TheChek News Flow

At Cannes Lions, marketers grappled with how to influence A.I. chatbots.

For a quarter-century, the economic engine of the digital world relied on a predictable transaction: Users searched, links appeared and traffic flowed.

That foundational contract was declared effectively dead this week at the Cannes Lions International Festival of Creativity in France.

As the global advertising elite gathered aboard superyachts and inside beachfront villas, the conversation shifted from how to win human attention to a much more urgent, existential question: How do you influence an A.I. model?

The concern carries major financial and systemic implications far beyond Madison Avenue: Fortune 500 corporations facing sudden commercial invisibility as chatbot answers bypass multibillion-dollar digital marketing funnels; publishers and creators watching A.I. models starve them of referral traffic; and political strategists being left blind about their donors.

“These are tectonic shifts,” said Shiv Singh, a former C.M.O. and marketing leader for brands like Pepsi, Expedia and LendingTree who runs the industry organization AI Trailblazers. “The way discoverability is changing is like an earthquake nobody expected.”

Google now serves up A.I.-generated summaries before search results. And many consumers are getting their information directly from chatbots. But compared with traditional search, the way large language models source and summarize content is far less predictable or consistent. That has some marketers feeling less in control and more worried about how the models explain, ignore or incorrectly characterize them.

One answer is to buy ads directly from A.I. platforms. OpenAI used the week in Cannes to pitch advertisers on placing ads within ChatGPT, through which the company reportedly expected to earn $100 million annually by 2030. (The New York Times has sued OpenAI and its partner, Microsoft, claiming copyright infringement of news content related to A.I. systems. The two companies have denied the suit’s claims.)

Another approach is to try to show up in A.I. content organically. What can brands do to make sure ChatGPT and Claude, Anthropic’s chatbot, tell users good things about them? Or to show up in lists of recommendations?

Brands have experimented with making the information easier for A.I. agents to read, including by adding more details about products and F.A.Q.s to their websites.

Established tech brands and start-ups alike are creating other solutions. At Cannes, Google showcased the ads in its A.I. summaries of search results. And Adobe, which paid $1.9 billion this year to buy the S.E.O. company Semrush, promoted a product powered by its technology for making brands more visible to A.I. models. At a “Semrush Villa,” Cannes attendees could see a “visibility report” on how a brand appeared to the models, how they compared with rivals and strategies for noticing and addressing content gaps.

A growing class of A.I. start-ups that aims to help marketers with A.I. visibility set up along the beach and in rented yachts during the festival. IV.AI, docked on “Yacht Row,” monitors external A.I. search outputs and suggests language that may improve citations. Jasper, which had one of the many smaller tents along the beach, helps brands generate content optimized for A.I. systems.

And Profound, which tracks the volume of mentions to show how A.I.-generated answers describe and cite brands across ChatGPT, Claude and Google’s Gemini, made its Cannes debut after reaching unicorn status this year when it raised a $96 million round of funding that valued the 2-year-old start-up at $1 billion.

James Cadwallader, a co-founder and the C.E.O. of Profound, said his team had already had nearly 300 meetings by Thursday morning. “It’s been a wild, wild week,” he said.

Many marketers said they were still experimenting with different approaches and solutions.

“It’s not like best practices exist right now,” said Kim Storin, the C.M.O. of Zoom, who is part of a group of C.M.O.s who meet monthly to talk about A.I.-related issues. “We’re creating the best practice ourselves.”

Some brands have concluded they need two versions of some content — one for the machines, one for the humans — said Anda Gansca, a co-founder and the C.E.O. of Knotch, a content marketing start-up.

She was in Cannes to promote her company’s new “agentic content engine,” which helped brands like Zillow and Ally Financial develop a second version of their existing content meant to be more visible and readable to machines.

The idea came after Knotch noticed some sites that were optimizing for A.I. readability were hurting their appeal to humans.

“We have to understand that the more LLMs change us as humans, and the more work they do for us, the more we’re going to want a different experience than agent-first content,” she said, referring to large language models. “We’re looking for an experience that’s multimodal and visual and relevant and personalized.”

Not everyone is buying Elon Musk’s vision for orbital data centers.

Masayoshi Son, the founder and CEO of Softbank, argued at a recent shareholder meeting that building data centers in space won’t do much to cut costs and will take too long when “in the battle for AI, the next few years will be far more important than what might happen a decade or so from now.”

On the latest episode of TechCrunch’s Equity podcast, Kirsten Korosec, Sean O’Kane, and I discussed Son’s remarks as part of a broader discussion that included OpenAI’s plans for custom chips, chipmaker Groq’s new $650 million funding, and much more.

