TheChek News Flow

China’s senior-most general is accused of leaking information about the country’s nuclear-weapons program to the U.S. and accepting bribes for official acts, including the promotion of an officer to defense minister, said people familiar with a high-level briefing on the allegations.

The briefing—attended on Saturday morning by some of the military’s highest-ranking officers—came just before China’s Ministry of National Defense made the bombshell announcement of an investigation into Gen. Zhang Youxia, once considered Chinese leader Xi Jinping’s most-trusted military ally. That statement gave few details beyond a probe of severe violations of party discipline and state laws.

But the people familiar with the briefing—which hasn’t been reported until now—said Zhang is under investigation for allegedly forming political cliques, a phrase describing efforts to build networks of influence that undermine party unity, and abusing his authority within the Communist Party’s top military decision-making body, known as the Central Military Commission.

Authorities are also scrutinizing his oversight of a powerful agency responsible for the research, development and procurement of military hardware. Those familiar with the briefing said Zhang was alleged to have accepted huge sums of money in exchange for official promotions in this big-budget procurement system.

The most shocking allegation disclosed during the closed-door briefing, the people said, was that Zhang had leaked core technical data on China’s nuclear weapons to the U.S.

Some evidence against Zhang came from Gu Jun, the former general manager of the China National Nuclear Corp., a state-owned company that oversees all aspects of China’s civilian and military nuclear programs, said the people familiar with the briefing. Beijing announced an investigation into Gu last Monday for suspected severe violations of party discipline and state laws.

During Saturday’s briefing, authorities revealed that the probe into Gu has linked Zhang to a security breach within China’s nuclear sector, the people said. Details on the breach weren’t disclosed during the briefing, the people said.

Zhang, 75 years old, and Gu couldn’t be reached for comment. In a statement to The Wall Street Journal, Liu Pengyu, a spokesman at the Chinese Embassy in Washington, said the party’s decision to investigate Zhang underscores that the leadership maintains “a full-coverage, zero-tolerance approach to combating corruption.”

Some analysts say Xi’s latest crackdown on corruption and disloyalty in the armed forces marks the most aggressive dismantling of China’s military leadership since the Mao Zedong era.

Like Xi, Zhang, a member of the party’s elite Politburo, is one of China’s “princelings,” as the descendants of revolutionary elders and high-ranking party officials are known. Zhang’s father fought alongside Xi’s father during the Chinese civil war that led to Mao’s Communist forces seizing power in 1949, and both men later rose to senior roles.

“This move is unprecedented in the history of the Chinese military and represents the total annihilation of the high command,” said Christopher Johnson, head of China Strategies Group, a political-risk consulting firm.

Saturday’s internal briefing also linked Zhang’s downfall to his promotion of former Defense Minister Li Shangfu. Zhang allegedly helped elevate Li in exchange for large bribes, said the people.

Li’s own downfall began in 2023, when he disappeared from public view and was later removed as defense minister. The party expelled him the following year for corruption. Li couldn’t be reached for comment.

In a sign of the depth of the current probe, Xi has commissioned a task force to conduct a deep-dive investigation into Zhang’s tenure as commander of the Shenyang Military Region, which spanned five years from 2007 to 2012, said some of the people familiar with the briefing. The team has now arrived in the northeastern city of Shenyang, they said, notably choosing to stay in local hotels rather than military bases, where Zhang would have a network of support.

Authorities have already seized mobile devices from officers who rose through the ranks with Zhang and Gen. Liu Zhenli, chief of the Joint Staff Department, whose investigation was also announced Saturday, some of the people said, as thousands of officers with ties to the men become potential targets. Liu, 61, couldn’t be reached for comment.

The downfall of Zhang and Liu expands Xi’s yearslong push to clear out officers deemed corrupt and politically unreliable.

The purging of even one of Xi’s personal friends, Zhang, shows that there now are no limits to Xi’s antigraft zeal, said Johnson, the China Strategies Group head who served at the Central Intelligence Agency for two decades, including as a top China analyst.

