Humanoids aren’t quite ready to replace factory workers, but the industry can’t wait. Faced with labor shortages, manufacturers have shown growing interest in startups that promise faster automation without the usual tradeoffs.

That’s the bet behind Theker, an AI robotics startup that aims to go beyond robots trained for a single task. “If you always have to put the same cookie in the same box, that works perfectly, but most processes aren’t like that,” co-founder Carla Gómez Cano told TechCrunch.

Theker is designed for that messier reality. Unlike humanoid robots designed around a fixed form — think Boston Dynamics — Theker’s machines are built to be reconfigured. Their hands, arms, and overall form can be swapped out or resized depending on the task, whether that’s sorting packages, packing clothing, or handling bottles and cans in a warehouse.

That Inditex, Zara’s parent company, signed on as an early backer is a signal of where Theker’s ambitions start, not where they end. The company’s broader goal is to move beyond retail into heavier industrial settings like manufacturing, where the complexity and scale of manual tasks is even greater.

This generalist ambition has helped cement Theker’s status as one of Europe’s hot startups to watch — and raise capital accordingly. The Barcelona-based startup has just raised $85 million in what it’s calling “Europe’s largest ever robotics Series A.” (We haven’t found a larger one in our records, either.)

Less than a year after a record seed round, this Series A was led by American VC firm CRV and backed by a mix of traditional and strategic investors, including Samsung and Aglaé Ventures, the investment vehicle tied to LVMH chairman Bernard Arnault.

Gómez Cano said Samsung is not a client yet but that the two are in advanced discussions. Theker would welcome having the Korean company as a customer, supplier, and investor simultaneously — a trifecta that would give the startup both revenue and credibility in manufacturing at scale.

She also noted that she and co-founder Jiaqiang Ye Zhu “didn’t build Theker to run pilots,” so the team skips innovation departments entirely and goes straight to logistics or operations, where deals are real and timelines are shorter.

To demonstrate that the company can actually deliver on that, Theker has a showroom in central Barcelona, and plans to open others as it expands across Europe, the U.S. and Asia. It will also grow its headcount across tech, deployment, and sales.  

“We already received 15,000 job applications and have to filter like crazy,” Gómez Cano said. She estimated that the team could grow from dozens to up to 120 people by the end of the year, then caught herself: “I am saying that, but I also said that we’d raise $30 or $40 million!” 

That Theker managed to raise twice its target also reinforces the startup’s conviction in keeping its HQ in Barcelona, a growing robotics hub, and in Europe’s tech ecosystem more broadly. “It has never been a barrier to acceleration for us, so we are making the most of it,” Gómez Cano said.

Prometheus, the physical AI startup co-founded by Jeff Bezos and Vik Bajaj, the former co-founder of Verily, Google’s life sciences unit, announced it raised $12 billion at a $41 billion valuation.

The new funds came from Bezos himself, as well as from JPMorgan Chase, Goldman Sachs, and BlackRock, among others.

This is the second fundraise round for Prometheus, which launched late last year with an initial raise of $6.2 billion, according to CNBC.

Prometheus is building what it calls an “artificial general engineer” — software capable of automating the design and manufacturing of complex physical systems, from jet engines to drug compounds.

The ambition is sweeping: replace large swaths of engineering work with AI. Although the startup will automate many aspects of an engineer’s job, Bezos told CNBC that the productivity gains AI delivers will lead to what he calls “labor scarcity” — his term for a world where demand for human workers outpaces supply.

That puts him at odds with a number of prominent voices in tech. While some AI leaders predict widespread job losses, Bezos sees it differently.

“Significant productivity in the economy is going to raise the standard of living,” he said. “People who today have two-earner households, they’ll become one-earner households. Maybe some people who are working overtime will stop working overtime.”

The company, which currently has 150 employees across offices in San Francisco, London, and Zurich, is keeping the specifics of what it has already built under wraps.

Bezos indicated that a large portion of the capital will go toward the company’s large compute needs.

Bezos knows something about labor at scale. Amazon — where he serves as executive chairman and is the largest individual shareholder — employs more than 1.5 million people worldwide and over the past year, under CEO Andy Jassy, has laid off tens of thousands of people as the company has accelerated its own automation push.

At $41 billion, Prometheus is one of the most richly valued AI startups ever funded, and one of the largest single bets on the physical AI sector. But it isn’t the only company attracting massive investor interest. In recent months, venture capitalists have increasingly poured capital into physical AI, a booming sector that investors and founders argue is inherently more defensible than pure software — because the physical world creates moats that code alone cannot.

Swiss cleantech startup GR3N has raised €15.5 million in a Series B round to build the world’s first commercial-scale microwave-assisted PET recycling plant. The round was led by 360 Capital, with new investor VP Textile also participating. The proceeds will fund MODUS, a 40,000-ton-per-year facility in Spain.