Kirsten noted that it’s “very ironic” that Son is playing the skeptic here, given SoftBank’s “long history of wild bets.”

Sean, meanwhile, said that when Musk talks about “making a constellation of satellites — satellites that need to be replaced every few years as well — to make up an ‘orbital data center,’” he’s just “guaranteeing that much more business” for SpaceX.

Keep reading for a preview of our conversation, edited for length and clarity.

Sean O’Kane: Listen, neo-clouds are the new oil, and everybody who wants to make money is pivoting to a neo-cloud. I’m proud to announce that TechCrunch is now a neo-cloud, give us all your money.

I mean, this is the thing you do. It seems like there are so many players that are compute constrained, so anybody who has a shot at being able to lease out that compute is taking it, whether that’s Groq, a company that was semi-hollowed out by Nvidia, or Allbirds, which went into bankruptcy and and emerged from it as a new neo-cloud provider instead of selling shoes — Tim Fernholz did an interview with the new CEO of of that new effort that I would definitely recommend people go read.

Or whether you’re SpaceX, where your idea was: I’m gonna build an AI platform that’s gonna have an addressable market the size of U.S. GDP, but before we get there, we’ll just rent out our compute. And we saw this continue to happen with SpaceX, where it’s not as big as the deals that they’ve struck with Google or Anthropic, but they just signed another deal, [their] first post IPO deal, to rent out compute to another smaller player. They’re continuing down that road.

You know, I can see this being a business for Groq in the near term. The question with all of these is how durable is it in the long term.

Anthony Ha: If we’re talking about SpaceX and their AI business and data center business, we also have to talk about these comments that Masayoshi Son, the CEO of SoftBank, made recently, where he basically said: What is the point of data centers in space? Which is a question we’ve asked on this show.

And it speaks to, again, this sense in the industry of being really, really compute constrained — they need to build as many data centers as possible, [and] there’s all kinds of reasons why that is proving to be challenging here on Earth, so maybe space is the answer. But I think Son makes some pretty fair points about: All this stuff we’re talking about, even if it all works — and the costs are going to be very, very serious to make it work — this is not happening for years and years and years, so this is not a solution to any immediate problem, as far the current need for data centers goes.

Kirsten Korosec: I just want to point out that SoftBank has a long history of making wild bets. I think it says something when Son comes up and asks the question that a lot of people have asked.

I mean, there are a lot of VCs and founders [who] have been swept up into the idea of orbital data centers and it seems like suddenly everyone’s on board. When just a couple of years ago, I think, if someone had mentioned that, it would get slapped down a little bit. So I do think it’s an important part of the process that someone who has a pretty high profile is asking that question. But it is very ironic to me that he is the one asking it, because if you look at his pitch deck, they’ve thrown a lot of money at some pretty bold ideas.

Sean: WeWork! Listen, we’re going to be saying this for a lot over the next couple years. The idea of putting these things in space is going to be an interesting engineering challenge and certainly an interesting economic challenge.

Anthony, what you said is definitely right to a certain extent. Elon Musk is a person who hates red tape and you know, there are no NIMBYs in space so of course he’s going to try and do that.

To me, it comes down to: The business as it stands now for SpaceX, especially its launch business, is just overwhelmingly reliant on Starlink. The reason that they are 80 or 90% of the launch market globally is not just because they’ve done all these things that are better than pretty much every other launch provider around the globe, it’s also because they have Starlink that is driving up that number. If you remove Starlink from the equation, they would be closer to — I don’t know, maybe 20% or 30% of the launch market, or 40%, but it certainly wouldn’t be 90%.

And when you talk about making a constellation of satellites — satellites that need to be replaced every few years as well — to make up an “orbital data center,” quote unquote, you’re just guaranteeing that much more business for your launch business. And I just can’t stop myself from coming back to that point.

Kirsten: I want to really quickly say that [SpaceX’s] other big business is renting out their compute, by the way. So back to the chip conversation. We’ve come full circle.

Anthony: One of the other themes that may run through this episode is this idea of talking your own book. This is not a new phenomenon. Executives at tech companies, or any other company, what they’re predicting for the future is ultimately the future that is going to be advantageous to their business.

But I think it’s something that’s just always worth remembering when we’re having these conversations about big AI companies, because it is this moment of incredible uncertainty, and we’re all wondering: What does the job market look like in the future? What effect is this going to have on the environment? What are the skills I need to learn?

