News

• Antares closes $470 million Series C, $370 million in equity and $100 million in debt, co-led by Paradigm and Caffeinated Capital
• The move follows the Mark-0 reaching criticality at Idaho National Laboratory and helps fund an electricity-producing Mark-1 in 2027 and military deployments in 2028
• Antares' military-first strategy targets a price-insensitive customer bound by an executive order requiring a reactor on a US base September 30 2028

Nuclear fission startup Antares has announced it has raised a total of $470 million in funding as it looks to supply a single client with a looming deadline: the US Military.

The Torrance, California company closed a Series C co-led by Paradigm and Caffeinated Capital, with Point72 Ventures, Shine Capital and Industrious Ventures among the participants, according to TechCrunch's look at its PitchBook data.

Antares' raise comes seven weeks after its Mark-0 microreactor reached criticality at Idaho National Laboratory under the Department of Energy's authorization in a demonstration run with the DOE, the lab, and BWX Technologies, with the US Army observing.

Antares calls it the "first privately developed non-light-water reactor to achieve criticality in the United States in more than four decades," underscoring how important the test was for both Antares and the US military.

The US military's presence is not by chance; It is mandated by an executive order ("Deploying Advanced Nuclear Reactor Technologies for National Security") signed by US President Donald Trump to ensure that it has an operating army-regulated nuclear reactor on a domestic military site by September 30, 2028.

Antares happens to be one of three finalists in the Pentagon’s Advanced Nuclear Power for Installations program that aims to test-deploy a small module reactor (SMR) on two different air force bases.

With Antares being the first private company to go through that gate under the DOE's Reactor Pilot Program, it plans to build an electricity-producing Mark-1 reactor in 2027, followed by initial deployments to defense customers by 2028.

The Mark-1 uses the same full-scale core, facility, and TRISO fuel as the Mark-0, which is the company's argument for why the 2027 date is credible rather than aspirational.

Antares CEO Jordan Bramble says the reactor is being engineered to run reliably for six-plus years deployed without refueling, at an output the company, while noting that firm contracts to build reactors and the capital raised would allow the company to build "the first microreactor producing useful electricity in a truly commercially viable design."

The claim might hold for the US military, but insights from financial advisory firm Lazard suggest the approach is more expensive for now than all but the most expensive gas turbines in operation, with renewables clocking in considerably cheaper.

More importantly, the US supply chain for advanced reactors remains immature; no company has yet to manufacture microreactors at a production scale, and a first-of-a-kind electricity-producing reactor may reveal problems that a zero-power core does not currently exhibit.

Half a billion dollars in capital buys Antares a serious attempt at the deadline. Whether it manages to buy a reactor running on a US base in time to meet the executive order remains to be seen for now, even as it remains committed to trying.

Spoilers for Stuart Fails to Save the Universe episode 2 ahead.

We've jumped into a completely different universe in Stuart Fails to Save the Universe episode 2, waving the Mad Max feel of the initial post-apocalyptic world goodbye (thankfully).

However, there's a new problem in the new HBO Max show. Stuart (Kevin Sussman), Denise (Lauren Lapkus), Kripke (John Ross Bowie) and Bert (Brian Posehn) have landed somewhere where everybody is incessantly nice to each other... a little too nice, if anything.

As Stuart quickly finds out, this is because they each have a device attached to their neck to prevent them from being mean or exhibiting any kind of bad behavior. If they try to fight it, they get tased from the inside, and eventually, shipped off to an AI correctional facility.

After years of rehabilitation (really, it's about two minutes), we find out that Penny (Kaley Cuoco) is leading the rebellion against the AI overlords, eventually helping the gang escape after she trusts them enough.

But while Penny is a Big Bang Theory cameo that fans have been waiting for, it's not the one that I think is the most interesting. For that, we need to look to the universe's AI overlord herself... film and TV's Christine Baranski.

Yes, that's right... Christine Baranski is the person in charge of this AI uprising. Via video, she explains to the gang what the neck devices are and why society needs them, giving creators Chuck Lorre, Zak Penn, and Bill Prady ample opportunity to criticize the growing use of AI in the entertainment industry.

