TheChek News Flow

Robots have long been seen as a bad bet for Silicon Valley investors — too complicated, capital-intensive and “boring, honestly,” says venture capitalist Modar Alaoui.

But the commercial boom in artificial intelligence has lit a spark under long-simmering visions to build humanoid robots that can move their mechanical bodies like humans and do things that people do.

Alaoui, founder of the Humanoids Summit, gathered more than 2,000 people this week, including top robotics engineers from Disney, Google and dozens of startups, to showcase their technology and debate what it will take to accelerate a nascent industry.

Alaoui says many researchers now believe humanoids or some other kind of physical embodiment of AI are “going to become the norm.”

“The question is really just how long it will take,” he said.

Disney’s contribution to the field, a walking robotic version of “Frozen” character Olaf, will be roaming on its own through Disneyland theme parks in Hong Kong and Paris early next year. Entertaining and highly complex robots that resemble a human — or a snowman — are already here, but the timeline for “general purpose” robots that are a productive member of a workplace or household is farther away.

Even at a conference designed to build enthusiasm for the technology, held at a Computer History Museum that’s a temple to Silicon Valley’s previous breakthroughs, skepticism remained high that truly humanlike robots will take root anytime soon.

“The humanoid space has a very, very big hill to climb,” said Cosima du Pasquier, founder and CEO of Haptica Robotics, which works to give robots a sense of touch. “There’s a lot of research that still needs to be solved.”

The Stanford University postdoctoral researcher came to the conference in Mountain View, California, just a week after incorporating her startup.

“The first customers are really the people here,” she said.

Researchers at the consultancy McKinsey & Company have counted about 50 companies around the world that have raised at least $100 million to develop humanoids, led by about 20 in China and 15 in North America.

China is leading in part due to government incentives for component production and robot adoption and a mandate last year “to have a humanoid ecosystem established by 2025,” said McKinsey partner Ani Kelkar. Displays by Chinese firms dominated the expo section of this week’s summit, held Thursday and Friday. The conference’s most prevalent humanoids were those made by China’s Unitree, in part because researchers in the U.S. buy the relatively cheap model to test their own software.

In the U.S., the advent of generative AI chatbots like OpenAI’s ChatGPT and Google’s Gemini has jolted the decades-old robotics industry in different ways. Investor excitement has poured money into ambitious startups aiming to build hardware that will bring a physical presence to the latest AI.

But it’s not just crossover hype — the same technical advances that made AI chatbots so good at language have played a role in teaching robots how to get better at performing tasks. Paired with computer vision, robots powered by “visual-language” models are trained to learn about their surroundings.

One of the most prominent skeptics is robotics pioneer Rodney Brooks, a co-founder of Roomba vacuum maker iRobot who wrote in September that “today’s humanoid robots will not learn how to be dexterous despite the hundreds of millions, or perhaps many billions of dollars, being donated by VCs and major tech companies to pay for their training.” Brooks didn’t attend but his essay was frequently mentioned.

Also missing was anyone speaking for Tesla CEO Elon Musk’s development of a humanoid called Optimus, a project that the billionaire is designing to be “extremely capable” and sold in high volumes. Musk said three years ago that people can probably buy an Optimus “within three to five years.”

The conference’s organizer, Alaoui, founder and general partner of ALM Ventures, previously worked on driver attention systems for the automotive industry and sees parallels between humanoids and the early years of self-driving cars.

Near the entrance to the summit venue, just blocks from Google’s headquarters, is a museum exhibit showing Google’s bubble-shaped 2014 prototype of a self-driving car. Eleven years later, robotaxis operated by Google affiliate Waymo are constantly plying the streets nearby.

Some robots with human elements are already being tested in workplaces. Oregon-based Agility Robotics announced shortly before the conference that it is bringing its tote-carrying warehouse robot Digit to a Texas distribution facility run by Mercado Libre, the Latin American e-commerce giant. Much like the Olaf robot, it has inverted legs that are more birdlike than human.

Industrial robots performing single tasks are already commonplace in car assembly and other manufacturing. They work with a level of speed and precision that’s difficult for today’s humanoids — or humans themselves — to match.

The head of a robotics trade group founded in 1974 is now lobbying the U.S. government to develop a stronger national strategy to advance the development of homegrown robots, be they humanoids or otherwise.

“We have a lot of strong technology, we have the AI expertise here in the U.S.,” said Jeff Burnstein, president of the Association for Advancing Automation, after touring the expo. “So I think it remains to be seen who is the ultimate leader in this. But right now, China has certainly a lot more momentum on humanoids.”

U.S. securities laws prohibit trading on “material nonpublic information.” But the SEC does not govern prediction markets contracts because they are not securities. Instead, it would be up to the Commodity Futures Trading Commission, which oversees futures trading, or the Department of Justice to pursue such cases, say lawyers.

Still, profiting off confidential information on a prediction market could violate an employee’s legal obligations to their employer. “It’s a form of fraud that’s akin to embezzlement because you’re secretly using the information for your own benefit,” said George Canellos, a lawyer for law firm Milbank LLP who specializes in corporate governance and securities law.

On Thursday, several firms including Kalshi and Coinbase said they formed a new industry group that will advocate for federal rather than state oversight. One of its first initiatives will focus on establishing national standards against insider trading.

Complicating matters is the fact that industry leaders have sometimes suggested there should be room for employees to bet on their own company’s activities. For example, Coinbase CEO Brian Armstrong said he was recently asked whether insider trading should be permitted in prediction markets.

Armstrong responded that it was not “clear-cut,” he recounted at The New York Times DealBook Summit last week. If people wanted to know whether the Suez Canal was going to reopen, he said, the market would be more accurate if an admiral on a ship in the canal was allowed to bet on it. On the other hand, “you want to preserve the integrity of those markets.”

In fact, some companies, including Google and Anthropic, have established their own internal prediction markets. Employees can bet—without using any real real money—on questions such as when a team will finish a project.

In these cases, the markets’ predictions are kept inside the company, so they do not harm the company, said Dan Schwarz, who built Google’s current prediction market and served as chief technology officer of forecasting site Metaculus. For these internal prediction markets, rather than discouraging insiders, “you’re trying to get insider trading,” he said. “You’re trying to get people to reveal what they know.”

In the increasingly risk-averse business of Hollywood, the best way to get a film greenlit is propose a rehash of something that’s been done before—a superhero sequel, perhaps, or a mash-up of two hit movies, like Elf meets The Godfather.

Maybe that’s why Hollywood’s latest all-consuming mega-drama—the war to take over Warner Bros. Discovery—has the media world hanging on every plot twist. It all rings so familiar, sort of like One Battle After Another meets Everything Everywhere All at Once. The fight between Netflix and Paramount Skydance to take over this grand old Hollywood name is both the latest chapter in a one-battle-after-another forever war and a massive everything-everywhere, all-hands-on-deck swirl.