PET is one of the most widely used plastics on the planet. But 98% of recycling relies on mechanical technologies that can only process transparent and light-blue bottles, roughly 15% of total PET waste. The remaining 85%, including textile fibres, films, and coloured resins, ends up in landfill or gets incinerated. It is one of the biggest gaps in the EU’s push to hit its emissions targets.

GR3N’s technology, called MADE (Microwave Assisted DEpolymerisation), can process all of it. Unlike mechanical recycling or alternatives like glycolysis and methanolysis, MADE has no feedstock limitations. It produces food-grade monomers that can be recycled repeatedly without losing performance, while cutting CO2 emissions by up to 80% compared to virgin PET production.

The MODUS plant has secured a €35 million grant agreement under the EU Innovation Fund’s large-scale industrial projects category. Intecsa Industrial, part of the Cobra IS group, is leading the engineering and construction. Financial closure is expected in Q4 2027, with commercial operations planned for Q2 2030.

GR3N was founded in 2013 by Italian inventor Maurizio Crippa, who recently handed the CEO role to Martin Stephan. Stephan brings two decades of experience in technology-driven businesses. Industrial shareholders now include Intecsa Industrial, Standex International, and Chevron.

The round is modest by venture standards, but the EU grant and industrial backing signal that the technology is past the lab stage. If MODUS delivers on its 40,000-ton annual capacity, it would demonstrate that chemical recycling can work at scale for the types of PET waste the industry has struggled to handle. Europe has twice as many climate tech startups as the US, but scaling deep tech from lab to factory remains the hard part.

Megaport spent a decade as a company you used to connect to other people’s clouds. On Wednesday it announced a plan to become one. The Australian networking firm secured four new AI infrastructure contracts worth a combined A$458.9M (about $329M) and launched a fully underwritten entitlement offer to raise A$827.3M (about $594M), according to its filing. The money funds a pivot from plumbing to compute.

The contracts come first. All four are with US-based technology providers running AI applications, are expected to start in the first half of 2027, and require nearly A$369.5M in capital expenditure, mostly for high-performance Nvidia GPUs alongside network and storage. That is a meaningful commitment for a company of Megaport’s size, and it explains why the raise is so large relative to the business.

What the capital is really buying is the strategy behind the contracts. Megaport says it will build a globally distributed AI inference cloud, anchored by an on-demand GPU pool backed by about A$350M in investment and offered to enterprise customers on both contracted and consumption-based pricing.

The pool is to be deployed across the company’s existing footprint of more than 1,100 connected data centres in 31 countries, with rollout over the next six to nine months.

The bet is geographic. Most GPU capacity today sits in a handful of enormous data centres optimised for training the largest models. Megaport is targeting the other half of the AI workload: inference, the act of running a trained model to answer a query, which benefits from being close to the user.

Its pitch is that a distributed network of smaller GPU pools, spread across the data centres it already connects, fits inference better than centralised mega-campuses, and slots into the gap between hyperscaler clouds and single-location GPU specialists.

It is a credible reading of where AI infrastructure is heading. As models move from research demos into products embedded in real applications, the economics shift from training to serving, and serving rewards proximity and distribution.

Megaport already owns the network that links the locations where that compute would live, which is a genuine structural advantage if the thesis holds.

The numbers around the raise were briefly muddled across early coverage, which is worth untangling. The four contracts are worth A$458.9M in total contract value; the capital raise is A$827.3M; the GPU pool commitment is about A$350M.

Several headlines collapsed these into a single figure. They are distinct: contract wins, the money to fund them, and the specific compute investment inside that money.

Megaport also tightened its 2026 revenue guidance to A$307M–A$315M and projected combined group pro forma annual recurring revenue of A$662.9M once the compute division is folded in. The shares were halted while the raise was arranged, a standard mechanism for a deal of this scale on the ASX.

The risk is the obvious one for any company spending heavily on Nvidia GPUs on the strength of contracts that begin in 2027: that AI infrastructure demand, and pricing, may look different by the time the hardware is installed and earning.

Megaport is committing capital now against revenue that lands later, in a market moving fast enough that 18 months is a long time. The contracts give it a floor. The inference-cloud ambition is the part that has to compound, and that is the part the A$827M is really betting on.

Before Zack Eakin sold investors on his new startup, he practiced on Palmer Luckey.

When Eakin left Luckey’s defense startup, Anduril, in 2024 to start a new composites company called Layup Parts, Luckey — along with Anduril co-founders Brian Schimpf and Matt Grimm — let him workshop the pitch.

He got different feedback from each, Eakin told TechCrunch. Grimm helped him think about how to pitch VCs, Schimpf (Anduril’s CEO) pushed him on strategy, while Luckey — ever the fundraiser — guided him on the storytelling.