TBBT fans will know that Baranski played Leonard's (Johnny Galecki) absent and standoffish mother Beverly in the original series. But in this universe, Beverly clearly doesn't exist.

"I think she was forced into doing it... not in real life, but in this universe," Penn tells me about Baranski's cameo. "In real life, she was totally willing. But in that universe, I don't think she had a choice but to record that message.

"Maybe it would be better if she was driving that universe... but she's not. She might end up getting accused of collaboration with the AI overlords, but I think that AI made her an offer that she couldn't refuse."

Prady adds, "We'll have to see how the rebellion goes and what happens afterwards. What do they do to her afterwards?"

"She'd be tried," Lorre agrees. "Well there we go — there's something we can do if we get a second season."

Lorre's comment suggests that this is a one-and-done appearance from Baranski in Stuart Fails to Save the Universe, but Penn's take on AI is particularly interesting in the week after fans accused the show's title sequence in episode 1 of being "AI slop."

the word slop is so overused today (like when referring to marvel movies) but this is absolute slop. literally ai slop. https://t.co/45cFGcMAsXJuly 26, 2026

Whether the titles were actually generated by AI is unconfirmed, but episode 2 suggests that the team clearly doesn't think the technology is a good thing.

As I tell the trio, if anybody can lead us through perilous technological times, I'd put all my faith in film and TV's Christine Baranski.

• Microsoft has won an extended stay at the Competition Appeal Tribunal while it seeks permission to appeal to the UK Supreme Court, having lost at both the Tribunal and the Court of Appeal
• Its argument that Office's icons and clip art placed it outside software exhaustion rules was rejected as producing "odd results", including CD software being resellable while identical downloads were not
• The jurisdiction ruling matters more than the £270 million claim, as it clears the way for a parallel collective action that could expose Microsoft to billions

Microsoft has bought itself time in its long-running fight with a British software reseller, securing an extended stay of proceedings at the Competition Appeal Tribunal while it prepares to seek permission from the UK Supreme Court to appeal.

While Microsoft has yet to appeal formally, its indication that it intends to do so buys it some breathing room against a judgment that could mean that parallel proceedings, including a £3.5 billion class action lawsuit that could compromise as many as 2.7 million members, could become a centerpiece of the software giant's future legal troubles in the UK.

The UK CAT (Competition Appeal Tribunal) has paused parts of the case as a result, and Chair Justin Turner KC noted that it was not yet known whether permission would be granted, and framed his decision as balancing the potential waste of costs if permission were granted and an appeal succeeded against the undesirability of delay in the proceedings.

Microsoft has been fighting this claim for years, dating back to 2021, when JJH Enterprises, a Derby-based reseller trading as ValueLicensing, sued Microsoft for £270 million.

The allegation is that Microsoft wrote terms into its agreements that offered customers better subscription pricing in exchange for surrendering the right to resell their perpetual licenses, thereby choking off the secondary market and pushing buyers toward Microsoft 365 even as the software giant denies wrongdoing.

ValueLicensing's case rests on the Court of Justice's 2012 decision in UsedSoft v Oracle, which held that a software vendor's distribution right is exhausted once it sells a perpetual license, including one delivered by download, so it cannot use copyright to block the onward sale of that copy.

Having contested the case as a competition matter for years, Microsoft changed its argument in 2025, recasting it as a copyright dispute. Office, it said, is not merely a computer program. It ships with icons, user interface elements, help files, and clip art, and those creative components should be governed by a different copyright regime in which exhaustion does not bite.

That argument has now failed twice. The Competition Appeal Tribunal (CAT) unanimously decided the preliminary copyright issues in ValueLicensing's favor in November 2025, and on July 7, 2026, the Court of Appeal dismissed both of Microsoft's appeals, one challenging the CAT's jurisdiction to rule on copyright questions at all, the other contesting the copyright findings themselves.

Microsoft's reading, the court observed, would produce "odd results": software supplied on a CD-ROM could be freely resold while the identical software delivered as a download could not, and a vendor could defeat established exhaustion principles simply by bundling some icons with its code.