I was struck by the former point while finishing up Barry Diller’s super-interesting memoir, Who Knew (makes a great holiday gift, according to Bill Gates), in which he recounts similar epic battles in which he engaged, most notably his struggle with Sumner Redstone to take over Paramount.

“It’s got similarities in some ways,” Diller says to me on the phone from Florida, noting the continued, inevitable consolidation of the business. “It’s going to be an auction. That’s what Paramount was with Viacom and QVC at the time—it’s just the numbers here are much greater, more blockbuster.” (Recall that Diller himself through his company IAC recently considered making a bid for Paramount but chose not to go up against David Ellison, as it would be “unwise to get in an auction with someone who has a pretty much unlimited balance sheet.”)

Just to level set: As of now WBD has accepted Netflix’s part-stock, part-cash offer for the movie studio and its streaming business (which includes HBO) of $27.50 per share, while Paramount is making an all-cash tender offer to WBD’s shareholders of $30 for the entire company, including its cable assets, such as CNN and TBS.

Diller isn’t the only mogul/billionaire with a deep-seated interest in how this battle royal plays out. Like some sort of celestial black hole swallowing up any and all nearby interstellar material, this takeover battle has sucked in an unprecedented cast of A-listers from the four power centers of America—Washington, Wall Street, Silicon Valley, and Hollywood—and beyond (never mind legions of directors, bankers, lawyers, flacks, and underlings).

Start with the principals David Zaslav, CEO of WBD; the Ellisons (père et fils), of Oracle and Paramount, respectively; Ted Sarandos and Reed Hastings of Netflix; and Gerry Cardinale, CEO of RedBird Capital (co-starring Jeff Zucker). And don’t think Comcast’s top brass, Brian Roberts and Mike Cavanagh, who dropped out of the bidding, aren’t still keeping abreast.

After that you have major Middle Eastern sovereign-wealth funds (Saudi Arabia’s PIF, Qatar’s QIA, Abu Dhabi’s L’imad), helping to bankroll Paramount’s latest offer, along with Jared Kushner’s Affinity Partners, and, of course, President Donald Trump—interested in what happens to CNN—is at the very least a keen observer, and at the very most a party who will have a say in the outcome.

Then there are debt commitments from Bank of America, Citigroup, and Apollo Global Management. Given the size and sensitivity of this deal, rest assured Brian Moynihan, Jane Fraser, and Marc Rowan are in the loop as well.

After that you have interested parties like Shari Redstone, who recently sold her controlling interest in Paramount; Jeff Bewkes, former CEO of Time Warner, who famously and dismissively referred to Netflix as “the Albanian army”; and Bob Iger of Walt Disney—who just plunked down a $1 billion investment in OpenAI this past week, nervously watching.

Then there’s John Stankey, CEO of AT&T, whose company, you may recall, owned Warners Bros. before spinning it off to Zaslav in April 2022. What does he think about what’s going down? “Not surprised,” Stankey said this week at the WSJ Leadership Institute CEO Council Summit in Washington. “When we made the decision to divest the asset, we felt there was no question there was going to be consolidation of media. If people were going to compete with what Netflix had built, they were going to have to have a different asset base to do that…. [I was] fully expecting there would probably be a follow-on to get the kind of scale that it needed. I wouldn’t say that four years ago I believed it was necessarily going to be this particular outcome, but that there would be a transaction. I held on to all my stock in the company during this period of time waiting for this day.”

At the time of the WBD spinoff, Stankey owned some 900,000 shares of AT&T. Using the exchange ratio of 0.241917 shares of WBD for each share of T, that would give him nearly 218,000 shares of WBD. If it winds up going with the Netflix offer of $30 a share, that would be worth about $6.5 million. (It’s worth noting that while WBD’s stock is up 20.7% since the spinoff, the S&P 500 is up 62.4%.)

So what will happen? Cardinale of RedBird, who is David Ellison’s partner in Paramount Skydance, has no doubt. “Our offer is better,” he told me. “I put $2 billion dollars into Paramount and am committing another $2 billion for the WBD deal. I’m betting my firm on it.”

Still, given that Trump has an interest in the deal—even though it’s one that appears to favor Paramount—handicapping the outcome is a fool’s errand. Trump and Zaslav may have different agendas, which for now, at least, look to be at odds with each other. The president seems to be intent on seeing through regime change at CNN. That is a more clear shot with Paramount owning WBD lock, stock, and barrel, rather than in the Netflix deal, where CNN—along with TNT, TBS, and the Discovery Channel—would be spun off into a separate publicly traded company called Discovery Global.

What about Zaslav? First, note that this battle is in a sense David (Ellison) versus David (Zaslav)—though they more closely resemble Goliaths—as Zaslav has been reluctant to turn over his baby to Skydance and Paramount, where he would become co-CEO with David Ellison. On the other hand, Ted Sarandos, the sometimes-lampooned in Hollywood (even by himself—see episode eight of Seth Rogen’s The Studio) co-CEO of Netflix, seems to have ingratiated himself with Zas.

“Zaslav cares about the outcome that delivers him big money and where he’s still a big macher with the house in Beverly Hills,” says a senior executive who is extremely well acquainted with Warner Bros. Discovery. While no role for Zaslav in a post-Netflix acquisition has been announced or publicly promised, here’s what Paramount lawyers wrote in a letter to Zaslav last week:

“Paramount has a credible basis to believe that the sales process has been tainted by management conflicts, including certain members of management’s potential personal interests in post-transaction roles and compensation as a result of the economic incentives embedded in recent amendments to employment arrangements.”

Read between them lines!

Both Netflix and Paramount face massive breakup fees: $5.8 billion to WBD from Netflix if the streaming giant backs out, $5 billion paid by Paramount to WBD if it wins but doesn’t get regulatory clearance, and $2.8 billion from WBD to Netflix if it ends up going with Paramount (which Paramount would essentially cover). Whew! Got that? “Even in today’s heady atmosphere of some giddy dealmaking, those are still really big numbers,” says Jeffrey Sonnenfeld, a professor at the Yale School of Management.

All that’s as of now. Don’t be surprised if Paramount and Skydance look to sweeten their offer more to Zaslav’s liking, or for Netflix to put out a more explicit plan for CNN. Either one of those moves could tip the balance.

There is another way out, coming from that next-media world of prediction markets, which is in itself worth noting. Polymarket recently had Paramount with a 48% chance of winning WBD, versus Netflix with 36%. That prompted a wag at Ramp Capital to note that “if I were Netflix, I would just bet on Paramount then drop out and double my money.”

