all hope as a society, as an industry, as a country.”

Money, talent and entrepreneurial spirit have turned DeepSeek’s hometown into a global AI hub

HANGZHOU, China—China is racing to develop world-leading artificial-intelligence technology. This city is paving the way.

More than two decades ago, Jack Ma launched Alibaba from a small apartment here, kick-starting his hometown’s transformation from scenic city to tech powerhouse.

These days, Hangzhou is an AI hub at the center of China’s global tech ambitions. Its breakthrough moment came earlier this year, when local company DeepSeek shocked the world with an AI model that rivaled American programs, at a much lower cost.

“All of Hangzhou went into a frenzy,” said Zhao Ji, a 41-year-old entrepreneur who is building an AI startup from his apartment in the city.

DeepSeek’s shot to fame was no fluke. The capital of the eastern Zhejiang province and a prosperous city of roughly 13 million people, Hangzhou has spent decades cultivating entrepreneurship. Supportive government policies, research universities, major tech companies such as Alibaba and NetEase and relatively low living costs compared with Beijing and Shanghai have made Hangzhou a mecca for tech talent.

Zhao moved to Hangzhou from Beijing in 2018 and was a marketing director for Alibaba for about three years. He is one of many Alibaba alumni now looking to make their mark in the AI industry.

On a recent Friday, Zhao and two of his employees huddled around a table at his two-story apartment in a gated community near Alibaba’s headquarters. Zhao’s 8-year-old son’s toys were scattered on the floor. A signed poster of Buzz Aldrin on the moon and a framed Life magazine cover of Abraham Lincoln on Broadway adorned the walls. A biography of Walmart founder Sam Walton sat on the table.

They are building an AI agent to help businesses generate video and audio content, with the ability to clone users’ voice and likeness. Zhao launched the startup, AlphaFin, last year.

Last October, Zhao was invited to speak about AI at an expo hosted by the Hangzhou government. After his presentation, local officials approached him to add him on WeChat, China’s do-everything app, and invited him to their districts. He didn’t yet have a working product.

“They really value tech and innovation,” Zhao said.

Known for its picturesque West Lake, a source of inspiration for poets and painters throughout Chinese history, Hangzhou has long been a center for trade, thriving far from China’s political centers. Family workshops in Zhejiang province were among the first to jump into the private sector when the Chinese economy opened up in the early 1980s. Then Alibaba helped lend Hangzhou its tech-forward sheen.

Drawing inspiration from Silicon Valley, the Hangzhou government built what it calls an innovation corridor in the city’s west, near Alibaba’s campus, where startups make AI models, robotics, brain-computer interfaces and AI-assisted software. Hangzhou is known for its “Six Little Dragons,” an informal cohort of six leading AI-related tech companies in the city that includes DeepSeek.

Hangzhou earlier this year said it approved two innovation-focused funds totaling more than $28 billion. Zhejiang ranks as the top province for science and technology investment: more than $12 billion in 2024.

Alibaba is also a major funder for highflying AI startups, in addition to its own AI business, which includes chips and models. The company has said it plans to invest more than $52 billion in AI and cloud infrastructure over the next three years.

Husband and wife Wei Dabao and Wu Xiaobao have had a front-row seat to the rapid development in Hangzhou. They own a bookstore near Alibaba’s campus where tech workers browse translated books and meet for reading events. When the shop opened five years ago, it was surrounded by construction sites. Now, the area is full of shiny buildings.

Wei said Hangzhou is a “gathering place” for tech talent, thanks to Alibaba. Most people in town have either worked at Alibaba or have friends who are current or former employees.

“Everyone is connected, more or less,” he said.

The northern suburb of Liangzhu has become famous in China’s AI community for events that bring tech founders and aspirants together. Many residents call themselves “villagers.” AI entrepreneurs present ideas and mingle in a shaded backyard at a monthly gathering called “Demo Day.” Independent programmers work on projects side by side at a local coffee shop during weekly meetups known as “Crazy Thursday.”

Then there is Hangzhou’s premier research university, Zhejiang University, which counts many entrepreneurs among its alumni, including DeepSeek founder Liang Wenfeng. In the early 2000s, Zhejiang started aspiring to become a kind of Stanford University for China.