All these AI CEOs or AI investors, they all have thoughts on that. And it’s not that they’re wrong or that they are being deliberately misleading, but in each case, there’s an asterisk to these predictions. In Musk’s case, he’s talking about something that would be very good for SpaceX’s business. In SoftBank’s case, they are very, very heavily invested in data center projects here on Earth. Sam Altman is the other notable figure who’s rolled his eyes a bit at the orbital data center idea — and again, he and Elon Musk obviously have a long and complicated history together.

All of which is to say that there’s just no objective, impartial observers here. It’s all these people with baggage and tremendous amounts of money at stake.

The Openness That Powered Germany’s Economy Is Now Its Biggest Weakness

The country that once led the world in exports has been stuck in neutral since before the Covid-19 pandemic

BERLIN—Germany’s famously open economy was its greatest economic asset, delivering almost 20 years of uninterrupted growth and turning it into one of the biggest winners of globalization.

Now that openness has become its biggest liability.

China, once a glutton for goods made in Germany, has turned into a mercantilist superpower. It produces many of the same things at a fraction of the price and often with equivalent—if not better—quality. Chinese imports aren’t only flooding Europe but also crowding out German companies in other countries.

Germany has found it is vulnerable to protectionist measures that have cut its companies off from critical resources and technology. Its economy is, meanwhile, being buffeted by external shocks beyond its control—from the rise in energy prices caused by the Iran war to President Trump’s tariffs.

“Germany was certainly a globalization winner,” said Dirk Schumacher, chief economist at the state-owned KfW development bank. “But interdependencies can be weaponized. In a world where the rule-based order is no longer guaranteed, being highly integrated in the global economy can make you more vulnerable.”

The latest illustration came earlier in June when the U.S. stopped the export of artificial-intelligence company Anthropic’s latest large language models on national-security grounds, leaving business AI users in Europe at a competitive disadvantage.

Beijing’s decision to restrict rare-earth exports in the midst of its trade dispute with Washington was another blow, affecting production across Germany in sectors ranging from automobiles to weapon manufacturing.

The government of Chancellor Friedrich Merz has tried to prime the growth pump with tax relief for business and cuts in energy prices, and has ramped up defense and infrastructure spending—but the efforts have yet to bear fruit amid the global headwinds. Berlin recently said it would gradually raise the retirement age to 70 from 67—a move that could eventually improve competitiveness by pruning a system that is funded by employers and employees.

Apart from unemployment and public debt, both of which remain comparatively low, Germany’s economic vitals look decidedly unhealthy. Most economists and the government expect gross domestic product to grow by 1% or less this year. GDP growth has underperformed that of the eurozone since 2019.

Investments have fallen since 2020 while they have risen in France, Italy and Spain. The number of manufacturing jobs in the economy has dropped to 6.6 million, its lowest in 10 years, according to a study published this month by the German Economic Institute, a think tank.

The external shocks have disrupted the government’s economic policy. When the Iran war broke out, it had to shelve work on overhauling its cash-starved welfare system and pivot to gas subsidies for commuters. A welfare-system reform project was scaled down so much that the government canceled the media blitz it had planned, according to a senior official. A more ambitious package of measures should be unveiled in the coming weeks.

The last time Germany’s economy was in a comparable rut was in the early 2000s. It was struggling to absorb the cost of reunification and unemployment was almost twice its current level, pushed up by rigid labor laws and the second-highest labor costs in the world.

In 2003, the government of then-Chancellor Gerhard Schröder took an ax to unemployment benefits, gave employers more say in setting wages and lowered taxes. Within two years, unemployment was falling, exports soaring and public coffers filling up. For six years in a row, Germany was the world’s largest exporter of goods, not just per capita but in absolute terms.

Economists say Germany has lost competitiveness since. But even if it recoups it, this may not suffice to persuade foreigners to buy German cars, medical equipment or tunnel-boring machines they no longer need. It would be like tapping a well that has long dried up.

“Schröder didn’t have to worry about the second China shock,” said Michael Hüther, director of the German Economic Institute.

Cutting the red tape that is stifling the economy could help. Survey after survey shows businesses consider bureaucracy the biggest drag on their activities, even ahead of the U.S. tariffs. Berlin has made some headway in this area but is still resisting calls from business to relax rigid labor laws that make it difficult to hire and fire to match demand.

“Let’s imagine the federal government announced that all business reporting requirements that cannot be re-justified will be eliminated on Jan. 1, 2027,” said Hüther. “That would have an enormous impact. Sometimes we need signals like these.”

Meanwhile, the Anthropic episode shows how German—and European—dependencies aren’t limited to rare earths. Europeans have become adept at building AI applications, but the foundational models, infrastructure and computing capacities underlying them are largely in U.S. hands.