How’s that for a post-Hollywood ending?

How Rafael Grossi risks his life tracking the world’s most dangerous material.

In September, 2022, Rafael Mariano Grossi, an elegant Argentinean diplomat with tousled salt-and-pepper hair, led a convoy of nuclear experts toward the sprawling Zaporizhzhia power plant, in southeast Ukraine, which had been seized by Russia in the early days of its invasion. It is the largest nuclear facility in Europe and the first ever to be on the front line of a war. The dangers of a radioactive catastrophe were unprecedented. Explosions near the site had already damaged a high-voltage power line; Ukraine feared the failure of cooling systems that prevent nuclear fuel from melting down. Neither Ukraine nor Russia had promised full access to Grossi, the director-general of the International Atomic Energy Agency, the U.N. watchdog that has a mandate to secure nuclear plants and report whether their materials are diverted to make bombs. Grossi had to argue his way through Ukrainian checkpoints that refused passage. Then, in the no man’s land between the Ukrainian and Russian militaries, his team—wearing blue helmets and bulletproof vests marked “United Nations”—came under fire. He had to make a snap decision, without knowing which side was shooting. “I asked the security people, ‘What is the worst that can happen to us?’ ” he told me recently. He didn’t want his experts to be trapped or killed, but the mission would establish a right for the I.A.E.A. to access nuclear facilities in any future conflict.

There are now about four hundred nuclear power plants worldwide, in over thirty countries on five continents, with sixty more under construction. Another thirty countries are considering or planning to build nuclear power plants as global energy demand soars, especially to fuel the data centers that support artificial intelligence. “We are on the brink of a renaissance of the nuclear industry,” Serhii Plokhy, a Harvard historian, writes, in “The Nuclear Age: An Epic Race for Arms, Power, and Survival.” Russia’s seizure of Zaporizhzhia suggests that reactors “can be used as weapons of war.”

Plokhy and other experts worry about two types of nuclear proliferation. Vertical proliferation is when the nations that already have nuclear bombs—there are nine—add to their arsenals, as China is doing now. In October, President Trump pledged to re-start nuclear-weapons tests, which haven’t been carried out by the U.S. since 1992, deepening fears about a new era of brinkmanship. But the I.A.E.A. is most concerned about horizontal proliferation, as additional countries, non-state actors, or terrorist groups develop nuclear weapons—by purchase, creation, or theft of the technology. Grossi warned, “At the risk of being provocative, horizontal proliferation is far more destabilizing than vertical proliferation.” At least thirty nations beyond the nuclear nine may have the capacity to develop nuclear weapons, according to the I.A.E.A. Plokhy puts it closer to forty.

Near Zaporizhzhia, Grossi encouraged his experts—all of whom had volunteered for the mission—to think about their families. “I will always remember this guy,” Grossi told me. “He was from Moldova—a short, stocky guy, but very determined. He said, ‘I’m with you.’ ” The convoy began to cross the no man’s land. “At certain moments, you have to push the envelope a little bit,” Grossi said. “People are not going to receive you with a red carpet. It’s a war. What we did had never been done before.” He recalled seeing “carcasses of cars, debris, shoes of someone who was blown up, rotten pigs, scorched houses.” The team members entered Russian-held territory and established a presence at Zaporizhzhia. They have monitored it ever since. Amid renewed U.S. diplomacy to end the war, Grossi has insisted that Zaporizhzhia get “special status” in any peace deal—and that I.A.E.A. inspectors, who rotate in and out every few weeks, continue to insure safety at the plant.

“That’s not a cookie-pushing bureaucrat,” Daniel Poneman, a former Deputy Secretary of Energy, told me. Grossi “really goes toe to toe” with difficult people on both sides of a crisis. He has met an equal number of times with the Russian President, Vladimir Putin, and with President Volodymyr Zelensky, of Ukraine. Zelensky initially had limited nuclear expertise; he was also angry that Grossi was dealing with Russia on Zaporizhzhia, which Ukraine claimed as sovereign territory. Putin, whom Grossi has talked with one on one, across a small table in the Kremlin, has displayed a “surprisingly high level of technical knowledge,” Grossi said. Laura Holgate, a former U.S. Ambassador to the I.A.E.A., told me that Grossi “is always very careful when moving into new space to say, ‘This is the authority under which I’m doing this.’ ” His predecessor, Yukiya Amano, would have been “very poorly equipped” to manage the Zaporizhzhia crisis, she added. “He would be hiding under his desk.” Amano, a Japanese diplomat, died on the job, in 2019; Grossi, his deputy, was elected by the I.A.E.A. board of governors to succeed him. Holgate noted that Grossi has since “boldly” expanded the agency’s mission more than anyone who came before.

Dwight Eisenhower proposed the I.A.E.A., in 1953, in his “Atoms for Peace” speech to the United Nations. Amid the escalating race for nuclear weapons, Eisenhower sought to promote global disarmament; he called for knowledge of the atom to serve “peaceful pursuits” in medicine, agriculture, and electric energy in “power-starved” countries. The I.A.E.A. was created four years later, and headquartered in Vienna. It had a tumultuous start, as member states quarrelled over leadership and mission. The Cold War “raged more fiercely in the Board room of the I.A.E.A. than in the halls of the U.N.,” the diplomat and Nobel Peace Prize winner Ralph Bunche reportedly said during a visit, in 1958, on behalf of the U.N. Secretary-General Dag Hammarskjöld.

Grossi is the son of Italian intellectuals—his mother was a sculptor, his father a journalist—who immigrated to Argentina. He was born in January, 1961, early in the nuclear age. Disarmament seemed distant at the time. In October of that year, the Soviet Union detonated Tsar Bomba over the Arctic. It still ranks as the largest bomb ever exploded—three thousand times more powerful than the one that the U.S. dropped on Hiroshima, in 1945, killing an estimated hundred and forty thousand people. The so-called Nuclear Club included the U.S., the Soviet Union, Britain, and France; China soon joined, too. By the end of the nineteen-sixties, the five nations had almost forty thousand warheads. Interest in nuclear arms increased in other countries, too.

Argentina was the first Latin American country to have a nuclear-research reactor, a product of the Atoms for Peace program. Under military rule in the late seventies and early eighties, it “amply demonstrated” interest in producing fuel that could be used for a bomb, according to the C.I.A. The I.A.E.A. surveillance system discovered “accounting gaps” in the plutonium fuel being loaded and removed at the Atucha I nuclear plant, outside Buenos Aires. After Argentina’s return to civilian rule in 1983, the government acknowledged the covert program and abandoned it.

