TheChek News Flow

Europe’s Mistral benefits from search for artificial intelligence alternatives

Start-up has embarked on expansion of its own infrastructure and secured significant contracts

Europe’s best-known artificial intelligence start-up Mistral AI has secured new contracts worth hundreds of millions of dollars, driving an upturn in its business that could help fuel a potential $1bn fundraising this year.

Paris-based Mistral faces intense competition from US and Chinese competitors but is starting to benefit from a European push to establish regional champions.

According to people familiar with its finances, its revenues have increased several times over since its last funding round a year ago and are on track to surpass $100mn a year for the first time, if it maintains sales momentum.

A relatively small number of large customers are driving much of that growth. Mistral had closed or was near to sealing a handful of commercial contracts, each worth at least $100mn over three to five years, the people said.

Corporate, public-sector and defence customers outside the US are increasingly looking for alternatives to US tech companies since Donald Trump returned to the White House.

“There’s a lot of European companies that want to reduce their dependency on US providers . . . There’s a growing demand for more strategic autonomy,” Arthur Mensch, Mistral’s chief executive, said last month.

This has spurred Mistral, which was valued at nearly €6bn in its last funding round a year ago, to embark on an ambitious programme to expand its own AI infrastructure, starting with a big data centre outside Paris, as well as collaborations with Abu Dhabi-based tech and investment groups G42 and MGX.

To fund that effort, the company — which has already raised more than $1bn since it was founded two years ago — was considering raising up to $1bn more, according to people familiar with the matter. It had begun to sound out potential investors, though a more formal fundraising process was not likely to begin until later this year, the people added.

Mistral declined to comment on its financial performance or funding plans.

The Nvidia-backed company, which was co-founded by three former Meta and Google DeepMind researchers, is far behind US rivals such as OpenAI and Anthropic in both fundraising and commercialisation.

Mistral’s “open” AI models, which customers can examine and customise for their own applications, also face competition from China’s DeepSeek and Meta’s Llama.

But tension between the Trump administration and Europe — as well as a desire by countries around the world to own and operate their own AI infrastructure — could be a boon for Mistral, which generated only tens of millions of dollars in revenue last year, according to people close to the company.

The latest contracts are modelled on Mistral’s €100mn deal with French shipping and logistics group CMA CGM. When that deal was announced in April, CMA CGM boss Rodolphe Saadé said the companies would work together on “bespoke” AI systems.

Mistral, which has about 250 employees, has significantly expanded its commercial team in recent months. It has adopted a sales model similar to that of US data analytics provider Palantir, employing a team of “solutions architects” who work like consultants with each customer on how best to deploy AI in their business.

That can mean a longer sales process than a typical enterprise software contract but with a bigger potential prize.

BNP Paribas, AXA, Stellantis and Veolia are among Mistral’s current customers. It also has a partnership with European defence tech start-up Helsing.

“Sovereignty is not our core business, and we’re a global company,” Mensch said in May. “But the last 100 days have tripled our business, in particular in Europe and outside of the US.”

Biotech behind Eli Lilly obesity pill aims for new standard of care

Chugai Pharmaceutical looks to combine orforglipron with its drug to tackle muscle wasting

The Japanese biotech behind Eli Lilly’s new blockbuster obesity pill believes it could be used with an anti-muscle wasting drug it has developed to create a new standard of patient care for weight loss.

Hitoshi Iikura, a senior executive at Chugai Pharmaceutical, told the Financial Times that the business had “high expectations” that its drug to treat spinal muscular atrophy could be combined with orforglipron, the once-daily weight-loss pill it has licensed to Lilly, or other similar treatments. Muscle loss is a common side effect of weight-loss drugs.

Chugai, which is majority owned by Roche, originally developed GYM329 to treat the degenerative muscle-wasting disease SMA. Studies suggest loss of “lean mass”, which includes muscle, could account for 25 to 39 per cent of the weight lost by people taking modern obesity drugs. Maintaining muscle is important for overall health, especially as people age.

Chugai’s success with new drugs and particularly with orforglipron has pushed its share price to all-time highs this year, making it the Tokyo stock exchange’s 11th largest company by market capitalisation.

Its stock had a big boost in April when orforglipron passed a significant milestone. Late-stage trial results showed it cut weight and blood sugar while being as safe as injectable treatments including Novo Nordisk’s Wegovy and Ozempic and Lilly’s Zepbound and Mounjaro. It is expected to be on the market from next year.

Iikura said that the company did not set out to treat weight loss. Its focus on obesity was possible, he added, because its scientists chose research targets that offered opportunities to develop drugs that could have wider applications.

In the case of GYM329, “it extended to treating obesity, which is a broader application than we had initially expected”, he said. “We now have high expectations for that. At first, we weren’t expecting a big market for obesity drugs . . . We didn’t anticipate it but it’s ballooning.”