“Xi sought to avoid a wholesale cashiering of the top brass in the early years of the anticorruption campaign” by not targeting active-duty senior generals and key parts of the military such as the strategic-missile force, Johnson said. “He later realized that was impossible, and this move is the denouement of that process.”

Analysts say the opacity of China’s political system makes it difficult to ascertain Xi’s precise motivations for liquidating a longtime ally in the Chinese military, known as the People’s Liberation Army, or PLA. Internal explanations provided to the party elite—such as those in Saturday’s briefing—don’t always reflect the complete or true motivation behind Xi’s decisions.

Even so, a Saturday editorial by the PLA Daily, the military’s flagship newspaper, indicated the importance of political factors in the case against Zhang, whom the newspaper accused of having “severely trampled on and undermined” the institutional basis of the CMC chairman’s authority. This suggests that “Zhang had too much power outside of Xi himself,” said Lyle Morris, senior fellow for foreign policy and national security at the Asia Society Policy Institute.

The party has emphasized that Xi, as CMC chairman, has ultimate authority over the military and represents the party’s command of the armed forces, Morris said. “So, highlighting such a violation suggests that Zhang was out of step with Xi’s chain of command.”

Regardless of the reasons, Xi’s decision to cast out Zhang shows he “is confident in his consolidation of power over the military,” Morris said. “It’s not a sign of weakness, but of strength, for Xi.”

By decapitating the command structure, said some of the people familiar with the briefing, Xi is signaling that rampant corruption, entrenched patronage networks and the compromise of state secrets are existential threats to his goal of gaining control over Taiwan, the democratically self-ruled island that Beijing claims as its territory.

However, some analysts say, the resulting hollowing out of the senior ranks is likely to affect combat effectiveness and could lower the immediate risk of a cross-strait invasion as Xi pivots to a more calculated strategy. Beijing appears to be looking first to make a deal with President Trump on Taiwan as the two superpowers prepare for high-stakes negotiations on trade and security matters this year. Any wavering by Trump on Taiwan’s security could shake Taipei’s confidence in American resolve.

Since the summer of 2023, the party has unseated top officers in China’s army, air force, navy, strategic-missile force and paramilitary police, as well as major theater commands—including the one focused on Taiwan. More than 50 senior military officers and defense-industry executives have been placed under investigation or removed from office in the past 2½ years, according to official disclosures reviewed by the Journal.

The Central Military Commission had six professional military members when its current term started in 2022. Now it has just one active uniformed officer, Gen. Zhang Shengmin, who was promoted to vice chairman only in October after the purge of another general who held that role. Unlike Zhang Youxia and Liu Zhenli, who are combat veterans, Zhang Shengmin has served as a political officer and discipline inspector for the bulk of his career, responsible for enforcing loyalty and boosting morale.

“Given the size and complexity of overseeing any large and sophisticated military organization, this vacuum at the top is untenable,” said M. Taylor Fravel, director of the Security Studies Program at the Massachusetts Institute of Technology. He added that it was “bound to have an impact on the PLA’s current readiness to undertake major, complex military operations in the short to medium term.”

The Takeaway

It’s a similar story with large language model developers Zhipu and MiniMax, which are trading at 178 and 252 times last year’s revenue, respectively. That’s much higher than the 57 times multiple of 2025 revenue implied by the $750 billion valuation at which the still-private OpenAI is currently reported to be raising money.

For fast-growing companies, investors typically look at projected future revenue when calculating multiples, rather than multiples of the last 12 months’ revenue. Estimates for future revenue aren’t available for most of the Chinese firms, which means forward multiples aren’t available. But in the one case where such data is available—Zhipu—the picture is the same. Zhipu is trading at 62 times forward revenue, according to S&P Global Market Intelligence. That’s more than double OpenAI’s estimated 2026 multiple of 25 times, assuming it raises money at $750 billion, and if we use its projection of $30 billion for 2026 revenue.

The valuations in China aren’t as frothy as they appear, given the chances that China’s AI industry will take off soon. That would underpin significant revenue growth in the not too distant future, according to Duane Kuang, founding managing partner of Qiming Venture Partners.