AI “is no longer a simple input into our value chains, but something that will influence all areas of the economy,” said Katharina Erhardt, head of the Industrial Policy Lab at the Kiel Institute for the World Economy think tank. “We have to make sure that this technology is developed here locally. That is much more important than protecting old industries.”

For Schumacher, the KfW chief economist, Berlin faces three priorities as it tries to protect strategic industries while still creating new ones: It should secure more sources of critical raw materials, boost the amount of capital available to grow startups into bigger companies and shield itself against cheap Chinese imports.

Between 10% and 30% of the value of products made by German manufacturers depends on raw materials, such as copper and lithium, that are imported from a handful of places. Berlin has created a raw material fund, but it is still small. Efforts to recycle rare earths or develop batteries that need fewer exotic components are still in their infancy.

Tax reforms and other incentives to encourage private-equity and institutional investors to fund scale-up businesses—as Sweden and France have done—could help boost the innovation that Germany has lacked. By channeling more savings into the market, the coming pension reform could help here too.

Berlin has warmed to proposals by Brussels to clamp down on state-subsidized Chinese imports. Yet with many companies still highly reliant on Chinese inputs, officials say Germany isn’t quite ready to withstand potential retaliation.

“There are costs and risks for sure but they depend on how the other side reacts. China also has something to lose,” said Schumacher. “Yes, you can improve your resilience, but the longer it takes, the more industrial know-how and value creation you might lose.”

Clampdown on top U.S. artificial intelligence is fueling concern that Washington is handing Beijing a cyberwarfare advantage

Chinese artificial-intelligence systems have matched the performance of Anthropic’s powerful model Mythos in some cybersecurity scenarios, a development poised to reset the global tech race and pressure the White House in its overhaul of U.S. AI policy.

Security researchers said that a new AI model, released this month by China’s Zhipu AI, also known as Z.ai, can match the latest U.S. models when it comes to finding security bugs, although it still lags behind Anthropic’s and OpenAI’s products in other tasks.

Overall, the capability gap between top U.S. models and those built by Chinese companies has narrowed significantly, and use of Chinese AI systems has surged as businesses seek to rein in runaway costs. A host of companies, including Microsoft, are weighing how they can offer Chinese models on their platforms, a development that is set to alter the balance of power among tech companies.

“China is making sure that the gap becomes smaller and smaller over time,” said Lior Div, chief executive officer of the cybersecurity company 7AI.

The ability of AI systems to find bugs in software has added urgency to efforts to use models to close quickly vulnerabilities that could be exploited by hackers. Otherwise, the world will face what some researchers have called a bugmageddon.

Unlike models from Anthropic or OpenAI, Zhipu’s GLM-5.2 is open-weight. That means it can be downloaded and run on hardware operated by anybody and can be modified and used without supervision. Open-weight models are ideal for users who want unfettered access to systems they control, but they are also ideal for hackers, who can run them in the shadows.

GLM-5.2 has ranked as one of the 10 most-used AI models, according to data from OpenRouter, a company that provides access to more than 400 AI models. In some benchmarking tests, according to the cybersecurity company Semgrep, GLM-5.2 bested Anthropic’s Claude Opus 4.8 model, which was released in May. When given further instructions, Opus 4.8 and GLM-5.2 can match Mythos in bug-finding ability, according to researchers.

On Wednesday, the Chinese cybersecurity company 360 Security Technology released a new bug-finding tool called Tulongfeng. The company said it was comparable to Mythos in finding bugs. Those capabilities have alarmed many national-security officials and CEOs.

“This kind of powerful weapon that can alter the landscape of cyberwarfare can’t remain solely in American hands,” 360 Security Chief Executive Zhou Hongyi said at a cybersecurity conference in Beijing. Zhou, an outspoken internet veteran and member of China’s top political advisory body, said China would face unacceptable risks if American entities could use advanced AI models to scan critical Chinese network systems while denying Chinese companies comparable capabilities.

China’s advances coincide with unprecedented U.S. government roadblocks to developers releasing models. On Friday, OpenAI said it was limiting access to its latest model, known as GPT-5.6, because of security concerns among administration officials. The company warned that the current case-by-case model-evaluation process wasn’t a long-term solution but said it is being used while a recent executive order focused on security and model oversight is implemented.

One of Anthropic’s latest general-use models has been shut down for more than two weeks after the Trump administration said no foreign entity or individual could use it because of security risks. The company closed all access to comply with the rule. The administration on Friday restored some access to a related Anthropic model called Mythos 5, which had previously been restricted.