Chugai is unusual in the pharmaceutical industry for the high degree of autonomy it has maintained under Roche, which owns just below 60 per cent of the company.

“Its relationship with Roche is a superpower,” said Stephen Barker, an analyst at investment bank Jefferies, which estimates orforglipron could be worth $40bn of sales a year at its peak.

Chugai took a “monozukuri” approach of precision and perfectionism to researching molecules and antibodies, Iikura said, while it does not have to worry about the costs of late-stage clinical trials thanks to the involvement of Roche.

Its operating profit margins, which were 48 per cent last year, are among the highest in the industry.

Since the 2002 deal with Roche, the Swiss group has the right of first refusal on all Chugai drugs. Hemlibra, a haemophilia treatment developed by Chugai, has become Roche’s second-biggest seller and dispelled any investor concerns about the partnership.

But Roche turned down orforglipron after difficult previous experience with a weight-loss treatment. Apart from Novo Nordisk and Lilly, which had long experience treating the related metabolic disease diabetes, the rest of the industry also showed little enthusiasm until recently.

“Since 2018, the market and scientific advances have evolved significantly and a new generation of [drugs] have established themselves as effective options to treat obesity,” Roche said in a statement. In 2023, it decided to go back into weight-loss drug development.

Chugai started research on orforglipron around 2003, according to Hiroshi Kawabe, who led the R&D team. The aim was to improve diabetes drugs that were in trials at the time to make them longer-acting, more stable and deliverable as tablets.

Analysts note that the idea of combining orforglipron and GYM329, which is licensed to Roche, is still at an early stage, as some researchers are sceptical about the risk of muscle loss from anti-obesity drugs.

But Miki Sogi of Bernstein said that “if this combination becomes a standard of care for obese patients, then that would be huge”.

Westinghouse targets $75bn US nuclear expansion after Trump order

Plans for 10 new reactors come as president tries to encourage an American atomic energy renaissance

Westinghouse is in talks with US officials and industry partners about deploying 10 large nuclear reactors to meet the goals of President Donald Trump’s executive orders, which aim to unleash an American atomic energy renaissance.

The Pennsylvania-based nuclear developer is one of a handful of western companies capable of designing and building large reactors, which typically have capacity to generate about 1,000MW of electricity, enough to power more than 500,000 homes.

The orders, which were published on May 23, set targets for quadrupling nuclear energy capacity in the US by 2050, starting work on 10 large reactors by 2030 and accelerating regulatory approvals.

They have sparked a rush among developers and utilities to fast-track plans, as they seek to tap into billions of dollars of federal incentives expected to be provided by the administration.

The orders have also propelled nuclear energy stocks to record highs this month in anticipation of a US construction boom.  

Dan Sumner, Westinghouse interim chief executive, said the company was “uniquely positioned” to deliver the president’s agenda because it had an approved reactor design, a viable supply chain and recent experience of building two of its AP1000 reactors in Georgia.

“There is active engagement with the administration, including key points of interface with the loan programmes office, recognising the importance of financing to the deployment of the model,” he said in an interview. 

“There are 10 large nuclear reactors in the executive order and we believe that we can do them all with AP1000 reactors . . . Our customers, the hyperscalers, the tech firms, the suppliers are all coming together to try to figure out exactly how to deploy.”

Based on estimates from the Department of Energy, building 10 large nuclear reactors in the US could cost $75bn without accounting for delays or cost overruns, according to TD Cowen, an investment bank.   

Westinghouse, which is jointly owned by private equity group Brookfield and uranium miner Cameco, has enjoyed success with its AP1000 reactor, a pressurised water reactor operating at several locations in the US and China.

At least a dozen additional plants are either in construction or under contract in Poland, China, Ukraine and Bulgaria.

Westinghouse faces limited competition in the US market as global industry leaders, including Russia’s Rosatom and China General Nuclear Power Group are unlikely to win contracts due to geopolitical factors.

GE Vernova, which has a joint venture with Hitachi, has not built a large reactor in decades in the US and has shifted its focus to small modular reactors (SMRs), a new type of reactor design that generates about a third or less of the power capacity of standard units.

Korea’s Kepco has a US approved reactor design but has never built a large-scale reactor in the country, say analysts, and France’s EDF withdrew from the US nuclear reactor market almost a decade ago.  

Adam Stein, a nuclear expert at the Breakthrough Institute, a Washington-based group, said the limited number of US approved designs was a boost to Westinghouse but he added that building 10 large reactors was very ambitious and challenging.   

“The US doesn’t have the most favourable market for building large new nuclear right now because of the type of the electricity market that doesn’t guarantee cost recovery in most cases,” he said.

“The executive order is not a direct mandate. It’s still a decision for the local utility to invest in new reactors and a state’s Public Utility Commission to consider that cost to the ratepayers. That makes it challenging to build large reactors right now.”