“The joke was ‘How to build a Silicon Valley? Build a world-class university and wait 30 to 50 years,’” said Duncan Clark, author of a book about Alibaba and a former visiting scholar at a Stanford entrepreneurship program. Zhejiang University, he said, “did it faster.”

In 2017, the school, the provincial government and Alibaba together set up an AI lab. The same year, Zhu Qiuguo, an engineering professor at Zhejiang, founded Deep Robotics, one of Hangzhou’s “Six Little Dragons.”

Hangzhou is also home to Westlake University, a new research institution supported by both private and public funding that aims to rival the California Institute of Technology. It has attracted many high-profile professors who have returned to China from American universities, as the U.S. cuts academic funding and heightens scrutiny of Chinese scientists. AI researcher Guo-Jun Qi and data scientist She Yiyuan are among Westlake’s newest hires from the U.S.

China’s research universities still have some catching up to do when it comes to invention. Nine of the top 10 institutions for international patent applications from academia are American, while the highest-ranked Chinese university is 29th, according to 2021-2024 data from the German Economic Institute.

Industry players and analysts also caution that China’s attempts to prop up specific industries have had mixed success because of potential resource misallocation or excessive political attention or restrictions. The scars of Beijing’s crackdown starting in 2020 on China’s tech companies are still fresh.

“The state has a deadening hand that just as often crushes entrepreneurship as it does to promote it,” said Dan Wang, a research fellow at Stanford’s Hoover History Lab.

For now, though, there are plenty of people eager to ride the AI wave.

Besides Hangzhou, Beijing and Shanghai, Guangdong province’s Shenzhen—home to tech giants such as WeChat parent Tencent and Apple rival Huawei—is another AI hub.

Liu Pei heard about Hangzhou’s AI scene and moved to the city from Beijing in May to found an AI media startup called White Whale Lab.

“When you come to Hangzhou, the people you meet make you feel really happy,” she said. “They’re really eager to seize the opportunities in AI.”

After sales fell and dealers revolted, the company reverses course and revives the popular model

Jeep is reintroducing its Cherokee after discontinuing it in 2023, a move that led to a significant sales decline.

The return of the Cherokee, now as a hybrid, is central to Jeep’s strategy to regain market share and boost sales.

Jeep aims to offer a competitively priced hybrid SUV, filling a gap between its Compass and Grand Cherokee models.

On a weeknight in mid-August, thousands of Jeep fans flocked to a waterfront park in Brooklyn, N.Y., where an ambitious turnaround campaign was underway. Rapper LL Cool J performed, and Jeep, facing an unsettling sales slump, unveiled the return of its Cherokee sport-utility vehicle that it branded “America’s Original Influencer.”

The Cherokee is credited with inventing the modern SUV and accounted for one out of every six Jeeps sold in America just before the pandemic, the sort of precious branding power companies dream of.

But in 2023, in what would be a major strategic mistake, Jeep stopped making the Cherokee as part of cost-cutting measures to support the automaker’s transition to making more electric vehicles. Sales fell and dealers revolted.

Now Jeep and its parent company Stellantis STLA -2.61%decrease; red down pointing triangle are trying to mount a comeback.

The “biggest piece of the puzzle” to fixing Jeep’s woes is simple: “You start with the Cherokee,” said Bob Broderdorf, a company lifer who was put in charge of Jeep earlier this year, in an interview.

Automakers are in the midst of big pivots. After they pushed to electrify their product portfolios for several years, goaded in part by stringent emissions regulations, consumer interest in EVs dramatically fell off. Now, recent policy changes have made American classics like the Cherokee more attractive for car companies to sell.

Midsize SUVs like the Cherokee are the most popular kind of vehicle in the U.S. Not having one factored heavily in the 36% slide in Jeep sales in 2024 from before the pandemic, according to data from industry-research firm Motor Intelligence.

Jeep stopped marketing the Cherokee model in the U.S. once before, in 2002, and replaced it with the Liberty, but brought it back in 2012 as Jeep’s popularity soared and it expanded globally.