“U.S. companies and their revenues provided a path for Chinese companies’ revenue growth. Based on the size of the market size and technology development in China, Chinese companies’ revenues will grow significantly and the multiples will come down and edge close to their U.S. peers. The current valuations reflect that confidence,” said Kuang, an early investor in Zhipu and Biren.

Edison Lee, an analyst at Jefferies, sees plenty of room for Chinese AI valuations to rise. He estimates that the combined market capitalization of China’s AI-related companies, including software firms and chip designers and manufacturers, is “now only 8% of its U.S. peers, but Chinese cloud providers’ capital spending is 18% of that of the U.S.,” he noted. That implies those Chinese businesses have a lot of room to catch up to the U.S.

China’s AI companies were mostly founded over the last five years, when U.S.-China relations started to sour, and Washington started to contain Beijing’s rise by restricting technology exports to the country.

The U.S. restrictions served as a wake-up call for Chinese leaders, who mobilized resources and formulated policies to nurture homegrown champions so the country can one day wean itself off American technologies.

It’s a work in progress. AI developers say China-made chips still lag far behind Nvidia’s, and LLMs developed by the likes of OpenAI, Anthropic and Google still top various industry benchmarks. But Chinese investors are optimistic about the local industry’s prospects.

Those investors demonstrated that optimism with their reaction to Zhipu’s release last Wednesday of an open-source AI image model trained entirely on chips from Huawei Technologies, the first time a Chinese state-of-the-art AI model had been trained on domestic chips. Zhipu’s share price soared 38% in the three days after the model’s release.

Still, challenges remain for the newly listed chipmakers. “As far as the newly listed GPU players are concerned, the biggest challenge is securing enough manufacturing capacity in China to boost their output,” Lee said. “We expect China’s GPU market to be very competitive in the next few years, even though demand will also be strong, driven by high growth in AI applications.”

Since the launch of Apple Intelligence over a year ago, Apple has added very few new AI features to its apps. Some recent additions are Apple Music’s AutoMix and Apple Watch’s Workout Buddy. However, that is set to change this year.

According to Mark Gurman in today’s Power On newsletter, Apple is planning to integrate the upcoming Gemini-backed version of Siri “deeply across its core apps”.

Previously, Apple’s plans had been much more ambitious. For a while now, Apple has been working on a “World Knowledge Answers” project, which aimed to compete with the likes of ChatGPT and Perplexity. However, this project has been scaled back, as Apple works to instead replace existing technologies with Google’s Gemini.

The World Knowledge Answers project is not the only project to be scaled back. According to Gurman, Apple had been working on a AI overhaul of Safari:

These plans are also now paused, though Gurman says there is a chance for the work to resume prior to WWDC26. Apple has also “returned to the drawing board” for AI Health features, Gurman reports.

According to today’s Power On, Apple had “envisioned standalone chatbot-style experiences embedded in apps such as Safari, TV, Health, Music and Podcasts”. But now, following the departure of SVP of Machine Learning and AI Strategy John Giannandrea, Apple is planning a more cohesive experience. Rather than having separate chatbots across multiple apps, they will instead “deeply” integrate the upcoming Gemini-based Siri across their core apps.

Unfortunately, we don’t yet know what this will look like, but we may not have to wait long to find out. Gurman also reports that the new Siri will be announced as soon as next month. In the past, Apple has also shown off how Siri might integrate with apps through App Intents.

Though it has certainly been a long wait, 2026 is shaping up to be an exciting year for Apple Intelligence. Apple is expected to finally launch an overhauled Siri, first shown off at WWDC24.

We’re in a unique moment for AI companies building their own foundation model.

First, there is a whole generation of industry veterans who made their name at major tech companies and are now going solo. You also have legendary researchers with immense experience but ambiguous commercial aspirations. There’s a clear chance that at least some of these new labs will become OpenAI-sized behemoths, but there’s also room for them to putter around doing interesting research without worrying too much about commercialization.