Adolfo Saracho, a senior diplomat and arms expert, soon created the Department for Nuclear Affairs and Disarmament in the Argentinean Foreign Ministry. “Saracho was a kind of Pied Piper, who was surrounded by young, smart, passionate kids he mentored,” Poneman, a nuclear-security expert who was in Buenos Aires at the time, recalled. Grossi was “a wet-behind-the-ears, newly minted diplomat” in Saracho’s orbit, Poneman said. “Rafa always had a kind of vision, even for a kid at that point, in his tender years, with a lively intellect, already charismatic, and with genuine gravitas. He stood out.”

Grossi has now spent four decades on the issues outlined in Eisenhower’s speech. In 2023, he addressed the U.N. General Assembly from the same dais where Eisenhower had spoken. “Atoms for Peace is more relevant than ever,” he said. “Every day on every continent, the I.A.E.A. supports nations in overcoming challenges like disease, poverty, hunger, pollution, and climate change by seizing opportunities to improve health care, agriculture, and energy systems through the power of nuclear science and technology.”

This year, Grossi persuaded the World Bank to end its decades-long ban on funding nuclear-energy projects; the agreement was signed in June, opening the way for the bank to support initiatives in developing countries. Grossi also created the Rays of Hope program, to expand global access to cancer detection and care. As a medical treatment, radiation had saved millions of lives “by turning cancers that were death sentences into curable diseases,” he said, in a speech in Ethiopia launching the initiative. “But these lifesaving advances have passed half the world by.”

Still, Grossi has generated more headlines in his role as the watchdog checking for cheaters—as Argentina once was. The Nuclear Non-Proliferation Treaty, or N.P.T., which went into effect in 1970, authorizes Grossi’s agency to monitor the nuclear facilities in all countries that have signed it; the I.A.E.A. can deploy cameras, conduct on-site inspections, and investigate suspicious activity. (The treaty currently has a hundred and ninety-one signatories.)

Iran was one of the original signatories. It is now the I.A.E.A.’s crisis case. A year ago, Grossi visited Fordo, the most advanced nuclear facility in the country. It was “very unassuming,” he told me. “Think about it as an underground parking garage. The difference is, instead of cars, it had labs and centrifuge halls and research-and-development places. It is a major piece of architecture.” Trucks could transport personnel and equipment into the complex; Grossi’s team opted to walk down a circular ramp almost three hundred feet underground. The facility is at the edge of the Alborz Mountains, a range considered in ancient times to be the home of mythical gods and an entrance to the afterlife. In the twenty-first century, it has hidden the centerpiece of Iran’s contentious nuclear program.

In June, the I.A.E.A. board of governors declared for the first time in two decades that Iran had violated the safeguard provisions outlined in the N.P.T. It cited the Islamic Republic for “many failures to uphold its obligations since 2019” on nuclear material and activities at multiple undeclared locations in Iran. I.A.E.A. declarations are based on reports prepared by Grossi. “That report did not say anything that we had not said before,” Grossi told me. “Of course, it was stern and serious about Iran’s lack of answers and coöperation on many fronts. At the same time, I said in black-and-white that there was no systematic nuclear-weapons program in Iran.” (The board includes representatives from the first five nuclear powers and thirty other rotating members. Nineteen countries supported the Iran resolution, eleven abstained, two declined to vote, and three—China, Russia, and Burkina Faso—opposed it.)

Shortly after the I.A.E.A. resolution, Israel bombed military, nuclear, and political headquarters across Iran, including Fordo’s surface facilities and access roads. U.S. B-2 stealth warplanes later dropped a dozen bunker-busting bombs, each weighing thirty thousand pounds, directly into Fordo. Ali Larijani, the head of Iran’s Supreme National Security Council, blamed Grossi personally for what would later be dubbed the Twelve-Day War; he vowed that Iran would “settle” with the I.A.E.A. director-general after it ended. Kayhan, a hard-line newspaper considered the mouthpiece of Supreme Leader Ayatollah Ali Khamenei, called Grossi, who is Catholic, a Mossad agent. It warned that he would be tried and executed if he returned to Iran. There have since been more graphic threats.

When I asked Grossi about all this, he sighed in frustration. “We live in a world of perceptions,” he said. “People who talk about my report have never read it in reality. And so they peddle things. So they chose to do this.” In Vienna, Grossi now has round-the-clock protection from Cobra, Austria’s antiterrorism police force.

Grossi is still dealing with the aftermath of the war—and the lingering questions about Iran’s nuclear assets. President Trump claimed that the U.S. had “obliterated” Fordo; separate U.S. airstrikes had also hit nuclear facilities at Isfahan and Natanz. Grossi, who is meticulous in composing his reports, told me, “There’s political language, and there is technical language, and they are not necessarily contradictory.” The damage at Fordo was “extremely serious, if not total,” he said. Centrifuges are used to separate liquids from solids, blood from plasma, fat from milk—or lighter uranium molecules from heavier ones to produce an enriched fuel. They are like “very sophisticated washing machines” that spin at furious velocity, he explained. “So all these systems, when you have such shocks, become unstable, unusable.”

The big unknown is the fate of Iran’s huge stockpile of uranium—some four hundred kilograms—which had been enriched before the war. Under the former nuclear deal between Iran and the U.S., France, Britain, Germany, China, and Russia, Tehran was allowed three hundred kilograms of uranium enriched to less than four per cent, limiting the material’s use to civilian power and medical research. (Under the N.P.T., the country has the right to produce nuclear energy.) But, in 2018, Trump abandoned the deal. Iran countered by developing more advanced and faster centrifuges and enriching uranium to sixty per cent—a short technical step away from ninety per cent, which is bomb grade.

In October, Grossi estimated that Iran could already fuel up to ten bombs with uranium enriched to sixty per cent, if the government made the political decision to weaponize. (Other steps are required to marry the fuel and warheads with delivery systems.) Grossi still communicates with the Iranian Foreign Minister, Abbas Araghchi, about I.A.E.A. inspectors returning to Fordo, Isfahan, and Natanz. He wants to determine if Iran is covertly using its stockpile. On October 29th, Grossi said that the I.A.E.A. had detected movement near where the stockpiles had been stored. Recent satellite images also reveal that Iran has continued to tunnel deep under the Zagros Mountains on a different facility—called Pickaxe Mountain, or Kuh-e Kolang Gaz La, in Farsi—near Natanz. It was not hit during the Twelve-Day War, and experts claim that it may be beyond the reach of U.S. bunker-busting bombs. Grossi has pressed Iran for an explanation and access, but has been repeatedly rebuffed.