At the same time, the CAT did not grant resellers unlimited freedom. Under UsedSoft, a buyer holding a single copy installed on a server and licensed for several users cannot carve off part of that user capacity for resale, because exhaustion attaches to the copy rather than the seat.

The Tribunal distinguished that from a genuine volume arrangement comprising multiple independent copies or licenses, in which reselling a corresponding number may be permissible once the seller has stopped using the copies it has passed on. That distinction sets real limits on how the secondary market can operate.

Microsoft has spent most of the period covered by the claim moving its customer base from perpetual licenses to Microsoft 365. If ValueLicensing wins outright, it wins the right to trade in a category that its opponent has been deliberately retiring for a decade. Should the Supreme Court decline to take the case, proceedings return to the Tribunal, with a trial mooted for 2027.

The jurisdiction ruling matters well beyond this claim. By confirming that the CAT can decide copyright questions where they are necessary to resolve a competition case, the Court of Appeal cleared a path for others, and The Register reported in April 2026 that Microsoft had viewed that appeal as a complete answer to Alexander Wolfson's collective action, on the basis that copyright issues could not be transferred to the High Court in a class action.

Wolfson was granted permission to intervene in support of ValueLicensing, and his claim, which could expose Microsoft to liability in the billions, now stands to benefit from the ruling.

Britain's regulators have taken an interest in the same broad shift, too. On July 29 2026, the Competition and Markets Authority opened a consumer protection investigation into how Microsoft marketed its Microsoft 365 Personal and Family plans.

From January 2025, existing subscribers were given Copilot and other new features at no extra cost for the remainder of their term, then automatically moved onto a higher-priced plan at renewal unless they chose an alternative or canceled.

A time-limited "Classic" plan preserved the old feature set at the old price, and for annual subscribers the Copilot version cost £25 more per year.

"When a business changes its subscription plans, customers need clear and timely information about their options," said Hayley Fletcher, senior director for consumer protection at the CMA. "Our investigation will consider whether Microsoft customers were misled and ended up paying more as a result."

Microsoft said consumer trust is a priority and that it is reviewing the claims while cooperating with the inquiry. The CMA has reached no conclusions and has not accused the company of breaking the law.

That probe concerns consumer subscriptions rather than the commercial agreements at issue in the ValueLicensing case, so the two are separate matters under different regimes. What links them is the same underlying transition from one-off purchases to recurring ones.

Microsoft already faces a CMA strategic market status investigation into its business software, and Australia's competition regulator is suing the company in the Federal Court over the same Copilot and Classic plan issue, affecting an estimated 2.7 million customers. Whichever way the Supreme Court application goes, the legal scrutiny of how Microsoft moved the world onto subscriptions is not ending soon.

The scrutiny also lands in a fraught transatlantic moment. On July 24, days after the European Commission fined Google $1 billion under the Digital Markets Act, Trump announced a Section 301 investigation into EU trade practices, declaring that the United States is not a "PIGGYBANK" for Europe and promising a substantial tariff.

Brussels has consistently maintained that regulating economic activity within the bloc is its sovereign right, and that its digital rules apply to all firms operating there. Britain sits awkwardly beside that fight. Its courts are no longer bound by EU case law, and its regulators answer to nobody in Brussels, but the president's stated targets have been digital taxes and digital markets rules in general rather than the EU alone, and the UK has both.

Spider-Man: Brand New Day has finally swung into theaters worldwide — and I bet you've got questions about Tom Holland's latest Marvel movie.

Haven't seen it yet? Read my Spider-Man: Brand New Day review for a spoiler-light insight into the newly released film. Then, bookmark this page, go and see it, and come back to carry on reading this article, because I'm not kidding when I say full spoilers immediately follow for Spider-Man: Brand New Day.

Of course, if you have watched it, by all means scroll on — it's worth the read, I promise!

Yes. One of the worst-kept secrets in Hollywood is finally confirmed about 70 minutes into Spider-Man 4: Sink is playing the Marvel Cinematic Universe's (MCU) version of iconic X-Men member Jean Grey.

It's a reveal that's been a long time coming. Sink was first rumored to appear in Spider-Man: Brand New Day last March, so we've had to wait 16 months to learn which Marvel character she was actually playing.