Delays in the construction of two AP1000 reactors at Georgia’s Vogtle nuclear power plant in recent years caused costs to more than double beyond their initial $14bn estimate, dimming US utilities’ enthusiasm for large reactors.      

Sumner said the first-of-a-kind construction challenges with the AP1000 in Georgia had been solved due to learning from deployments in the US and China.

“The design is frozen . . . We are the only firm in the world that has done modular nuclear construction and we have all of [that] real life learning now embedded in our go forward delivery models.”

But it is not clear if utilities and tech groups, such as Microsoft, Google and Amazon, are prepared to invest tens of billions of dollars to jump start construction of large scale nuclear plants in the US.

SMR developers are talking to US officials and utilities about co-locating multiple reactors on a single site to provide similar generation capacity, which they claim will reduce construction risks.

NuScale, which has an SMR design approved by US regulators, told the Financial Times it could deploy 12 of its 77MWe (megawatt electrical) reactor modules to support a plant with 924MW of capacity.

Kelly Trice, president of Holtec International, an US-based SMR developer, said grouping two or three of its 320MWe reactors together would enable it to compete against any large-scale nuclear plant.

“We think we can do the same for less costs, for fewer people, to operate, less maintenance and simpler. So, we fully intend to compete with the big plants,” he said.

Dubious pledges and uncertainty cast a shadow over the president’s manufacturing boom

If there is one thing US President Donald Trump obsesses over even more than the size of crowds at his rallies, it is the level of investment pledges he has secured from companies and countries.

The White House maintains a running list of the president’s second-term deals on a webpage entitled “Trump effect”. The commander-in-chief recently claimed that commitments had already exceeded $10tn, supposedly proving that his plans to spark a tariff-driven manufacturing boom in America will pay off. This week, I stress-test his claims.

The White House’s list is undoubtedly impressive. However, the $10tn figure — which Trump cited early last month — is misleading.

For starters, in mid-May, Goldman Sachs estimated that companies had announced plans to invest over $2tn across multiple years, with foreign governments pledging in excess of $4tn. That comes to a still chunky number of at least $6tn — roughly 20 per cent of the US economy. (Further commitments have also been made since Goldman’s analysis.)

The cumulative value of total greenfield foreign direct investment announcements — which captures capital earmarked for new facilities and operations — is also already well ahead of where it was at the equivalent point in both Trump’s first term and Joe Biden’s administration, according to data from fDi Markets, an FT-owned database.

But how much of all this will actually happen?

First, pledges are, of course, not the same as realised investments. It is typical for companies and countries to announce projects early in a president’s term to curry favour with their administration.

Looking at promises made during Trump’s first term, Goldman Sachs estimates that 80 per cent of investment commitments actually materialised. Though solid, that did include some high-profile projects falling short of their goals. Alibaba scrapped a plan unveiled in January 2017 that was slated to create 1mn jobs. Foxconn watered down a manufacturing investment plan in Wisconsin from $10bn to only $672mn.

The incentive to embellish investment plans is even stronger in Trump’s second term, given his broad tariff threats. “The US’s partners have a record going back to the 1980s of allaying trade tensions and US concerns by committing direct investment,” notes Matt Gertken, a chief strategist at BCA Research. “After reviewing the major items on the White House list, we find that many of them are indeed hyperbolic and meant for political effect.”

This is supported by data from fDi Markets. The group also tracks signals that a foreign company may be considering investment, such as a new investment strategy. It finds that signals to invest in the US, as a share of all global signals made by businesses, have jumped to their highest in well over a decade so far this year.

This suggests an 80 per cent conversion rate is highly unlikely.

But the White House’s tally is deceptive in other ways too, according to analysis by the Cato Institute. “The list includes previously planned projects that are already under way,” says Scott Lincicome, vice-president at the think-tank. “Other items have classic wiggle room, such as ambiguous timeframes, or are conditioned on the economic environment.”

Some of the larger commitments are indeed curious. Apple has promised to invest $500bn in the US over the next four years. Yet it only spent $10bn on capital expenditure and $31.4bn on research and development globally last year, notes Goldman Sachs. Nvidia’s $500bn commitment is similarly dubious.

For companies, the chasm between promised and current investment suggests the headline-grabbing figures could be bolstered by partnerships, acquisitions or costs of production. As for promises from countries, both Saudi Arabia and Qatar’s “deals” includes purchasing US goods, which boosts American exports, not investment.

This raises a broader point: big investment numbers do not necessarily translate into GDP or job gains.

Tellingly, the investment announcements have not driven significant upgrades to capital expenditure expectations at a company level. Close to 70 per cent of Goldman Sachs’ equity analysts who cover companies with recent investment promises say the pledges mostly overlap with prior plans.