This miniature boot camp appears to have worked. Two years ago, Eakin raised a $9 million seed round. The startup announced Tuesday it has raised another $42 million in a Series A funding round led by dual-use venture fund Marlinspike, with participation from new investors Cerberus Ventures and Pinegrove Venture Partners, and existing backers Founders Fund and Lux Capital.

It’s a tidy sum for the Huntington Beach, California, startup, which employs just 60 people or so. And much of it will go toward people. Layup Parts used most of its seed money on capital expenditures. Eakin wants to use the new funding to grow the startup’s ranks and move into a bigger facility this year. The goal is to make ordering custom parts made of carbon fiber or fiberglass as easy as if they were sold on Amazon.

Eakin has been working with composite materials for around two decades, dating back to his time in motorsports, he told TechCrunch. The engineer started his professional career at Chip Ganassi Racing, where he worked with carbon-fiber structures and bodywork, especially for the company’s IndyCar entries and the radical (and radically controversial) DeltaWing prototype.

Eakin took a bit of a detour to become the first engineer at Elon Musk’s Boring Company in 2017. But by 2021, he was once again elbow-deep in composites when he took the role at Anduril.

It was at that point Eakin realized how, during his time digging tunnels, something of a revolution had started in the worlds of industrial fabrication and manufacturing. Startups like SendCutSend and Protolabs had dramatically reduced the time and cost required to prototype and ship parts to customers. But no one was doing this for composites, he said.

“It just kind of dawned on me that, like, all these other manufacturing verticals are getting better, [and] we are struggling to find people to make our composite parts for us,” Eakin said. “Why is there nobody trying to make this better?”

It’s not that Eakin didn’t know the answer. Composites tend to be harder to deal with in general — or, as he put it, there are “a lot more fingers and eyeballs involved.” Plus, there had been a lot of consolidation among composite companies, according to Eakin.

This meant bigger firms were less likely to try to innovate and risk their dependable revenue streams. And even if they wanted to, he said, these companies don’t have the software talent to build the tools required to reach that goal of getting to a one- or even zero-click solution.

“If we have stock materials, and you have a good understanding of those materials, we can build software that has an order of magnitude reduction in the amount of clicking it takes for an engineer to produce those — and ultimately gets to zero clicks, where it just takes customer data and poops out shapes,” he said with a smile.

Eakin said it became obvious that the best way to do this was to start a whole new composites company and that these challenges made the idea all the more valuable.

“I just decided this might be the best thing I can do for Anduril, is to go fix this part of the supply chain, because I don’t think it’s just an Anduril problem,” he said.

So far, he’s been right. In the two years since Eakin founded Layup Parts, his team has been rapidly prototyping and producing parts for a variety of customers, including motorsports, design studios making show cars, and even pickleball paddle companies. The company has already cut the time between receiving customer data and manufacturing a part down from weeks to hours in some cases.

The biggest business lines, unsurprisingly, are aerospace and defense. That includes both startups and the more traditional defense primes, according to Eakin.

The opportunity is evident in the cap table. There’s the lead backer Marlinspike, which is already invested in Anduril and a number of other defense-focused manufacturing companies. Cerberus Ventures was started in 2023 by Chris Darby, who spent nearly 20 years running the CIA-backed venture firm In-Q-Tel.

While Eakin looks back fondly on what he learned from Anduril and its leaders, he’s also carrying over skills he learned at The Boring Company. Despite not working with composites there, he said a lot still applies to a startup. Working at The Boring Company involved a lot of “first-principles engineering stuff, very similar to what we would do in racing,” he said.

“Elon has a very high sense of urgency, so as much as it was a new type of thing to make, it felt familiar with the crazy deadlines and just developing stuff as fast as you can,” he said.

The company’s developer conference opens in San Francisco with another round of AI tooling expected, against the awkward backdrop of how few people pay for Copilot.

Microsoft opened its annual Build developer conference in San Francisco on Tuesday, with Satya Nadella due to take the stage at 9:30am Pacific for a keynote that, by every signal the company has sent, will be about putting artificial intelligence into as many corners of its products as it can.

The conference runs June 2 and 3, in person and online, and is pitched squarely at the developers Microsoft needs to build on top of its platforms.

The broad direction is not in doubt. Microsoft has spent the past year reframing Windows as a host for AI tools and agents rather than a passive operating system, and Build is where it courts the developers who would write for that vision.

Reporting ahead of the event, including from Reuters, pointed to new PC-side and cloud AI tooling as the centrepiece, the latest instalment in a strategy Nadella has pursued relentlessly since the company’s OpenAI partnership began.

The specifics of what was unveiled on stage should be read against the official Microsoft announcements rather than the preview chatter, which this year has been unusually heavy on speculation.

What gives the event its tension is not the technology but the take-up. On its most recent earnings call, in late January, Microsoft said it had 15 million paid seats of Microsoft 365 Copilot. That is a striking number until it is set against the 450 million commercial Microsoft 365 seats the company also reported, at which point it becomes a conversion rate of roughly 3.3%.