Many have called the administration’s attack on a leading U.S. AI company counterproductive and criticized its decision to allow exports of AI chips to China in light of the nation’s recent advances.

“Banning Fable while selling chips China needs to develop its own version is a gift to China,” said Saif Khan, a distinguished technology fellow at the Institute for Progress think tank who worked on export restrictions in the Biden administration. The U.S. needs to maximize the use of Mythos and comparable models to harden its cyber defenses while it can, he added.

Among the Mythos 5 and Fable 5 users that had lost access before Friday’s decision to restore Mythos 5 access for some trusted entities: the National Security Agency, which had been testing the tools and found them impressive in trials, according to people familiar with the matter.

Critics of the White House approach have said it has been lax in restricting use of Chinese open-weight models from companies such as DeepSeek and Zhipu, which are popular among U.S. businesses.

Some companies have evaded existing chip-export restrictions, while others have used distillation—in which a new system learns from an existing one by asking it hundreds of thousands of questions and analyzing the answers—to benefit from U.S. advances.

“Our administration is very much focused on Chinese open-source models,” said Jacob Helberg, who is undersecretary of state for economic affairs and a former tech adviser and investor. “It’s something that we’re tracking very closely.” He spoke in a recent interview at a summit for a coalition of countries working to secure supply chains and counter China’s influence in AI.

In one sign the administration wants to boost U.S. open-weight companies, the Pentagon recently announced a deal with one of the few domestic open-weight developers, Reflection AI, for use in classified settings along with a host of similar agreements.

At the same time, AI users said that U.S. efforts to rein in the worrisome capabilities of recent cybersecurity-focused models have added to concerns that access to needed systems could eventually be cut off.

“It is incentivizing companies across the globe to use cheaper but very capable Chinese open-weight models, while at the same time undermining the U.S. AI industry,” said Niels Provos, a researcher who led security teams at Google and Stripe. “I don’t understand it.”

A Rising Star’s Abrupt Fall Shows Xi’s Crackdown on Overambitious Officials

Purges aimed at rooting out corruption and disloyalty push officials to focus on pleasing Beijing

When Jiang Duntao shot up the ranks to become the top Communist Party official in Zibo, China, in 2019, he moved fast, pivoting the local economy from heavy industries toward higher-end manufacturing and services—while raising debt and tapping private investment to fund the projects.

The city in eastern China reported strong economic growth under Jiang and went on to win viral fame as a vacation spot. State media portrayed Jiang’s pro-business policies as bold and dynamic, and he was quickly promoted.

But his career collapsed abruptly within a few years when the party accused him of saddling Zibo with onerous debt.

Jiang was “eager for quick success, and acted blindly and recklessly,” the party’s top disciplinary agency said in April when announcing his expulsion.

Such projects would have won plaudits in China’s go-go years of red-hot growth. But as the country struggles with a slowdown and enormous debt, Chinese leader Xi Jinping is cracking down on officials who try to boost their careers with flashy projects and rash policies. Such practices “exhaust the people and drain the treasury,” Xi said in February.

This turnaround is also exacerbating a culture of fear and uncertainty in the ranks of Chinese officialdom. The waves of purges aimed at rooting out corruption and disloyalty have pushed many officials to focus on finding ways to please Beijing and avoid punishment.

Some officials have responded by putting off decisions while awaiting clearer signals from higher-ups on what to do. Others have focused on traditional benchmarks such as economic growth, seeking high-profile investments and “prestige projects” to impress superiors. In some cases, bureaucrats resorted to ostentatious gestures and outright fraud, feigning adherence to unrealistic targets and masking underperformance.

Party authorities punished nearly 160,000 people for offenses related to policy inaction, recklessness or deceit last year, a 16% increase from the 2024 total, according to official data.

Such behavior has grown more troubling as China confronts economic problems ranging from sluggish demand to a depressed property market. Some analysts said Beijing exacerbated the problem by issuing looser economic targets and more-ambiguous demands—such as delivering “high-quality development”—which made officials less certain about how to implement Xi’s policies.

“Replacing a hard metric like growth with softer goals such as improving business confidence or social welfare is easier said than done,” said Lizzi Lee, a fellow at the Asia Society Policy Institute. “Hard targets are at least visible and enforceable. Softer objectives are harder to measure and harder to embed in the cadre evaluation system.”

Party officials are trying to clarify Xi’s message through a new indoctrination campaign. The program features seminars, field trips and inspections emphasizing that party members must work to improve people’s lives instead of protecting their own careers.