Things changed after Stellantis was created in early 2021 through the merger of Fiat Chrysler Automobiles and France’s PSA Group. Faced with stringent tailpipe emissions standards enacted around the world, Stellantis’s then-chief executive Carlos Tavares began cutting costs to prepare the company to make more electric vehicles.

Stellantis kept sticker prices high during the pandemic amid a new-car supply crunch, even after competitors started to dial back prices. It also launched a new branding campaign for Jeep’s luxury models. Both moves backfired.

Then in early 2023, Stellantis canned the Cherokee, and this time, it didn’t have a replacement.

Nixing the Cherokee wreaked havoc and annoyed U.S. dealers. Between 2020 and 2024, Stellantis’s U.S. market share dropped from 12.5% to below 8% after the Cherokee was discontinued.

The dealer outrage bubbled over in a letter that a national group fired off to Tavares in 2024. They argued that higher prices relative to competitors had seriously eroded the company’s market share.

“Reintroduce the Jeep Cherokee ASAP,” the May 2024 letter said.

By the end of last year, Stellantis realized it needed to change course, and parted ways with Tavares, the CEO who orchestrated the merger. And earlier this year the carmaker assembled a new leadership team, installing Broderdorf as the leader of the Jeep brand. Broderdorf had spent the last two decades leading some of Stellantis’s most important North American brands like Dodge and Ram.

For Stellantis, the Jeep turnaround plan is vital, and the Cherokee is a big piece of it. It also plans new luxury Jeep models, and a second EV model later this year.

Since the start of 2025, the company’s share price is down 30%, and it reported losses of more than $2.5 billion in the first half of the year. The North American region—where along with Jeep its main seller is Ram trucks—has historically contributed a majority of profits. More than 80% of its total retail shipments in the U.S. last year were pickups and larger SUVs, Jeep’s bread and butter, according to a company filing.

“I have a great deal of respect for how important this brand is for the company, and the heritage of it and what it means for so many people,” Broderdorf said.

Broderdorf’s approach to right the ship amounts to getting back to basics for a brand with a devoted fan base: Jeep drivers are known to wave to one another on the road and leave rubber ducks on door handles of other Jeeps. Jeep knows it needs a midsize SUV in its product lineup to boost sales and draw in customers, he said.

So far, Jeep has seen a glimmer of hope: In the second quarter of 2025, Jeep notched a slight 1% sales increase compared with a year earlier.

The new Cherokee, when it goes on sale late this year, will only be offered with a hybrid gas-electric motor.

Dealers think the Cherokee relaunch should help. The loss of a midsize SUV like the Cherokee threw a wrench in long-established patterns of car buying, they say: customers who start out with a more affordable, entry-level model will, over time, move into a larger vehicle of the same brand as life changes occur, such as marriage or the birth of a child.

Without the Cherokee, Jeep customers didn’t have a step between the entry-level Compass SUV and the larger Grand Cherokee, which now starts around $37,000 and was much higher during the pandemic.

“It was a big jump to go from Compass to Grand Cherokee,” says Steven Wolf, a Houston dealer that sells Stellantis brands, including Jeep. “We didn’t have something in between, so we lost a lot of business.”

Jeep’s decision to make the coming Cherokee a hybrid-only should be a selling point for customers, Wolf says. The vehicle is expected to carry an estimated 500 miles of range on a single tank of gas.

Jeep says the new Cherokee will start at $35,000.

Joseph Yoon, an analyst for car-shopping website Edmunds, says the Cherokee’s larger size and hybrid powertrain should make it a competitive model for midsize SUV shoppers looking for something bigger than Jeep’s Compass but who don’t need a third row like its Grand Cherokee can offer.

“Right now, people want car-based SUVs that can pretend to be rugged, not rugged actual SUVs,” he says.

Cheaper sticker prices could help. Jeep brought pricing down across most of the brand’s portfolio over the past year. The move has been welcomed by dealers, who complained vehemently throughout 2024 about sticker prices they said made Jeep uncompetitive. Now, its cheapest model, the Compass SUV, starts under $27,000.

Heading into the ever-shifting regulatory environment, Broderdorf believes Jeep should have an edge in part because of Stellantis’s newer vehicle platforms, essentially the common structural foundations on which different types of vehicles can be built. The company’s platforms were designed to support cars powered by gasoline-only, hybrid models, or all-electric vehicles.