“The consensus forecasts for investment and economic growth have not changed in response to these announcements”, says Mark Zandi, chief economist at Moody’s Analytics. “If anything, the fundamental drivers of investment have weakened considerably because of the global trade war.”

Whether they are genuinely new investments or not, many projects are likely to be scuppered because of uncertainty. Indeed, BCA Research’s indicator of business investment intentions across the US is in recessionary territory.

Indicatively, America’s construction spending boom, linked to the Inflation Reduction Act and the Chips and Science Act, appears to have peaked. Trump’s well-known dislike of the Biden-era initiatives has created uncertainty around the status of their tax credits and subsidies. Investment under both programmes tapered off as the 2024 election came into close view.

The lack of clarity over tariffs in particular significantly reduces the chances of successfully pulling off any fresh building plans. Import duty rates are subject to the White House’s negotiations with America’s trade partners, and now the courts too.

Relocating manufacturing production involves substantive fixed costs, often at multiples of average gross operating surplus. Rational boardrooms will not risk years of profit by breaking ground on new facilities if tariff rates shift again and render investment less competitive. Across many industries, US production costs are significantly higher than the top three countries currently exporting to the US, according to Goldman’s analysis.

Both Trump’s sector- and countrywide levy plans matter here. For example, car manufacturers weighing up whether to jump the tariff wall need to keep tabs on levies concerning their inputs, such as steel and copper.

Beyond tariffs, there are other limiting factors. Over two-thirds of manufacturing businesses considering producing more goods in the US cite the availability of qualified labour as a significant concern, according to a recent Bank of America Global Research survey.

The administration’s broader assault on universities and research risks undermining access to high-skilled workers, while a clampdown on undocumented migrants would hit the construction labour force. Permitting processes are also notoriously slow.

And then there is Section 899, a provision in Trump’s One Big Beautiful Bill Act that would give the Treasury secretary the power to set retaliatory taxes on inbound foreign investment. The Tax Foundation estimates that over 80 per cent of current US FDI stock comes from nations covered by the legislation.

All said, the obstacles to completing any proposed factory build, product launch or hiring spree during Trump’s second term are unprecedented. And, even if projects do occur, they may not deliver the desired results.

Maurice Obstfeld, a senior fellow at the Peterson Institute for International Economics, looked at consumer goods company Kimberly-Clark’s $2bn pledge last month, which he says is linked to a strategic restructuring and cost-cutting plan unveiled a year ago. “While Kimberly-Clark certainly wants to expand its US manufacturing capacity, the new facilities announced appear highly automated and will use high-skilled labour — not necessarily the blue-collar jobs Trump has been promising.”

Indeed, since production costs and labour availability are primary concerns, US reshoring is unlikely to drive significant employment. “Those companies that do end up reshoring are likely to offset associated costs via increased automation”, adds BofA Global Research.

Trump is not alone in overegging investment pledges. Most presidents do it. The Biden-Harris White House touted over $1tn in private sector commitments, even though many projects have been delayed or paused.

Yet the discrepancy between this current administration’s running list of investment announcements and what actually happens is likely to be one of the most exaggerated, given Trump’s transactional and capricious policymaking.

Net total investment could even look dire by the end of Trump’s second term, if it continues as it has started. Right now, domestic capital expenditure projects are largely on hold or cancelled. Outbound FDI projects may even pick up as companies try to avoid retaliatory measures on the US. Automakers are considering shifting to China, given its stranglehold on rare-earth magnets.

Over time, there is reason to be optimistic. For some companies and nations, announcing plans to invest in the US is a longer-term strategic play to gain exposure to its unrivalled market and technology (even if the conditions aren’t quite right now). “Some pledges suggest a desire to leverage the promise of a future upside to stave off the reality of near-term threats,” says Clayton Allen, US director at Eurasia Group. If policy stabilises, it could catalyse dormant projects.

For now, however, Trump’s investment deals are largely a mirage.

Apple’s struggles to update Siri lead to investor concerns over AI strategy

iPhone-maker hit by technological challenges that have led to delays to the full rollout of its ‘Apple Intelligence’ features

Apple is struggling to deliver upgrades to its artificial intelligence voice assistant for the iPhone, with investors downbeat about the potential for major AI announcements at its flagship annual event next week.

Recently departed employees told the Financial Times that the Silicon Valley giant has been hit by challenges with updating Siri using cutting-edge large language models that can deliver more sophisticated responses to spoken prompts.

Apple has been attempting to build its own LLMs over the machine learning technology that currently powers Siri, a product already used in hundreds of millions of its bestselling devices, with the aim of creating a truly conversational assistant.

Former executives said that the process of integrating the technologies has led to bugs, an issue not faced by competitors such as OpenAI which have built generative AI-based voice assistants from scratch.