Some agencies have party members read offenders’ confessions and watch documentaries depicting cases of self-serving officials who worked for their own gain. Beijing has also sent “central guidance teams,” typically led by retired senior officials, into some provincial governments, state agencies, businesses and universities to promote Xi’s ideas on how to be a conscientious official.

The party is naming and shaming offenders. One case featured a vice mayor in southwestern China who allegedly chased quick results by approving titanium plants and signing billions of yuan of investment deals without proper due diligence. Production at some plants was halted for regulatory violations, while less than 7% of the investment agreements he oversaw materialized, according to party authorities, who expelled the official and transferred his case to prosecutors.

Another case involved officials in the central China city of Leiyang who built an industrial park for making strollers. Party inspectors said these officials inflated reports on the park’s output, claiming that more than 60 enterprises had moved into the industrial park when it only had five businesses with a combined workforce of fewer than 100 people.

Beijing has tried to guide officials by directing them to study Xi’s ideas on what constitutes good political performance and how to make China’s economy more innovative, environmentally friendly and equitable.

Party publishers issued a series of guidebooks as required reading, including one title that listed 72 positive and 74 negative practices that officials should follow or avoid. It promoted positive practices such as visiting local communities more often, while negative behavior included “thinking that doing more means more mistakes, doing less means fewer mistakes.”

Jiang, the former Zibo party chief, appeared at first to be fulfilling the party’s wishes. He halved the number of chemical companies in the city to around 500 to reduce its reliance on older, more-polluting industries. He started projects to build higher-education and science hubs, as well as an “entrepreneurship and innovation valley.”

Aiming to rehabilitate Zibo’s image as an industrial backwater, Jiang promoted the city as a production center for precooked meals, aiming to “allow the ‘Zibo flavor’ to spread its fragrance around the world.” He promised to boost the city’s nightlife with new bars, cafes and music lounges—efforts that paid off when Zibo became the center of a barbecue craze in 2023 that drew tourists from across China.

Jiang became the party chief of a neighboring city in 2022 and became vice mayor of the inland megacity of Chongqing the next year.

But Zibo’s finances grew strained. The city’s debt by 2022 reached about 108 billion yuan, the equivalent of about $16 billion at current rates, double its levels in 2018—the year before Jiang was appointed as the city’s chief.

His fortunes turned abruptly in October, when the party’s Central Commission for Discipline Inspection announced an investigation against him. Then, in April, the commission said the party had expelled Jiang and transferred his case to prosecutors. He couldn’t be reached for comment.

Party inspectors said Jiang set up government funds without proper assessments or consultation and allowed private entrepreneurs to abuse them, resulting in significant losses to state assets. He also allegedly raised large amounts of debt to fund prestige projects, even though he knew Zibo’s fiscal revenue couldn’t sustain such spending.

Jiang advocated the idea that “a government that doesn’t borrow is one that doesn’t act,” a notion that contradicts Xi’s calls for pursuing sustainable growth, the commission said. “He seized political achievements for personal gain and ultimately set back the party’s cause,” the commission said.

Surging appetite for advanced models is turning computing power into the tech industry’s scarcest commodity

Google has put limits on Meta’s use of its Gemini AI models after the social media giant sought more computing capacity than the rival tech group could provide, in the latest evidence of the infrastructure constraints facing even the world’s largest AI providers.

Google told Meta around March that it could not provide all of the Gemini capacity the company wanted to purchase, according to three people familiar with the matter, in a move that has disrupted and delayed some of Meta’s internal AI projects.

Owing to the restrictions, which remain in place, as well as a broader push to streamline AI costs, Meta has encouraged staff to be more efficient with AI tokens — the units that measure AI usage, several people said.

Several other Google clients have been affected by the restrictions, although to a lesser extent, according to one person familiar with the matter. Meta has been particularly impacted because of its exceptionally high demand for Google’s models, the person said.

The decision by Google to cap a large customer’s access to its models offers a rare glimpse into the infrastructure pressures and bottlenecks building across the AI industry.

Despite spending tens of billions of dollars on chips, data centres and power, even the largest tech companies are struggling to secure enough computing power to support surging demand for advanced models and AI services.

As a direct result of the demands, particularly from big corporate customers such as Meta, Google has raced to secure additional capacity, according to one person familiar with the matter. Google earlier this month signed a $920mn-a-month deal to lease computing capacity from Elon Musk’s SpaceX.

Google and Meta declined to comment.

At its first-quarter earnings in April, Google chief executive Sundar Pichai said that the company’s cloud revenue exceeded $20bn for the first time, while its backlog of signed — but not yet delivered — cloud contracts nearly doubled quarter on quarter to more than $460bn.