“For us, what’s super important is that we remain as flexible as humanly possible,” Broderdorf says. “Those are the manufacturers that I think will win.”

Sweating in elite new supersize saunas—with anywhere from one to 90 other people—produces a unique sort of connection.

Ambassador’s meeting with chancellor Rachel Reeves came days before AstraZeneca and Merck paused projects in Britain

The US ambassador to the UK has urged Rachel Reeves to offer a better deal on drug pricing to global pharmaceutical groups, piling pressure on the government in the same week that Merck scrapped a £1bn London development and AstraZeneca paused a Cambridge project.

At a private dinner at the ambassador’s residence in London on Sunday, Warren Stephens raised with the chancellor the US government’s concerns about the amount the UK pays for drugs as well as the value for money formula used for drug approvals, said multiple people briefed on the meeting.

One of the people said Stephens sought to convey to Reeves how strongly the US government feels about the need for the UK to improve the environment for pharmaceutical companies. Another described the discussion about drug pricing and approvals as “frosty”.

News of the US intervention comes as AstraZeneca said it was pausing a £200mn investment that would have expanded its research and development operations in Cambridge.

The Anglo-Swedish drugmaker’s announcement came two days after US pharma group Merck — known as MSD in Europe — scrapped a £1bn research centre in London, accusing the UK of not being internationally competitive.

Pharmaceutical companies have been pushing ministers to pay more for drugs while complaining that a clawback tax on their UK sales unexpectedly soared this year.

The groups had previously warned that investment in UK research and development was at risk because of what they see as a refusal to pay enough for innovative drugs.

President Donald Trump has railed against European countries that do not pay as much for drugs as the US, accusing them of being “foreign freeloaders”. In May, he described how a friend was able to buy an obesity medication in London for far less than he could in the US.

He has given companies a deadline of the end of September to come up with ways to lower US drug prices and threatened to impose a “most favoured nation” policy that would peg US prices to those in other developed countries.

A senior British official said this week’s meeting was “not just about the UK, but about the rest of the world generally”, adding that the Trump administration believed “everyone else piggybacks on US pharma and their R&D and then keeps prices low so that US pharma doesn’t get a fair return”.

Trump is now “live” to the issue and “wants compensation”, the person added.

A person close to Reeves said Stephens acknowledged wider challenges for the pharmaceutical industry and showed understanding about the tough fiscal environment in the UK.

The US ambassador did not respond to a request for comment.

A spokesperson for the UK Treasury said: “This government is open to working collaboratively with the pharmaceutical industry.”

They added that the government had “put forward a generous and unprecedented offer as a part of the VPAG [voluntary scheme for branded medicines pricing, access and growth] review, which is worth approximately £1bn over three years”, and cut “clinical trial approval times to under 150 days, cutting the delays that deter investors”.

The meeting came ahead of Trump’s three-day state visit to the UK next week, where the two countries are planning to announce agreements ranging from nuclear reactors to artificial intelligence data centres and whisky.

In the trade deal with the US, the UK promised to improve the “overall environment” for pharmaceutical companies, but the specifics are still to be thrashed out.

As the sector does battle with the UK government over drug prices, AstraZeneca on Friday became the latest drugmaker to say its plans to invest in Britain were in jeopardy.

“We constantly reassess the investment needs of our company,” the Anglo-Swedish drugmaker said as it paused expansion of its Cambridge R&D operations.

The move by AstraZeneca, the largest company in the UK by market capitalisation, comes after MSD earlier this week said it was scrapping a £1bn research centre in King’s Cross. Eli Lilly has also paused an investment in a laboratory site.

Earlier this year, AstraZeneca ditched plans to expand a vaccine factory in Speke near Liverpool after the UK reduced the subsidy offer. The plans to expand the Cambridge site were announced by the previous Conservative government as part of the same investment package as the vaccine facility.

Last month health secretary Wes Streeting walked away from the negotiating table with pharmaceutical companies over drug pricing. But health department officials are seeking to reopen talks over drug pricing and market access after MSD’s move to scrap its London research centre.