One former Apple executive said: “It was obvious that you were not going to revamp Siri by doing what executives called ‘climbing the hill’,” meaning to incrementally develop the product rather than rebuilding it from the ground up.

“It’s clear that they stumbled,” the person added.

The updates to Siri form a key part of “Apple Intelligence,” a suite of AI features announced at the company’s Worldwide Developer Conference last year and intended to boost hardware sales.

The FT reported this week that Apple’s attempt to rollout the AI features in China, powered by models made by Alibaba, is being held up by a Beijing regulator. Sensitive deals in the country involving American tech companies have come under closer scrutiny in response to US President Donald Trump’s trade war.

Repeated failures to release Apple Intelligence features that have already been announced has meant expectations are low for this year’s WWDC, which kicks off next week.

“We’re at the point where investors already know what the good news potentially is, and it’s about: let’s first have you deliver what you promised last year,” says Samik Chatterjee at JPMorgan. 

The AI struggles have weighed on the tech giant’s stock. It has been worst-performing of the so-called Magnificent 7 tech stocks in 2025, down around 18 per cent since the start of the year and below the tech-heavy Nasdaq which is largely flat. 

Trump’s tariffs, competitive threats in China, and legal pressure on Apple’s high-margin services business have also led to investor concerns about its long-term growth.

At the core of Apple’s AI troubles is Siri, its legacy voice assistant which is seen as critical to unlocking true “agentic” abilities on the iPhone and other Apple devices.

When ChatGPT launched in late 2022, “the way companies were doing conversational interaction was changing rapidly, and it was clear Siri was coming up short,” said another former senior Apple employee who worked on the technology ahead of the launch.

The person added that they were “surprised” to see features announced last year that were ultimately “not going to make it” in time for Apple Intelligence’s initial release.

As well as operating much larger and more powerful models, the likes of OpenAI, Google and Perplexity all have launched voice assistants that are widely viewed as smarter than Apple’s.

The iPhone-maker’s answer was to focus last year’s annual developer conference on its own AI push, where it teased an AI-upgraded assistant able to read the user’s screen, draw on their contextual information and take actions within their apps.

A group of AI features such as writing aids, image and emoji generation and camera-based search have already hit the market.

The heralded changes to Siri are yet to be released, however. Chief executive Tim Cook recently admitted the technology did not meet the company’s “high quality bar” and was “taking a bit longer than we thought.”

The delays led to Apple pulling TV ads featuring The Last of Us star Bella Ramsey that promoted the new Siri update. The company drew a number of false advertising lawsuits from consumers.

The current delays to Siri mean that Apple is essentially three years or more away from delivering “a truly modern AI assistant, long after Google and others have integrated such tech,” Bank of America analysts wrote on Monday.

The failures have led to changes at Apple. John Giannandrea, its AI guru poached from Google in 2018, saw the Siri product division removed from his remit earlier this year and transferred to Mike Rockwell, the executive behind the Vision Pro headset.

A former Apple executive said that fragmented leadership teams led to a lack of a unified strategy around AI, made worse by an initial lack of appetite on the part of top executives to allocate a big enough budget for the build-out of the technology.

Another challenge is Apple’s focus on user privacy and security. It has prioritised running its AI features through smaller models and user data staying on the device, which former employees said adds another layer of complexity to the challenge.

This stands in contrast to larger LLMs such as those that power OpenAI’s ChatGPT, which run through the cloud on powerful servers. Apple has leaned on OpenAI by releasing ChatGPT integration with Siri.

Since then, OpenAI has signalled its own ambitions in the hardware space, with chief executive Sam Altman announcing a $6.5bn deal to acquire IO, the company founded by former Apple designer Jony Ive, who will now be creating products for a potential rival. Apple shares fell about 2 per cent on the news.

Companies are freezing hiring and investment to deal with shifting tariff policies. ‘Even Trump doesn’t know what Trump will do next.’

Key Points

The U.S. economy, which weathered false recession alarms in 2023 and 2024, is entering another uncomfortable summer.

Job growth held steady in May, with the economy adding 139,000 jobs. The unemployment rate has stayed in a tight range, between 4% and 4.2%, over the past year.

But there are cracks beneath the surface. Businesses are warning that constantly shifting trade policies are interfering with their ability to plan for the future, leading to hiring and investment freezes.

Policy uncertainty has unfolded against the backdrop of an economy with slower job growth and a cooling housing market. Compared with last year, the Federal Reserve is more reluctant to cut interest rates because officials are worried about new inflation risks.

John Starr, the owner of UltraSource, an importer and manufacturer of meat-processing technology in Kansas City, Mo., said he is hunkering down—no hiring, no more capital spending—until he has clarity on tariffs.

The company is waiting for suppliers in Europe to finish work on $20 million in orders it placed before 10% tariffs took effect on April 9. That means he faces a $2 million levy if tariffs stay at that level.