“Obviously, we are compute-constrained in the near term,” Pichai said. “And as an example, our Cloud revenue would have been higher if we were able to meet the demand.”

Demand for AI computing has risen sharply as companies deploy chatbots, coding assistants and AI agents across their businesses.

The resulting increase in inference workloads — tasks required to run models after they have been trained — has emerged as one of the industry’s biggest challenges.

AI lab Anthropic, the maker of the popular Claude chatbot, last month struck a deal with SpaceX that is similar to the deal it has with Google.

The constraints illustrate the extent to which Meta has relied on rival models such as Gemini, as the social platform spends aggressively to become a leader in AI and improve its own models. Chief executive Mark Zuckerberg has been pouring billions of dollars into tapping talent and securing infrastructure in order to develop what he dubs “personal superintelligence”.

Unlike Google, Meta does not have a cloud business and is racing to build out its fleet of data centres for its own training and inference needs. As part of the push, Meta has committed to investing $600bn in the US by 2028.

Gemini has been used internally at Meta as part of a push to automate some of its safety processes, such as rooting out scams and taking down harmful content, as well as for its customer services and advertising help chatbots. It is also used internally for some workflows and coding, alongside other models such as Anthropic’s Claude.

Meta initially chose to use Gemini because it performed better than the social media company’s own Llama open-source models, according to people familiar with the matter.

More recently, Meta has begun to shift to prioritise its new Muse Spark model, several people said, which is viewed as more competitive with Gemini and reduces the company’s dependence on external models for some applications.

Conno Christou doesn’t leave things to chance. He tracks his sleep with a Whoop band, cross-references it with an Oura ring, and gets nearly 100 biomarkers checked every year. He had been doing the annual bloodwork for four consecutive years, following the protocols of longevity researchers like Peter Attia and Rhonda Patrick. He was optimizing his supplements, his circadian rhythm, his protein intake.

At 35, building his second company, he was as dialed-in on the latest in health research as anyone he knew. His last checkup, in 2025, was green across the board. “It was the best I’d had in years,” he says.

Then, after a workout, his arm swelled.

He didn’t think much of it at first. A week passed before he saw a doctor, who found two blood clots in his veins and scheduled surgery. But the pre-op exams changed everything. A doctor walked back into the room and told him the procedure wasn’t happening.

“We see an 11-by-11-by-8 centimeter mass behind your sternum,” the doctor said.

A biopsy confirmed what Christou had never before even contemplated. He had an aggressive, fast-growing form of non-Hodgkin’s lymphoma — a rare diagnosis affecting roughly one in 420,000 people, caused by a random genetic mutation with no connection to lifestyle, diet, or stress.

The tumor had only existed for about three months. In three more weeks, it would have reached stage four.

“Lucky in my unluckiness,” Christou told this editor this week from his home in Athens, where he lives part time. “It was only found because I went in for something else entirely.”

What followed was an education in the limits of the medical system, and in what a determined patient can do about that with tools now available.

His first oncologist, a renowned specialist, recommended the lighter of two available chemotherapy regimens. Christou booked his first infusion three days out. Then, the night before, he sought a second opinion.

That second doctor didn’t hesitate. He recommended the harder regimen — continuous in-hospital infusion, cycling every three weeks across six months — citing Christou’s specific pathology. The lighter treatment carried roughly a 60% success rate for his presentation. The aggressive one brought that number to around 85%. Two world-class doctors. Diametrically opposite recommendations.

“As founders, we hold the wheel,” Christou says of the propensity of many people to accept what they are told — and why more should not. “You hear many things. You don’t have to follow the first advice.”

He didn’t opt to just follow the advice of the second physician, either. Over the next two days, he gathered 12 opinions in total — drawing on his professional network, reaching out to hematologists and oncologists in the US and abroad, calling in every favor he could. Eleven to one voted in favor of the harder path. He took it. The decision, he says, didn’t feel brave so much as logical. He was already a data-driven person, and now the stakes felt existential to him.

Over six months of treatment, Christou approached chemotherapy the way he approached building a company, as a marathon of sprints — each of them with a finite cycle and each week filled with data points. He had done a mandatory 25-month military service in Cyprus at age 18 and he borrowed from that experience, too. He was going to be a good soldier, he told himself. Trust the process. Six cycles. Get through it.

He wore his Whoop throughout, and found it remarkably accurate at predicting the days his immune system would bottom out, sometimes flagging them before symptoms arrived. He kept a symptom journal using voice transcription, logging every shift, every side effect, every medication and counter-medication. He narrowed his focus to three variables: sleep, nutrition, and, first and foremost, psychology. (“It moves the needle more than anything,” said Christou. “I never asked ‘why me’ — not once. That question has no useful answer.”)