Grossi, who speaks seven languages and dreams in several of them, too, keeps a regimented schedule. He wakes up at 5 A.M. and runs for two hours. He has one coffee at eight, then eats nothing until 1 P.M., when he has what staff call his “leaves.” Grossi nibbled on a small kale salad when I met him in D.C. He has tea—loaded with sugar—at 5 P.M., and then a late dinner, his only full meal. He used to run outdoors in Vienna, and in the city’s annual marathon. Now, owing to safety concerns, he uses treadmills in a converted garage, where he listens to classic jazz, like John Coltrane; podcasts on history and philosophy; or, under the influence of his only son, Benjamin, contemporary music, like Bad Bunny.

On weekends, Grossi insists on being in Vienna, where he is the assistant coach for Benjamin’s soccer team. Grossi is a self-described soccer fanatic, and still wears two V.I.P. wristbands from the 2022 World Cup, won by Argentina. At his son’s games, he is charged with preventing parents of both teams from fighting. “So he’s kind of the security guard,” Mariela Fogante, an arms expert who heads his office, told me. “I always laugh because, oh, my God, he’s so fragile.” Grossi insists he’s just coaching and keeping stats. The I.A.E.A. staff know to leave him alone during games.

Grossi also has seven daughters, now adults, who have influenced his agenda: one of the major changes at the I.A.E.A. under Grossi, as Holgate, the former U.S. Ambassador, noted, has been the creation of new opportunities for women. When he joined, the head of human resources told Grossi that twenty-eight per cent of the staff was female. “I said, ‘That’s not possible,’ ” he recalled. “ ‘What’s the problem?’ ” The staff is now fifty-two-per-cent female.

On November 26th, Argentina nominated Grossi to be the next U.N. Secretary-General. The Foreign Ministry cited “his deep knowledge of the multilateral system,” and a “proven performance” in dealing with conflict and international crises. The Secretary-General position—which is held for five years, with potential reëlection to a second term—rotates among regions. This time, it’s Latin America and the Caribbean’s turn. The election will take place next year. Grossi wants the job, even though the international body has been hamstrung by infighting and failed to use political leverage to prevent or end recent conflicts. “There is a widespread idea that the U.N. has become unwieldy and, having a thousand things to do, doesn’t do one well,” he told me. “It cannot be that out of all the crises that you see in the world, the U.N. is completely absent.”

In his “vision statement” after the nomination, Grossi wrote, “The world does not need more declarations. It needs a United Nations capable of responding to the real demands of our time.” Since the U.N. was founded, eighty years ago, it has spawned “overlapping mandates and fragmented functions,” he added. Long-sought reforms have become “stuck in self-serving bureaucratic loops.” The world still needs the U.N., he said. “But it must be a United Nations that works.”

In some ways, the I.A.E.A. is a microcosm of the U.N. Both tackle crises around the world but neither has the military means to enforce decisions. Both navigate among rival powers with opposite positions. The I.A.E.A. board of governors is akin to the fifteen-member U.N. Security Council; both include the five original nuclear powers. “There is no such thing as an intractable problem,” Grossi told me. “I don’t want to sound arrogant, but I know what the organization needs.” What works at the I.A.E.A. can be applied to the United Nations, he says. “I’m really convinced about that.”

Ernest Moniz, a former United States Secretary of Energy and nuclear negotiator, said that Grossi would bring a “very interesting new perspective” to the U.N. Scott Roecker, a former director for nuclear-threat reduction at the U.S. National Security Council, cited Grossi’s “boundless energy and optimism in solving some of the most difficult problems in the world.” Given the U.N.’s current state, “new blood with a spirit of coöperation and ability to have a positive impact is sorely needed,” Roecker said. But even long-standing colleagues are cynical about Grossi’s potential impact. “I don’t think God himself could restore the credibility of the U.N.,” Gary Samore, the former czar for weapons of mass destruction at the National Security Council during the Obama Administration, said. Divisions among the major powers, which all have vetoes, are too deep. The formal campaign for Secretary-General does not begin until next year, but Grossi told me that he would not challenge core U.N. pillars—peace and security, development, and humanitarian missions. “You don’t need to change the symphony,” he said. “The notes are the same. But the way you interpret the music can make a helluva difference in the way you play it.

Anyone with an upturned glass to the door of the conference room we met in last week could have been forgiven for thinking we were rehearsing a scene from Macbeth.

“Bubble.” “Bubble.” “No.” “Not yet.” “No trouble.”

In fact, I had asked the panel of five experts gathered for FT Money’s annual investment roundtable the Big One: whether the valuations of AI companies are irrationally overinflated — and what were the chances it would all go pop next year?

Before a twinkling Christmas tree, with the dome of St Paul’s framed in the picture window behind it, we discussed what retail investors should look out for in 2026.

Joining the discussion were FT columnists Stuart Kirk and Katie Martin; Simon Edelsten, a fund manager at Goshawk Asset Management; Niamh Brodie-Machura, chief investment officer of the equities team at Fidelity International; and Iain Stealey, chief investment officer of the global fixed income, currency and commodities team at JPMorgan Asset Management.

Perhaps unsurprisingly, the room was divided along professional lines — with the two journalists happy to use the B-word, and those responsible for running clients’ money more optimistic about the state of the market.

“There are two things to know about bubbles, historically,” said Edelsten. “The first one is that bubbles are notorious because everyone is a buyer at the top.” The current level of concern and dissent was absent, he argued, from the tulip mania, for example, or the crash of 1929.

“But the more important, less cynical point, is the valuation point,” he said. “I think it’s very different from 2000, when basically everything was nuts: there was no valuation basis for a huge number of very large companies, across Wall Street and Europe.” That is not true of the best AI stocks today.

Nvidia has a price-to-forward-earnings ratio of about 25; Amazon and Microsoft are in the high 20s — not cheap, but nowhere near as wild as some telecoms companies at the height of the dotcom boom.

“These are extremely strong companies from a fundamental perspective,” said Brodie-Machura. “The demand is across all parts of the supply chain, and it is in excess of what can be supplied at the moment, which means you have volume growth and pricing power at the same time.”

Stealey said that, from a bondholder’s point of view, he was “pretty comfortable” with the state of the market — though he made a distinction between some of the hyperscalers with huge cash piles and companies such as Oracle, which have much higher debt levels.

“And I know we had this indigestion over the past six weeks or so, where we’ve had big issuance from Meta [and Alphabet], and it did cause spreads to widen a little bit — but, to us, that’s a buying opportunity,” he said. “These are very highly rated [bonds], and there is liquidity sloshing around in the system, and people are desperate for yield at the moment.”

Next year, Fidelity analysts predict 28 per cent growth in tech sector earnings and for growth to broaden out.

Nobody wants to be the naive investor that got in at the top, said Brodie-Machura. “But then nobody wants to be the cynic sitting on the side . . . at a point in time where many participants think this may be the biggest inflection point for human productivity since the industrial revolution.”