Of course, there's been all manner of speculation about the identity of her character, with everyone from Mayday Parker — aka Peter Parker and Mary Jane Watson's daughter in the comics — and Typhoid Mary to Firestar and Shathra mentioned as potential candidates.

Now that we know Sink is playing Grey, she's a shoo-in to appear in Marvel's new X-Men movie (more on this later). Her next outing as Jean Grey is expected to be in Avengers: Secret Wars, though, so we'll see her again when that film arrives next December.

That'll be William Metzger, aka the head of the Department of Damage Control (DoDC).

Metzger is played by Tramell Tillman, the Severance star signing on to appear in Spider-Man: Brand New Day last August. It's unclear when or where he'll appear in the MCU next, but I have my suspicions about his next appearance, which I cover later in this article.

Anyway, Metzger is revealed to be the Marvel Phase 6 flick's actual villain in its third act. Initially positioning himself and the DoDC as an ally of Spider-Man's, we eventually learn that Metzger was the one who kidnapped, experimented on, and ultimately killed Jean's older sister Sarah. A fellow telepath, Sarah was taken to the DoDC's main testing facility on Staten Island, where she was used as a lab rat for the company's investigations into new tech powered by superhuman abilities.

Of course, as soon as Spidey learns the truth, our friendly neighborhood hero sets out to rescue Jean, whom he naively helped the DoDC to capture earlier in Brand New Day. Little does Peter Parker know, though, that it could've been his last mission...

No, but he comes mightily close to losing his life.

After he infiltrates the DoDC's Staten Island headquarters, Spidey uses his upgraded powers to defeat members of The Hand controlled by an antagonistic Jean — remember, she still thinks Spider-Man is the enemy at this point.

Eventually, Peter tracks down Jean, who's holding Metzger hostage. Metzger tried to experiment on Jean as he did with Sarah, but the test not only goes awry, it unlocks Jean's full powerset. No longer limited by the 33-foot radius that her telepathic abilities had, Jean utilizes her powers to paralyze a large part of New York's human population, and threatens Peter that she'll hurt them all — including his former girlfriend M.J. and best friend Ned, who are caught in her paralyzing telepathic wave — if he tries to stop her from enacting revenge on Metzger.

Being the good guy that he is, Peter manages to talk her down. He does so by allowing the souped-up Jean to enter his mind, where he shows her a very specific memory of Aunt May that helps Jean understand that, despite Sarah's death, she's not alone.

Happy ending, right? Not exactly. Like Spider-Man, Frank Castle, aka The Punisher, sprang into action when Jean used her upgraded powers on New York's residents. Unable to reach DoDC HQ, he heads to a nearby rooftop outside of the impacted zone. There, he uses his marksmanship skills to take a shot at Jean with a heavy-duty sniper rifle.

Sensing the incoming bullet, Spidey shields Jean from the sniper round and is critically injured by it. As he bleeds out, Jean employs her powers to keep him alive before Frank races him to a nearby hospital to undergo life-saving surgery.

Some time later, Peter wakes from his coma. After a heart-to-heart — and some light banter — with an apologetic and remorseful Frank, Peter checks himself out of the hospital.

A crowd of dedicated Spider-Man fans, who've congregated outside the hospital to hold a vigil for his swift recovery, awaits him. Fortunately, Frank stays behind, dons the Spider-Man mask, and waves to the assembled crowd from Spidey's hospital room, allowing a profoundly moved Peter to walk away without anyone knowing he's the real wallcrawler.

No — she doesn't need to, because the pair end up finding out that Peter is Spider-Man.

M.J. learns who he is when she discovers the note Peter had written back in 2021's Spider-Man: No Way Home. Meanwhile, Ned appears to remember Peter when the duo performs their secret handshake in Brand New Day's final scene. Ultimately, then, Jean's possible intervention wasn't necessary.

We already knew the vast majority of Spider-Man: Brand Day's cast heading into the film, but the MCU's latest movie has a couple of surprises up its sleeve.