The end result? It’s getting hard to tell who is actually trying to make money.

To make things simpler, I’m proposing a kind of sliding scale for any company making a foundation model. It’s a five-level scale where it doesn’t matter if you’re actually making money – only if you’re trying to. The idea here is to measure ambition, not success.

Think of it in these terms:

The big names are all at Level 5: OpenAI, Anthropic, Gemini, and so on. The scale gets more interesting with the new generation of labs launching now, with big dreams but ambitions that can be harder to read.

Crucially, the people involved in these labs can generally choose whatever level they want. There’s so much money in AI right now that no one is going to interrogate them for a business plan. Even if the lab is just a research project, investors will count themselves happy to be involved. If you aren’t particularly motivated to become a billionaire, you might well live a happier life at Level 2 than at Level 5.

The problems arise because it isn’t always clear where an AI lab lands on the scale — and a lot of the AI industry’s current drama comes from that confusion. Much of the anxiety over OpenAI’s conversion from a non-profit came because the lab spent years at Level 1, then jumped to Level 5 almost overnight. On the other side, you might argue that Meta’s early AI research was firmly at Level 2, when what the company really wanted was Level 4.

With that in mind, here’s a quick rundown of four of the biggest contemporary AI labs, and how they measure up on the scale.

Humans& was the big AI news this week, and part of the inspiration for coming up with this whole scale. The founders have a compelling pitch for the next generation of AI models, with scaling laws giving way to an emphasis on communication and coordination tools.

But for all the glowing press, Humans& has been coy about how that would translate into actual monetizable products. It seems it does want to build products; the team just won’t commit to anything specific. The most they’ve said is that they will be building some kind of AI workplace tool, replacing products like Slack, Jira and Google Docs but also redefining how these other tools work at a fundamental level. Workplace software for a post-software workplace!

It’s my job to know what this stuff means, and I’m still pretty confused about that last part. But it is just specific enough that I think we can put them at Level 3.

This is a very hard one to rate! Generally, if you have a former CTO and project lead for ChatGPT raising a $2 billion seed round, you have to assume there is a pretty specific roadmap. Mira Murati does not strike me as someone who jumps in without a plan, so coming into 2026, I would have felt good putting TML at Level 4.

But then the last two weeks happened. The departure of CTO and co-founder Barret Zoph has gotten most of the headlines, due in part to the special circumstances involved. But at least five other employees left with Zoph, many citing concerns about the direction of the company. Just one year in, nearly half the executives on TML’s founding team are no longer working there. One way to read events is that they thought they had a solid plan to become a world-class AI lab, only to find the plan wasn’t as solid as they thought. Or in terms of the scale, they wanted a Level 4 lab but realized they were at Level 2 or 3.

There still isn’t quite enough evidence to justify a downgrade, but it’s getting close.

Fei-Fei Li is one of the most respected names in AI research, best known for establishing the ImageNet challenge that kickstarted contemporary deep learning techniques. She currently holds a Sequoia-endowed chair at Stanford, where she co-directs two different AI labs. I won’t bore you by going through all the different honors and academy positions, but it’s enough to say that if she wanted, she could spend the rest of her life just receiving awards and being told how great she is. Her book is pretty good too!

So in 2024, when Li announced she had raised $230 million for a spatial AI company called World Labs, you might think we were operating at Level 2 or lower.

But that was over a year ago, which is a long time in the AI world. Since then, World Labs has shipped both a full world-generating model and a commercialized product built on top of it. Over the same period, we’ve seen real signs of demand for world-modeling from both video game and special effects industries — and none of the major labs have built anything that can compete. The result looks an awful lot like a Level 4 company, perhaps soon to graduate to Level 5.

Founded by former OpenAI chief scientist Ilya Sutskever, Safe Superintelligence (or SSI) seems like a classic example of a Level 1 startup. Sutskever has gone to great lengths to keep SSI insulated from commercial pressures, to the point of turning down an attempted acquisition from Meta. There are no product cycles and, aside from the still-baking superintelligent foundation model, there doesn’t seem to be any product at all. With this pitch, he raised $3 billion! Sutskever has always been more interested in the science of AI than the business, and every indication is that this is a genuinely scientific project at heart.