Nadella told investors that Copilot was “becoming a true daily habit,” citing daily active users up tenfold year on year, but the gap between people who can use Copilot and people who pay for it remains the most stubborn fact in Microsoft’s AI story.

The company has been acting on that gap in ways that complicate the upbeat keynote framing. Earlier this year it began letting users and administrators uninstall Copilot from Windows 11 outright, a concession to the many who never wanted it bundled in. At the same time it has been building out its own MAI model family, a move widely read as an effort to depend less on OpenAI for the intelligence under Copilot’s hood.

So Build arrives as both a showcase and a sales pitch. Microsoft has the distribution, the cloud, and the developer base that almost no rival can match, and it has a product millions of people have access to and decline to buy. The keynote will be heavy on what AI can now do inside Windows and Azure.

Whether developers, and the customers behind them, decide it is worth paying for is the question the announcements are really meant to answer. The conference will say what Microsoft is building. It will not, by itself, settle who is buying.

The Swiss post-quantum-crypto firm has lifted its WeCan stake from 28% to majority and committed CHF 5M to accelerate AI-compliance tooling for Pictet, Lombard Odier and Barclays.

---

Geneva-based SEALSQ has acquired a majority equity stake in WeCan Group, lifting its position from the 28% taken in October 2025 and committing a further CHF 5m ($6.1m) to accelerate the joint development of an AI-powered compliance co-pilot built on post-quantum cryptography for the global financial industry.

The investment is being made through SEALSQ’s dedicated Quantum Fund, an internal strategic vehicle the company set up to back early-stage quantum-computing and quantum-security companies.

The customer base WeCan brings to the platform is the part that makes the announcement substantive. The company already counts Pictet, Lombard Odier, Edmond de Rothschild, Syz and Barclays among its private-banking clients for its existing know-your-customer infrastructure, which has been operational across several Swiss tier-one wealth managers since 2020.

The WeCan compliance stack uses blockchain-based document verification to handle the KYC and AML data-sharing that private banks have historically struggled to coordinate, even when serving overlapping client books. SEALSQ’s majority position now puts that customer base alongside SEALSQ’s own quantum-resistant Hardware Security Modules (HSMs) and post-quantum-cryptography (PQC) IP.

The product the two companies are now building is the more strategically interesting half. The joint platform will use SEALSQ’s PQC primitives to handle the cryptographic layer, WeCan’s existing private-bank customer integration to handle the operational layer, and an AI compliance co-pilot to absorb the rules-based-to-judgment work that private-bank compliance teams currently do by hand.

The pitch, on the company’s own framing, is a compliance system that survives both the EU AI Act’s 2027 enforcement window and the anticipated post-quantum cryptographic transition that NIST’s August 2024 PQC standards have started to make mandatory.

The post-quantum dimension is the part most likely to be under-priced by readers outside the cryptography community. Existing public-key cryptography, the RSA and elliptic-curve protocols underlying almost every banking-and-government secure-communications system in use today, will be broken by sufficiently capable quantum computers.

The estimates of when that breaking becomes practical range from five to fifteen years, but the “harvest now, decrypt later” threat, in which adversaries capture encrypted financial traffic today and decrypt it once quantum hardware matures, means the transition needs to start now rather than then.

NIST finalised its first set of post-quantum standards (CRYSTALS-Kyber, CRYSTALS-Dilithium) in 2024, and the financial-services sector is reportedly the first commercial vertical where mandatory PQC migration is being seriously planned.

SEALSQ’s positioning sits cleanly inside that timeline. The company, listed on Nasdaq under LAES, has built a portfolio of quantum-resistant HSMs and is currently constructing what it calls a fully integrated quantum vertical-sovereign stack through its Quantum Fund investments in carefully selected quantum-computing companies across the US and Europe.

The WeCan deal is the most concrete commercial test of whether the SEALSQ stack can actually be deployed inside operational banking systems rather than remaining a regulatory-aspiration story.

The European banking sector is visibly building out a multi-layered AI-compliance-and-security stack in parallel, with SEALSQ-WeCan focused on the cryptographic-integrity layer, Bayshore on the legal-rule layer, and the BNP-Mistral programme on the AI-attacker-defence layer.

What the SEALSQ-WeCan announcement leaves under-specified is timing. Neither company has disclosed a launch date for the compliance co-pilot or specified which of the named banks will be the first deployment customer.

The CHF 5m commitment, set against the cumulative engineering work the post-quantum-AI integration will require, is best read as Series-A-equivalent runway rather than the cost of the full build-out. Whether further capital follows from SEALSQ’s public-equity base or from outside investors is the next concrete question.

SEALSQ’s shares were modestly higher in pre-market trading on the announcement.