For one, Florence Pugh's Yelena Belova makes an unexpected yet scene-stealing cameo. Spider-Man meets her at a New York bathhouse to obtain information about the attack on the DoDC in the film's first act. Later on, she appears once more when the eponymous hero starts to suspect Metzger isn't all that he seems and seeks details on Jean's sister Sarah.

The other big cameo is Marisa Tomei's Aunt May. Peter's surrogate mom, who died in No Way Home, appears by way of the memory I mentioned earlier in this article.

"But what about Daredevil?", you might be asking. Given what happened in the final episode of Daredevil: Born Again season 2, some fans had expected Spidey to come across Charlie Cox's Matt Murdock and join forces to take on The Hand in Spider-Man 4's prison-set action sequence. Unfortunately, Cox doesn't show up in Brand New Day, so the wait for a Daredevil and Spider-Man MCU team-up goes on.

Yes. Right at the end of the film's final credits, the Spidey Tracker — an app built by Ned Leeds to, well, track Spider-Man's whereabouts — flashes up on the screen.

As an automated message that says "Location unknown" starts to play over and over, the app zooms out from planet Earth and eventually pinpoints that the webslinger is somehow in space.

What does this mean? I have a wild theory that Spider-Man: Brand New Day's post-credits scene might not mean what we think it does, but the most likely scenario is that Peter Parker has been transported somewhere due to events that'll go down in Avengers: Doomsday or Avengers: Secret Wars. Speaking of which...

In short: I don't know. Spider-Man 4's end credits scene could be related to either of the aforementioned movies, so until Avengers: Doomsday is released this December, your guess is as good as mine.

That said, at this moment in time, Tom Holland isn't slated to appear in Doomsday, and that film is already packed with enough characters as it is. You can find out who'll appear and more by finding out everything we know so far about Avengers: Doomsday.

If Holland doesn't show up in Avengers 5, he'll be a dead cert to star in its sequel. Marvel and Sony won't miss the opportunity to put one of their most bankable characters in at least one of the next two Avengers movies, so the smart money is on him having a substantial role to play in Avengers 6. For more on that flick, find out what we know so far about Avengers: Secret Wars.

It doesn't. With Doomsday and Secret Wars set to be the next two MCU movies to hit theaters, that pair is far more likely to impact the MCU's first mutant-based film.

Nevertheless, Spider-Man 4's final minutes do show Jean Grey traveling on a bus to an unknown location. Maybe she's attempting to track down more of her fellow mutants? If she is and she's successful, it's possible that whoever she locates will not only have roles to play in Secret Wars, but also form a prototype version of the X-Men ahead of their actual founding/formation in their first dedicated MCU adventure.

While we wait for more news on one of the few confirmed post-Multiverse Saga projects, find out what we know so far about Marvel's new X-Men movie.

• Post-quantum encryption is now live on Norton VPN's Mimic and WireGuard
• Rebuilt iOS and Android apps have slashed connect times by more than 50%
• Norton has added 12 new countries across Asia and Africa

Competition for the title of best VPN is fiercer than ever, and Norton is making a strong case for the crown. The cybersecurity veteran has just unveiled the final releases from its first-half 2026 roadmap, delivering a robust mix of future-proofed privacy tools, significantly faster mobile applications, and an expanded global footprint.

The standout feature of this latest rollout is the introduction of post-quantum encryption. This comes alongside improved Software Development Kits (SDKs) for its mobile apps and an expanded server network across Asia and Africa.

At first glance, these releases may appear to be separate features. But for the provider, they all address the same need — customers must count on Norton VPN, wherever they are.

As Himmat Bains, Product Lead at Norton VPN, told TechRadar, post-quantum encryption is crucial to protect users' data from future threats. Faster, more reliable mobile connections mean VPN protection is there exactly when they need it. And coverage in 12 new countries means that more people can connect locally.

"Different problems, same answer — we want this to be the VPN you can simply trust and forget about," said Bains.

Norton VPN – best for beginners
The cybersecurity company's latest upgrades mean it's now challenging some of the best VPNs on the market. It offers great speeds, strong unblocking and straightforward apps, but some of its features are still quite basic. Subscriptions start from the equivalent of $3.33 per month and are all protected by a 30-day money-back guarantee.View Deal

As quantum computing advances, cybersecurity experts have increasingly warned of "harvest now, decrypt later" attacks. This is a tactic where cybercriminals or state actors steal encrypted data today, hoarding it with the intention of cracking it when quantum computers become powerful enough to break traditional encryption standards.