“How am I supposed to pay this?” said Starr, a third-generation owner of the company. “That could wipe out profits for a year.”

Whether the economy again bends, rather than breaks, turns on how the U.S. consumer handles the latest curveball—this time from President Trump’s desire to reorder America’s trading relationships and reduce reliance on imported goods. For months, the president has announced one large tariff increase after another, at times wavering from escalation to temporary resolution.

“Where this goes all depends on what Trump decides to do next, and candidly, even Trump doesn’t know what Trump will do next,” said Christopher Thornberg, founding partner at Beacon Economics in Los Angeles. “So it’s almost impossible to see where this thing is heading.”

Economists largely agree that for the U.S. economy to slide into recession, the American consumer needs to falter.

“As long as the consumer is doing OK, it’s not going to change our world,” said Ric Campo, chief executive of Camden Property Trust, a Houston-based developer and owner of 58,000 apartment homes.

Most economists think the prospects of a recession are higher than they were at the beginning of the year but lower than in April and early May, when tariffs on China had been increased by 145%.

The U.S. agreed to roll back the tariff increase to 30% last month. Most other nations face 10% tariff increases, with higher rates on dozens of countries paused until early July.

Three risks loom large.

• First, the U.S. labor market has been in an uneasy equilibrium where companies aren’t hiring but are reluctant to fire workers that they hustled to find three or four years ago. Like a beach ball that shoots skyward after being held underwater, joblessness can quickly jump once companies decide demand is too soft to keep those workers.

“It starts with one large firm. Then competitors might say, ‘Well, listen, we have to do the same,’ ” said Gregory Daco, chief economist at consulting firm EY.

Bill Hutton, president of Titan Steel, a Baltimore-based distributor and processor of mostly imported steel for products such as paint cans, said he is going to “err on the side of caution” when it comes to reducing the size of his workforce. Having worked so hard to staff up, he said, he is “very, very leery of making assumptions that, ‘Oh, we can dial up or down our workforce at a moment’s notice.’ ”

• Second, consumers could finally push back against rising costs, forcing companies to tighten their belts.

Delinquency rates on consumer debt have been on the rise for a year, raising fears that deteriorating finances for low-income borrowers could lead to a more pronounced slowdown in consumer spending.

For the housing market, the spring sales season has been a bust. The U.S. market now has nearly 500,000 more sellers than buyers, according to real-estate brokerage Redfin. That is the largest gap since its tally began in 2013. Home prices could fall 1% this year, said Redfin economist Chen Zhao.

“The market has been at rock bottom for the last 2½ years and there was some hope that we’d get a little bit of a turnaround this year. And it’s just actually been worse than expected,” said Zhao.

• Third, financial-market shocks or abrupt sentiment changes remain a wild card. The Fed reduced short-term interest rates by 1 percentage point last year, providing a measure of relief to borrowers with credit cards or variable-rate bank loans.

Officials hit pause on rate cuts this year amid concerns that tariffs might create new inflation risks. Longer-term borrowing rates, which aren’t set by the Fed and which influence many borrowing costs such as mortgages, have been elevated as investors around the world pay more attention to how governments will finance rising deficits in the years to come.

Any sudden and sustained rise in borrowing costs could spill over to the stock market, hurting companies’ earnings and making stocks less attractive. Lofty asset prices have supported business investment and high-income consumer spending.

For many companies, the uncertainty triggered by Trump’s sudden and seemingly arbitrary announcements of tariffs has upended the outlook for sales this year.

Starr, at UltraSource, orders equipment that has a lead time measured in months. Because those products are made to each customer’s specifications, Starr can’t resell them if clients refuse to eat the cost of the tariff.

“I have to take action now,” Starr said. “We’re going to be very careful about any cash expenditure just because we need that cash to pay the tariff.”

White House officials have said they are confident the president’s approach will lead to better trading relationships. “In order to get another country to eat the burden of the tariffs, we have to have a credible threat to move our supply chains across the border…It can take some time to make that threat credible,” Stephen Miran, chairman of the president’s Council of Economic Advisers, said in an interview.

Miran said he couldn’t produce a forecast for inflation this year because “we’re still waiting for policy details to be fully fleshed out right now.” He also said businesses would benefit from a tax-cut package moving through Congress.

Starr, who said he has already racked up $300,000 in unanticipated expenses because of tariffs, said the prospect of business tax cuts is of little use if his profits are zeroed out from tariffs. He said he doesn’t object to paying a 20% tariff on prospective orders as long as he has certainty the duty won’t suddenly change after he has negotiated purchase orders with customers and vendors.

Steel and aluminum tariffs, which Trump this past week raised to 50% from 25%, could boost domestic metal producers while squeezing profits for carmakers, can manufacturers, and companies such as Titan Steel. Hutton, the steel company’s president, said customers have been understanding about accepting some price increases because his competitors have also had to raise prices.