“The stakes are really high,” she added. “If it doesn’t work, the risk of capital loss is real. If it does work, this [growth] can absolutely continue.”

But it’s not an argument that landed with Kirk — who recently liquidated his equity fund holdings, and is now 100 per cent in cash. “Having worked through four or five bubbles, where people told me ‘it’s different this time’, it never is,” he said. “So I discount the upside for AI: it’s just another ‘this time it’s different’ argument.”

Martin said she didn't doubt AI was a potentially transformative technology, but there was clearly “a very thick layer of froth” that needs to come off the market — and some of that would dissipate in 2026.

“Look, I believe that what’s happening with Nvidia is real. I don’t think this is a Pets.com phenomenon,” she said, referring to a poster child of dotcom-era hype. “But you look at some of the things going on in markets and you have to think: if this isn’t bubbly behaviour, then what is it?”

She cited the moment in October when Nvidia chief executive Jensen Huang dined at a fried chicken restaurant in Seoul, causing the share price of one of South Korea’s largest fried-chicken chains to jump 20 per cent (though not the one Huang actually visited, which is privately owned). Even the share price of a local chicken supplier jumped 30 per cent.

“And you look at this weird, circular-financing, cross-shareholding thing that’s going on at the top of the market, and you think: what is this if it’s not behaviour that tells you there’s excess here?” Martin said. “And I don’t have another word for it other than ‘bubble’.”

While the room was divided on the bubble question, there was broad agreement that inflation was a major risk for 2026.

Stealey said he still believed the most likely outcome for next year was that inflation in major western economies would stay roughly where it is or trend downwards, allowing central banks to ease rates, which would benefit all financial markets. But interest rate expectations have changed significantly over the course of this year.

“The big risk now is that we see a little bit of what is happening in Australia occur globally,” said Stealey. The Reserve Bank only started its rate-cutting cycle back in February. But with inflation rebounding, governor Michele Bullock has more or less ruled out any more cuts, and warned that hikes may be needed in 2026.

If borrowing becomes more expensive in the US, and this triggers a market sell-off, AI could follow the US housing bubble of the early 2000s, the dotcom bubble of the late 1990s and Japanese asset price bubble of the late 1980s, which all popped when central bankers started raising rates.

While this risk may feel remote, given that the Fed chose to cut rates this week to their lowest level in three years, Edelsten believes there is already more inflation in the system than is being modelled: “I get the feeling that the cost of living is going up much faster than the figures say.”

“Inflation has been tracking a lot higher and is becoming a political issue in America — I was surprised by how little it was mentioned in our UK Budget, because it’s high here too. And this is all against a background where the oil price is weak,” he said.

The ability — and political will — for many western economies to deal with rising prices has become a concern, with the US, UK and Japan all criticised for showing signs of fiscal dominance this year, where monetary policy is determined by budgetary concerns.

At the same time, Kevin Hassett has emerged in recent weeks as the frontrunner to become the next Fed chair in May. If appointed, many expect him to be obedient to President Trump. Joseph Wang of Monetary Macro even described Hassett as “the most political choice that you can get next to appointing Don Junior”.

“If [Fed] independence gets called into question, right around the time of the midterms, then that might cause a bit of an upset in all financial markets,” said Stealey. But he questioned how likely this would be, pointing to the fact that, even with Hassett the favourite in the betting markets, the 10-year US Treasury yield is still around 4.1 per cent. “At the moment, I think the market is still of the mindset that this is going to be a reasonably independent Federal Reserve,” he said.

If there is a problem with inflation next year, what would be the first signs that something was wrong? “If you started to see a re-acceleration in services [inflation], that would be a concern,” said Stealey. “How would that happen? It would be wages.”

At the moment, concerns about the US labour market are that it’s slowing. “But if you started to see a little more stability and wages start to pick up — and you combine that with the fiscal support that is coming from the One Big Beautiful Bill Act — that’s the concern,” he said.

Investors should have their eyes trained on the second quarter of 2026, since this is when US consumers could be in line for an income tax refund season, thanks to the OBBBA, which could boost demand and put yet more inflationary pressure into the system.

There are other risks that could take a pin to the AI balloon. One is that, for all the technology’s razzle-dazzle, investors will demand to see more evidence of its useful application in commercial settings.

“The spend is definitely happening, the capital deployed is going up,” said Brodie-Machura. “Now we’re looking for the ROI to come through . . . If it falters, or there are risks that it falters, or if the lead players change, [expect] a lot of volatility in the stock market.”

The other big risk, she said, is competition from Chinese AI. In January, DeepSeek stunned Silicon Valley rivals with the release of an AI model that worked at a fraction of the cost of those produced in the US, prompting a major tech sell-off. Another breakthrough like that next year could seriously undermine US stock valuations.

There are also more general risks to markets posed by the possible unwinding of the yen carry trade, due to economic conditions in Japan, and then there are the black swan risks — which the panel thought could come from non-bank lending or crypto or some other overlooked or poorly understood corner of the financial markets.

“It’s usually a product we’ve never heard of before that blows up,” said Kirk. “And it usually has something to do with debt.”

The room was divided on how investors can protect themselves from a downturn next year, should one occur.

The first thing, said Martin, is that the sheer size of the US tech companies and the concentration in US and global markets, means that were the AI bubble to pop “the blast radius would be enormous”.

But the secondary problem, she said, is how asset managers have told her that their hedging strategies involve moving away from AI stocks into Asian tech, energy infrastructure or copper, which is required for the data centre build-out. “And I’m like: ‘am I going mad here? This is the same thing!’” Martin said. “You’re hedging out into a different part of the same value chain of the same thing.”

“And so I worry about dominoes,” she added. “And I worry about how deep the damage could be.”

But both Edelsten and Brodie-Machura were keen to point out that diversification strategies have worked in 2025, citing big gains made in the FTSE 100, European defence and Japan.

“In the past year, you would have made as much money — in sterling — investing in the UK, because the banks have gone up, as you would have made investing in the S&P 500,” said Edelsten. “You would have made as much money investing in Japan, despite the yen going down, because the market recovered a lot — and none of that was AI.”

For those looking for diversification, there was much discussion about the excitement in so-called quality stocks, which offer a high return on equity, stable earnings growth and low debt levels.

Typically, the quality stocks that people concentrate on are consumer staples, such as Procter & Gamble or Nestlé — but they need to be cheap, said Edelsten. That’s crucial. “I’m afraid a lot of people have said: ‘oh, for cautious investors, just come and buy quality’ and [they] miss out the ‘cheap’ bit.”

Unfortunately, he added, a lot of these stocks became very expensive during the pandemic, because they were considered safe. “So everyone piled into them, but they don’t grow very fast.”