That said, the AI world moves fast — and it would be foolish to count SSI out of the commercial realm entirely. On his recent Dwarkesh appearance, Sutskever gave two reasons why SSI might pivot, either “if timelines turned out to be long, which they might” or because “there is a lot of value in the best and most powerful AI being out there impacting the world.” In other words, if the research either goes very well or very badly, we might see SSI jump up a few levels in a hurry.

Plus, our columnists get on the Claude Code bandwagon, why Elon Musk wants a SpaceX IPO and Intel’s crash back to earth, in this edition of the Technology newsletter

Of all the fronts in the AI wars—funding, electricity, chips—the messiest is the fight for human capital.

Top AI researchers can earn hundreds of millions of dollars and jump between labs with ease, all while carrying around the knowledge that they may be writing humanity’s future. It’s a recipe for high-stakes drama.

Barret Zoph was one of six OpenAI employees who left to start Thinking Machines Lab a year ago. On Jan. 14, Thinking Machines Chief Executive Mira Murati—the person who replaced Sam Altman as OpenAI’s CEO during his brief ouster—posted on X that the company had “parted ways” with Zoph.

Within hours, tech reporters Keach Hagey and Meghan Bobrowsky learned that Zoph had been fired—and immediately hired by OpenAI. They found out that Zoph had had a romantic relationship with another Thinking Machines employee, and that Murati had described Zoph, in a message to colleagues, as having a series of performance and conduct issues while at the company. Zoph said Thinking Machines fired him “only after it learned I would be leaving the company” and never cited performance or conduct as a factor.

Then the reporters began to hear even more intriguing details about the backstory behind Zoph’s exit—details that included a surprise meeting, an attempted power grab and secret discussions with Altman and executives at OpenAI competitor Meta Platforms.

Keach and Meghan spent the days that followed nailing down those details. They ended up with a story that will have you wondering: With all this workplace drama, how do these companies ever get around to making world-changing AI?

NBA to share financial plans with potential backers of European basketball league

Private equity groups and football clubs invited to submit bids to own a team in the new competition

The NBA will next week give private equity firms and top European football clubs access to financial projections for its proposed European league, as the world’s top basketball competition pushes ahead with plans to expand internationally.

Blackstone, CVC Capital Partners, RedBird Partners and General Atlantic are among the groups that will be given access to a data room as soon as Monday, according to people familiar with the situation.

BlackRock, BC Partners and Oaktree Capital are also set to be given access to the documents as investors assess whether to back teams that would participate in the new league.

In US sports leagues, including the NBA, private equity firms and other institutional investors are barred from owning teams outright.

However, speaking last week in London, NBA commissioner Adam Silver said that the European league would take a different approach, noting that sovereign wealth funds and private equity firms already owned and operated teams in Europe.

“We are open to different kinds of investors that, as principal investors, aren’t currently allowed in the NBA and the US,” he said. “With an opportunity to begin a league from scratch, we’re taking a fresh look and we’re realists. We understand that there are different norms in different parts of the world in terms of sports team ownership.”

After initially receiving more than 300 expressions of interest, the NBA has narrowed down the list of potential financial backers for the European league to several dozen, according to people familiar with the matter.

Nonbinding team franchise bids are due by the end of March, with the NBA hoping to launch the competition in late 2027.

Team franchises are likely to be valued at between $500mn and $1bn, according to people briefed on the discussions.

Executives at several top European football clubs, including Real Madrid, FC Barcelona, Bayern Munich, Turkish side Fenerbahçe and Athens-based Panathinaikos, will also be given details of the NBA Europe project, including its governance structure and revenue forecasts. While known for their football teams, all are multisport clubs that already operate basketball teams.

The NBA plan threatens to reshape European basketball and poses a major threat to EuroLeague, the long standing regional competition. Barcelona recently renewed its licence to play in EuroLeague but said it would always seek to compete “in the best competitions available”.