“It feels like we’re muddling through,” he said. “Nobody—neither us, nor our customers, nor our overseas supplier—is in any position to do any long-term thinking.”

The Fed aggressively raised rates in 2022 and 2023 to combat inflation. But the U.S. economy was insulated because many households and businesses had already refinanced at ultralow rates during the pandemic. Later, the economy benefited from an unexpected boom in capital spending on artificial intelligence.

Any pullback could be abrupt. “It’s very rare that you have a technology shock of this sort that doesn’t lead to overbuilding,” said Jason Thomas, chief economist at private-equity manager Carlyle Group.

Some companies have held back from raising prices now until they can see how tariffs settle out. “They just said, ‘We cannot take the risk of souring relations with our customers, with our suppliers, over a policy that in two months’ time may not even be in place,’” Thomas said.

He expects businesses eventually will have to pass along some cost increases, however, because they will have depleted inventories acquired at pretariff rates.

One tailwind—recent declines in energy prices—could help offset some of the inflationary impulses from tariffs.

While the president often gets undue credit for what goes right or wrong in the economy, this time could be an exception.

“The economy has a lot of momentum, and so if Trump truly backs off on tariffs and just calms down, you could see this expansion going another two, three years, honestly,” said Thornberg of Beacon Economics. “Then again, if he keeps rocking the boat, you can blow it up by the beginning of next year.”

Donald Trump and Elon Musk’s allies urge reconciliation after damaging split

Tech industry fears commercial and political damage from breakdown in relationship between Tesla boss and US president

Allies of Donald Trump and Elon Musk have urged the US president and his billionaire backer to repair their relationship, seeking to limit the political and commercial damage from this week’s spectacular split.

The fissure between the powerful pair, ostensibly over the president’s signature tax bill, threatens to derail the White House’s legislative agenda and wreck a hard-won alliance between Silicon Valley and Washington.

“It is unfortunate . . . I hope they will come back together,” Texas Senator Ted Cruz, who was in the Oval Office when Trump slammed Musk, said on Friday. “A lot of conservatives are feeling like this is not good, let’s hug and make up.”

The Tesla chief executive, who had spent Thursday launching more explosive attacks on Trump, seemed to be open to a détente, responding positively to hedge fund manager Bill Ackman, who urged the duo to “make peace for the benefit of our great country”.

Trump on Friday claimed he was “not even thinking about Elon”, before adding on CNN: “The poor guy’s got a problem . . . I won’t be speaking to him for a while I guess, but I wish him well.”

He repeated that line on Saturday, telling NBC he had no intention of mending their relationship. He warned of “serious consequences” if Musk funded candidates to run against Republicans.

Tech figures who backed the administration hoping it would usher in an era of tax cuts and deregulation have been racing to contain the quarrel, with limited success.

“Elon isn’t taking calls from anyone,” one Silicon Valley financier and big donor to Republican candidates told the Financial Times. “Not from people who have billions invested in his companies . . . The Valley is losing their shit.”

Attempts to get Trump to reconcile with his former “first buddy” were set back on Friday morning by news the president planned to sell or give away the Tesla he had bought in March as a show of support for Musk. The White House also dismissed reports of a conciliatory phone call between the two men.

Billionaire Tim Draper, who invested in Tesla and SpaceX, urged Trump and Musk to reunite to save the so-called Department of Government Efficiency, an initiative that was backed by many in the tech world and staffed by Silicon Valley executives.

“They seemed to be making good progress together,” Draper told the FT. “My advice: Don’t throw the baby out with the bath water.”

However, cracks in Silicon Valley and Washington’s marriage of convenience had been appearing for weeks, particularly over Trump’s tax bill that irked Musk. Deficit hawks have balked at the legislation adding trillions to the US debt pile, while more socially progressive tech figures have bridled at proposed cuts to entitlement programmes such as Medicaid.

“I am fully for pursuing the elimination of waste and fraud,” said Jon McNeill, a former Tesla president who worked alongside Musk and runs start-up incubator DVx Ventures. “But at the same time, I don’t want a tax break so badly as to make the most vulnerable suffer. And from what I’m hearing, a lot of my peers feel the same way.”

Musk’s allies aligned with Trump fear several Silicon Valley figures who followed him into government could find their roles are in peril.

The sudden deselection of Jared Isaacman, a tech founder and friend of Musk’s who had been nominated to lead Nasa, was the start of an expected “purge”, said one person close to the administration.

Among those considered to be at risk were crypto and artificial intelligence tsar David Sacks, policy adviser Sriram Krishnan, and Michael Grimes, Musk’s former banker at Morgan Stanley who is now an official at the Department of Commerce.