Edelsten said he was more interested in healthcare stocks. He mentioned United Healthcare — whose former chief executive was shot in Manhattan a year ago — and Johnson & Johnson. The company has had its problems, he said, but has a good pharmaceutical division and is well placed to develop automation inside healthcare, which could boost productivity.

At 18 times earnings, he said, it’s not very cheap, but it shows that not everything in the US is expensive. “The American equity market is massive. And although it’s dominated by a small number of companies on some really fancy ratings, as soon as you go down [the list], if you do work, you’ll find plenty of stocks that you’ll think are reasonable.”

But playing defensive is not easy, he added. “This is one of the reasons that tech investing has carried on. When you look at the alternatives . . . they have had an awful lot of bombs going off. And so switch out of a tech stock, where you hope to make a lot of money, into something where you don’t expect to make an awful lot of money, and then it has a slight wobble on its profits and you’re 20 per cent down and you think, ‘why did I bother to do that, then?’”

Kirk said he thought government bonds were interesting again. “Whereas maybe nine, 12 months ago, we all thought that no one’s ever buying govvies ever again, that all governments are bust and the returns are awful, but I think if there’s a sell-off next year, government bonds are going up.”

Both Stealey and Brodie-Machura liked the idea of some alternative assets, such as gold. But Martin disagreed, at the risk of sounding like the ghost at the feast.

“Weirdly,” she said, “I’d point to gold as one of my pieces of evidence that markets are bubbly.”

She said the Royal Mint had told her it had introduced a queueing system to its website, because it was becoming so overloaded with people wanting to buy gold “while they’re sitting in front of the telly watching Strictly on a Saturday night”.

“Now ,” she added, “these are not people who are sitting around looking at central bank reserve accumulation and thinking about fiat currency debasement or global debt sustainability. They’re hopping on a momentum-driven bubble.”

‘ChatGPT for Doctors’ Startup Doubles Valuation to $12 Billion as Revenue Surges

The Takeaway

If the deal closes, the Miami-based company will become one of eight AI application developers to both reach a valuation higher than $10 billion and generate annualized revenue of more than $100 million, according to The Information’s Generative AI Database. It makes money by selling advertising space on its chatbot to pharmaceutical companies, similar to the way Google sells ads on its search engine.

That business is growing quickly. OpenEvidence is generating around $150 million in annualized advertising revenue, which implies it’s currently generating more than $12 million in revenue per month, the person involved in the deal said. That’s up three times from the advertising revenue it was generating in August.

OpenEvidence has told potential investors that it’s only selling one-tenth of its ad inventory, suggesting it could generate more than $1 billion in annualized revenue if it sold the rest.

The company’s gross profit margin is currently higher than 90%, putting it above many other AI startups in that regard. Gross margins measure how much companies earn from their sales after subtracting the direct costs of serving its revenue-generating products, known as cost of revenue. OpenEvidence’s costs of revenue include what it pays to run servers, including AI models that power the product, and money it spends to license content from medical journals, the person said.

The round comes as consumers increasingly use chatbots to look up medical information. That’s driven chatbot-makers to put resources towards healthcare-related features. OpenAI, for instance, hired longtime Facebook executive Ashley Alexander in August to develop health-related products.

OpenEvidence targets physicians, not consumers, who use its chatbot to answer medical questions or analyze peer-reviewed studies. The company has said its chatbot is more accurate than general chatbots such as ChatGPT because it was developed using information from medical journals whose content it licensed, such as the New England Journal of Medicine. It’s not clear whether OpenAI has similarly struck licensing agreements with medical journals.

OpenEvidence can also use user feedback to improve the ranking of its search results, the person involved in the deal said. It can also reach out to users who are physicians and experts in niche fields to help it train its AI to respond more accurately to niche questions, they said.

The company’s product uses open-source AI models to search through medical journals and return useful information and citations, as well as models it trains for tasks like retrieval and search ranking. It also uses AI from Google, OpenAI and other providers to summarize its chatbot’s responses, the person said.

OpenEvidence’s chatbot seems to be catching on with physicians. According to an October survey of 1,000 U.S.-based physicians by OffCall, a company that allows physicians to compare their salaries, around 45% of physicians use OpenEvidence, compared to 16% for ChatGPT and 5% for medical scribe startup Abridge, the next most popular healthcare-specific AI product. (There are about 1 million physicians in the U.S. total, according to the Federation of State Medical Boards.)

The company’s chatbot answers around 20 million questions from U.S.-based physicians per month, according to the person involved in the round.

OpenEvidence previously raised more than $500 million in funding from investors including GV, Sequoia Capital and Kleiner Perkins. Its CEO and cofounder, Daniel Nadler, previously sold an AI startup to financial research firm S&P Global for hundreds of millions of dollars in 2018.

Startup investors this week demonstrated continued willingness to write big checks for promising companies in sought-after areas. Leading the pack was Saviynt, an AI identity security platform that picked up a fresh $700 million.

The next three largest rounds — Unconventional AI, Fervo Energy and Boom Supersonic — all shared some focus on energy. The rest of the list included companies in AI, biotech and space tech.

1. Saviynt, $700M, AI identity security: Saviynt, a provider of AI-optimized identity security tools, announced that it secured $700 million in Series B financing. KKR led the round, which set a $3 billion valuation for the 15-year-old, El Segundo, California-based company.

2. Unconventional AI, $475M, AI energy efficiency: Unconventional AI, a startup designing a computer to optimize energy efficiency for AI, raised $475 million in seed funding. Andreessen Horowitz and Lightspeed Venture Partners led the financing for the San Francisco Bay Area company.

3. Fervo Energy, $462M, geothermal energy: Houston-based Fervo Energy, a developer and operator of geothermal energy projects with a focus on technologies to scale this power source, picked up $462 million in Series E funding led by B Capital. The funding will go toward a geothermal project in Western Utah as well as other projects in its pipeline.

4. Boom Supersonic, $300M, fast airplanes, turbines: Boom Supersonic, a Denver company working to build what it says will be the world’s fastest airliner, closed on $300 million led by Darsana Capital Partners. In addition to its planes, Boom is also attracting investor interest for its Superpower natural gas turbine that can also have applications in delivering energy to AI data centers.

5. K2 Space, $250M, space tech: Torrance, California-based K2 Space, developer of a platform for building large, high-power satellites, landed $250 million in Series C funding. Redpoint led the round, which set a $3 billion valuation for the company, which was founded in 2022.

6. Harness, $240M, software development: Harness, a developer of tools to automate and simplify software delivery processes, raised $200 million in Series E funding and $40 million in a planned tender to provide liquidity to long-term employees. Goldman Sachs led the Series E, which set a $5.5 billion valuation for the 8-year-old company.