Silver said he hoped the involvement of football teams would help bring to basketball the kind of “tribalism” and “religious fervour” typically associated with European sports fans.

The NBA plan threatens to reshape European basketball and poses a major threat to EuroLeague, the long-standing regional competition.

Barcelona recently renewed its licence to play in EuroLeague but said it would always seek to compete “in the best competitions available”.

Existing European basketball teams, including Lyon-based ASVEL, owned by French NBA legend Tony Parker, and Olimpia Milano have also been shortlisted as potential participants in the new league.

The NBA has been seeking to build interest in its proposed European expansion, meeting potential investors, broadcasters and sponsors in recent days after staging games in London and Berlin.

The association has been making the case for establishing its own European league based on basketball’s growing popularity outside the US, and the region’s strong record of producing top players. A European has been named the NBA’s Most Valuable Player five times in the past seven years.

More than 40 per cent of the country under winter storm warning with emergency orders in place

A severe winter storm ripping across the south and east of the US has caused mass power outages and the cancellation of thousands of flights.

More than 500,000 people across the country were left without power, according to tracking website Poweroutage.com. The hardest hit were the states of Texas, Mississippi, Tennessee and Louisiana.

More than 13,500 weekend flights were canceled across the US, according to flight-tracking site FlightAware, with as many delayed, leaving passengers stranded at airports across the US.

Some 40 per cent of the country was under a winter storm warning by Saturday evening.

President Donald Trump announced emergency declarations to deal with the impact of the storms in South Carolina and Virginia. “With the help of FEMA and our State partners, we will keep everyone safe, and make sure both States have the support they need,” he posted on Truth Social.

As Winter Storm Fern travels eastward, power companies were braced for snow, freezing rain and sleet to hit the grid system, with helicopters and line construction crews being dispatched to identify damage and carry out swift repairs.

Georgia Power urged its customers to prepare for two to three-day outages, while Dominion Energy in Virginia warned that icy conditions could knock out power in the region for days.

Governors in states including Virginia, North Carolina, South Carolina and Georgia declared a state of emergency. On Thursday Energy Secretary Chris Wright ordered grid operators to make more than 35 gigawatts of backup generation available to prevent blackouts.

Power outages cost the US $44bn last year, according to the Department of Energy’s National Laboratories.

Fern is drawing comparisons to Winter Storm Uri, which led to mass blackouts and loss of life across Texas in 2021. Temperatures are expected to drop to the low teens and single digits Fahrenheit, with wind-chill factors of minus 10 to zero Fahrenheit.

Ercot, the grid operator for Texas, has said it is ready for the storm and has sufficient generation to meet demand.

“The same risk of power outages due to lower natural gas production still applies, as natural gas wells can freeze in temperatures below freezing,” said Mark Callahan, director of Americas Natural Gas Pricing at S&P Global Energy. “However, conditions for this storm are not as bad, and work has been done since 2021 to winterise the grid for such events.”

While the state is better prepared due to grid modernisation efforts and the rise of home generators, “the buffers in the system are much more constrained” by rising power demand from data centres and electrification of household appliances, said Didi Caldwell, chief executive of Global Location Strategies.

PJM, the grid operator which serves more than 67mn people in the north-east and Midwest, said temperatures could reach single digits Fahrenheit across the region and dip below zero in the west.

“This is a formidable arctic cold front coming our way,” said Mike Bryson, senior vice-president of operations at PJM. “We will be relying on our generation fleet to perform as well as they did during last year’s record winter peak.”

Natural gas was trading at elevated levels, with prices around $28 at Henry Hub, Louisiana, $34 in Chicago and $60 to $100 in the Northeast and East Coast on Friday, noted analysts at S&P.

However, this is “much lower than the $1,000 prices we saw during Uri”, said Callahan.

Retailers Amazon and Door Dash said they would balance deliveries of consumer staples while prioritising the safety of their workers. At least 21 Walmarts across the country were closed.