Sacks, who is also one of the co-hosts of the All-In podcast, a sounding board for the Trump-aligned tech world, was “shell-shocked”, said one person familiar with the matter, and needed to be protected from public scrutiny until things calmed down.

Usually a frequent poster on social media, Sacks has remained silent since the Musk-Trump relationship imploded. He did not immediately respond to a request for comment.

Other prominent tech figures have been debating whether reconciliation is possible and, if not, what life beyond the break-up would look like.

Ryan Selkis, founder of a crypto platform who became a prominent Trump backer, told the FT: “Elon will be back in the fold in a matter of weeks, but it will be a chastened Elon.”

But Delian Asparouhov, a space tech founder who co-runs the Hill & Valley Forum, which links Silicon Valley and Washington, said: “I don’t think there is going to be a de-escalation here.”

He told tech news site TBPN smaller space companies that work with Musk’s SpaceX could encounter “more resistance” from the White House.

Others bemoaned the souring of the tech community’s Trump bet. “Maybe Silicon Valley got played by Trump. He got what he wanted,” said one west coast venture capital founder, citing Musk’s $250mn donation to Trump’s campaign.

The person lamented the economic volatility — caused by tariffs and Trump’s unpredictability — during a presidency that they had been promised would be a boon to business. “We’re all experiencing a liquidity crunch,” they said. “We need public markets to open.”

The public bust-up between Trump and his benefactor could open the door for others in Silicon Valley to become tech’s de facto ambassadors in Washington, especially Musk’s arch-rival, OpenAI’s Sam Altman.

“Tech is not represented by one person,” an investor in Musk’s companies said. “This doesn’t end because of one person, even if he’s the most prominent person in the world.”

Executives prioritising break-ups or continuation funds as trade war kills off potential floats

Private equity groups are overhauling their exit strategies after accepting that a years-long downturn in initial public offerings is unlikely to end soon.

Buyout executives at the industry’s annual European conference this week said they were prioritising other options for exiting their investments, including breaking up businesses to sell them off in smaller parts or selling companies to themselves via “continuation funds”.

“I can’t remember in my 20 years of growth equity investing, not having an IPO window open for this kind of long period of time,” said General Atlantic co-president Gabriel Caillaux at the Berlin SuperReturn event. “That is obviously calling us to rethink not strategy, but some tactical aspects.”

Buyout firms have a record backlog of ageing and unsold assets, as higher interest rates and market turmoil have made it harder to float companies or sell at acceptable prices, putting pressure on them to find other ways to return cash to their investors.

The volume of private equity-backed IPOs has slumped since the frenzy of 2021, with only nine across Europe and the US this year compared with 116 in the same period in 2021, according to Dealogic.

The head of private equity at a large international firm said IPOs now ranked behind break-ups and minority stake sales as an exit option.

“The IPO is number three on the list these days,” they said.

Permira in January sold a minority stake in its €2.2bn luxury sneaker company Golden Goose after abandoning an IPO. EQT, which was last year reported to be considering a listing for its schools business Nord Anglia, eventually cashed out its older fund by selling to a consortium that included one of its newer funds.

Sellers were increasingly securing sales by offering buyers greater protection against risks, including through earnouts — where part of the price is linked to future performance, the private equity executive said. “The toolbox is really being opened now,” they added.

Executives had hoped the election of US President Donald Trump would lead to a revival in IPOs, but instead his policy volatility has closed the capital markets to most potential issuers.

In March, Permira and Hellman & Friedman postponed a planned IPO of US software group Genesys, while Bain Capital and Cinven did the same with their listing of German pharmaceuticals company Stada.

The head of private equity at a large global asset manager said that in the wake of Trump’s April 2 tariff announcements, listings were “gone”.

A top dealmaker at another of the world’s largest private capital firms said “the only thing that’s worse” than the current IPO market was “the perception of how strong it was supposed to be compared to how it’s turned out”.

Structural changes in the markets were also making it harder to list businesses, they added, including the rise of passive exchange traded funds that do not typically buy IPOs.

Daniel Lopez-Cruz, head of private equity at Investcorp, said the IPO market “for all intents and purposes is closed for private equity companies”.

The secondary market — where buyout firms sell assets to themselves with so-called continuation funds, or investors in private equity funds sell on their stakes in those funds — had become “a great help”, he said.

Continuation vehicles have soared in popularity in recent years as a means to return cash to fund investors. Private capital firms sold $75bn of assets on the secondary market last year, up 44 per cent from the previous year, according to Jefferies. The vast majority of that went into continuation funds.

Some executives remained positive about the possibility of IPOs making a comeback, however.

“Things can change very, very fast,” said the head of a major European buyout firm. “We have businesses in our pipeline that we’re considering IPOs for in nine or 12 months. It’s about being well prepared and going for it when you can.”