7. Impulse Dynamics, $158M, medical devices: Marlton, New Jersey-based Impulse Dynamics, a medical device company focused on patients with heart failure, secured over $158 million in a funding round led by Sands Capital Ventures and Braidwell. The financing follows an announcement from the Centers for Medicare and Medicaid Services extending coverage for its device.

8. Fal, $140M, generative AI: Fal, developer of a real-time generative AI platform for video, images, 3D and audio, picked up $140 million in a Series D financing Led by Sequoia Capital. The round was the third this year for San Francisco-based Fal, which closed a Series C in July and a Series B in February.

9. Sanegene Bio, $110M, biotech: Sanegene Bio, a biotech startup focused on developing RNAi-based therapeutics, said it raised over $110 million in Series B funding from a long list of venture investors. Founded in 2021, the startup is headquartered in Boston with significant operations in China.

10. BlossomHill Therapeutics, $84M, biotech: San Diego-based BlossomHill Therapeutics, a developer of medicines to treat cancer, pulled in $84 million in Series B extension funding. Janus Henderson Investors, Brahma Capital and BioTrack Capital led the financing for the 5-year-old company.

Donald Trump’s Ukraine envoy, Jared Kushner, has approached Marc Rowan’s Apollo Global Management and Henry Kravis’s KKR to assist with post-war reconstruction in Ukraine, The Post has learned.

Sources familiar with the matter said Kushner, President Trump’s son-in-law who has no formal role in the administration, was in New York City on Wednesday for exploratory discussions with the top asset managers in Manhattan’s swanky Hudson Yards development.

Insiders said the 44-year-old held meetings at the offices of both financial giants on how to rebuild the war-torn nation.

The talks, which are still at a very early stage and may not come to fruition, came on the same day that he took part in a call with Larry Fink of BlackRock, the world’s biggest money manager, Special Envoy Steve Witkoff, and Treasury Secretary Scott Bessent, as well as Ukrainian President Volodymyr Zelensky.

The names of the VIP participants were released in a statement by Ukraine’s presidential office that evening.

Reps for BlackRock, KKR, Affinity Partners, the Treasury Department, and President Zelensky did not return The Post’s requests for comment.

A spokesman for Apollo declined to comment.

BlackRock CEO Larry Fink halted his firm’s own multi-billion-dollar Ukraine Development Fund in July amid uncertainty over how President Trump would try to end the war.

But he is set to revive the idea, which would also include involvement from the World Bank, that aimed to drum up some $400 billion in fresh investment.

“This is not a huge money maker for whoever does it. They are not going to charge incentive fees. It’s a public service,” one person briefed on the situation added.

“But you are going to ingratiate yourself with the president and Mr Kushner,” the source added. “They are going to f–k the Europeans.”

Indeed, the move could raise fears amongst Ukraine’s allies in Europe that they will be cut out of any lucrative contracts to rebuild the war-torn country.

The Trump administration, via the US International Development Finance Corporation (DFC), in September pledged $75 million in equity intended to jumpstart the fund’s investment in critical minerals, hydrocarbons, defense, and related infrastructure in Ukraine.

That investment is set to be matched by Kyiv to bring it to an initial combined total of $150 million, but DFC stressed that much of the work would be done by “aligned private sector partners.”

Ukrainian government officials, including Prime Minister Yulia Svyrydenko, are set to hold what one source described as “marathon talks” about the US-backed peace proposal this weekend.

They will also focus on how to rebuild the nation that has been subjected to a brutal four-year attempt to wipe Ukraine off the map by Russian invaders.

Kushner, who is not formally part of the administration, has emerged as a central figure in the process of negotiating a peace settlement between Ukraine and Russia.

Alongside Trump envoy Steve Witkoff, he attended five hours of Kremlin talks last week with Russian President Vladimir Putin and then later discussions with a Ukrainian delegation in Florida.

Kushner, married to the president’s daughter, Ivanka Trump, previously played a major role in negotiating a ceasefire agreement in Gaza between Israel and Hamas and has also put forward proposals for reconstruction in Gaza.

A key point in plans for the reconstruction of Ukraine has been the establishment of an investment fund for sectors including rare metals – sought by the United States, with Svyrydenko playing a key role in setting it up.

Kyiv sent its own revised draft of the now 20-point plan to the White House on Wednesday, according to German Chancellor Friedrich Merz.

It reportedly includes the creation of a demilitarized zone along the line of contact, security guarantees for Ukraine in line with NATO’s Article 5, and membership for Ukraine in the European Union by 2027.

It removes language barring Ukraine from ever joining NATO, but does not mention if the country will ever join the military alliance — a key Kremlin talking point.

Coca-Cola holds last-ditch talks in bid to salvage Costa Coffee sale

Owner and preferred bidder TDR Capital are trying to rescue deal after failing to agree on price

Coca-Cola’s proposed sale of Costa Coffee is at risk of collapse and the soft drinks giant is holding last-ditch talks with private equity firm TDR Capital this weekend in an attempt to salvage it.

Asda owner TDR was selected as Coke’s preferred bidder for Costa earlier this week after a board meeting in New York, according to people familiar with the matter. However, talks with Coke and its advisers at Lazard have stumbled over the price, according to one of the people.

Coke will make a decision next week on whether to shelve the sale process altogether, the people said.

The deal on the table includes Coke retaining a minority stake in Costa, one of the people said, adding that the size of the stake could be adjusted in Coke’s favour in order to get a deal done.

Coke wanted to get roughly £2bn for Costa, the Financial Times previously reported.

The drinks group paid £3.9bn to acquire Costa from Premier Inn owner Whitbread in 2018. But the business has since struggled against competition from independent operators and mass-market rivals such as Greggs.

Rising costs of everything from coffee beans to staff wages have pushed the business’s finances into the red. In 2023, Costa reported an annual loss of £13.8mn on revenues of £1.2bn, according to the most recently available set of accounts at Companies House.

TDR, which co-owns petrol forecourt business EG Group, is seeking to buy Costa’s UK and international business, excluding its operations in China, one of the people said.

Other bidders for Costa included Bain Capital’s special situations division, which owns Gail’s and PizzaExpress. Centurium Capital, the private equity owner of China’s Luckin Coffee chain, has also been involved in the auction, according to people familiar with the process.

Private capital firms Apollo and KKR have both dropped out of the process in the past few months.

Coke announced this week that chief executive James Quincey will be replaced by chief operating officer Henrique Braun in March. Quincey, who told analysts in July that Costa had “not delivered”, will become executive chair.

Coke did not respond to a request for comment. TDR and Lazard declined to comment.