He fed all of it — blood results, scan data, wearable output, journal entries — into Claude. He’s far from alone in turning to chatbots for medical guidance. A public opinion poll released in March found that a third of American adults now use them for health information and advice. The stories accumulating online suggest that for some patients, AI is delivering what the system couldn’t.

Experts urge caution; Danielle Bitterman, clinical lead for data science and AI at Mass General Brigham, has told the New York Times in recent months that general-purpose chatbots are frequently wrong and “have not been thoroughly evaluated” for personalized diagnoses.

Christou doesn’t disagree. “It didn’t replace the doctors,” he says, but it “helped me ask the right questions.”

On Wednesday, Chinese cybersecurity firm 360 reportedly unveiled Tulongfeng, an AI tool it says can go head-to-head with Anthropic’s Mythos. That’s the cybersecurity-focused AI model that is reportedly so powerful, the Trump Administration has currently banned it and its more restricted version, Fable 5, from the hands of non-Americans.

Earlier the same week Sakana AI, a Tokyo-based AI startup launched Fugu, a model named after the Japanese word for blowfish. The company says this frontier AI model “stands shoulder-to-shoulder with leading models like Anthropic’s Fable 5 and Mythos Preview.” It is also designed for agents, with an ability to orchestrate access to other models though their APIs.

The two new Asian model products come as the U.S. government’s ban drags on. It’s order that prevents Anthropic from global access to Mythos and Fable occurred two weeks ago.

A spokesperson at Sakana AI told TechCrunch that release of its new model was “entirely coincidental,” yet that hasn’t stopped it from capitalizing on the moment. It’s website advertises “delivering frontier capability without the risk of export controls.”

“Sakana Fugu is something we have been building since last year — the research behind it was presented at ICLR this spring, and it reflects an approach that is central to how we deliver frontier-level value at Sakana AI. We were confident in the product on its own merits; the timing simply happened to coincide with a moment that brought it more attention than we expected,” the spokesperson said about launching during the Mythos/Fable export ban.

Sakana, co-founded in 2023 by former Google researchers Ren Ito, Llion Jones and David Ha, makes affordable generative AI models that work well with small datasets and are optimized for the Japanese language and culture.

While the company is targeting Fugu at Japanese businesses and government agencies looking to reduce their exposure to tightening export controls, it isn’t yet proclaiming a lasting shift away from U.S. AI in Asia.

“U.S. models remain important to Asia,” the spokesperson said, a view consistent with remarks co-founder Ren Ito made at the G7 summit in Evian last week, where AI access and export controls were one of the central topics. “We’d characterize the current moment in those terms rather than as a permanent realignment toward any one set of players.”

Sakana co-founder Ren Ito elaborated on that view in an op-ed published in the Project Syndicate last week. He urged the US federal government, that consider that its “first priority should be to preserve access,” for America’s closest allies, and argued that “AI should not become a technology that is hoarded; it should be one that is developed together.”

David Ha, co-founder and CEO of Sakana, described Fugu as more than just a land grab during a vulnerable moment for a US competitors. It is designed to coordinate agent usage among many models.

“Orchestration Models are the next frontier, beyond bigger models,” he wrote on X. Relying on a single provider for national infrastructure, he argued, is a risk the recent export controls made impossible to ignore.

“Access to top models can disappear overnight,” he wrote. “Collective intelligence is the practical hedge against this concentration of power.”

While Tokyo-based Sakana positioned Fugu as a hedge strategy, a way to preserve access to frontier AI, not replace it, China’s 360 wasn’t hedging.

The Chinese firm reportedly unveiled two AI security tools. Tulongfeng is designed to automatically discover software vulnerabilities, and Yitianzhen is built to automate cyber defence and incident response.

The product launch, however, came with a message. According to Reuters, 360’s founder Zhou Hongyi described vulnerability-finding AI as a national strategic asset, and flagged what he called the risk of “one-way transparency”, a situation in which some actors could access advanced vulnerability-detection capabilities while others could not.

Anthropic had been on a historic growth trajectory. The US AI lab said its run-rate revenue crossed $47 billion in May 2026. How much of that depends on Asian enterprise customers is not publicly known.

But in the weeks since the export order took effect, at least two companies, one in Tokyo, one in Beijing, have stepped into the space it left behind. Even if US companies could win back trust should this ban ever end, local alternatives, trained to better understand local language and nuance, are already filling the gap.

360 did not respond to a request for comment.