To combat this long-term risk, Norton VPN has upgraded its proprietary Mimic protocol to utilize the ML-KEM-1024 encryption algorithm by default across all platforms.

The company is also rolling out the exact same post-quantum protection to the popular WireGuard protocol on both desktop and mobile devices, ensuring that user traffic remains fully secure against tomorrow's technological threats.

While future-proofing encryption is crucial for long-term privacy, everyday users will immediately feel the benefits of Norton’s revamped mobile experience. The provider has entirely rebuilt its iOS and Android SDKs, the underlying code architecture that powers its mobile applications, from the ground up.

The performance gains are impressive. According to Norton's release notes, the new mobile architecture reduces connection times by more than 50% and connection errors by over 60%. For mobile users who constantly hop between home Wi-Fi, public hotspots, and cellular networks, this means a much smoother, almost instantaneous protective shield.

Just as importantly, this rebuilt framework will allow Norton to ship bug fixes and new features much faster going forward. The overhaul has already paid immediate dividends, leading directly to a rebuilt mobile widget that patches previous usability issues users had been facing.

"

Beyond the apps themselves, Norton is continuing to scale up its physical infrastructure. The provider has added 12 more countries to its network, specifically targeting underserved regions across Asia and Africa.

This expansion brings Norton's total reach to an impressive 102 countries and more than 140 global locations, seamlessly building on the momentum of previous network upgrades earlier this year.

This latest announcement caps off a remarkably busy start to the year for the Norton VPN team.

The provider brought a native Amazon Fire TV app (Fire OS 8 and newer) in March and introduced the industry's first AI-native VPN for Agents in April. It also successfully added Split Tunnelling on Mac in June, joining its Android and Windows counterparts.

Furthermore, WireGuard is now live across Mac, iOS, and tvOS, with a new Pause VPN feature on iOS closing the platform's last functional gap. The company also rolled out 25 Gbps servers across busy cities and added Manual IP Rotation.

"Our goal is simple: deliver a VPN for the masses," Bains told TechRadar, arguing that these releases weren't about shipping a longer list; "it was about closing the gaps customers actually feel."

Looking ahead, Bains confirmed that Norton's second-half roadmap pushes even further, noting that "we'll have third-party audits for the third year launching later in the summer, along with meaningful updates coming to our streaming and performance.

The telecoms industry is all a flutter.

Satellite connectivity was once seen as useful but peripheral. Starlink is changing that.

With scale, money and ambition behind it, the company is starting to look less like a niche rural broadband provider and more like a serious question for the wider connectivity market.

Kester Mann, Director, Consumer and Connectivity at FDM CCS Insight, said this is one of the more interesting developments of 2026.

“The industry is nervous. Operators are currently working to shore up their defenses against potential offerings from a company like Starlink, but the long-term plan is not obvious yet. Starlink has the financial muscle to be disruptive if it wants to be.”

Starlink is clearly a very ambitious business. The recent IPO of its parent company, SpaceX, shattered global financial records, raising $85.7 billion in total capital and valuing the company at $1.77 trillion at listing.

Just weeks after its IPO, the company completed a $25 billion bond sale. One thing it is not short of is money.

Clearcut data is somewhat difficult to come across, but the Department of Agriculture puts the total U.S. population in nonmetro counties stood at 46.2 million, accounting for 14% of the total population spread across 72% of the landmass.

This demonstrates why closing not spots in the U.S. has become such a significant challenge.

“It’s very difficult to make connectivity economical in the U.S. outside urban areas,” Mann added. “I understand why we have this situation, and partnerships between telecoms operators and satellite companies make a huge amount of sense.”

The telecoms industry has often envisioned a patchwork landscape for connectivity, with each technology serving the purpose to which it best suits. This works today, mobile, Wi-Fi and broadband interlink nicely, and satellite could fill the spaces left behind by traditional telecoms. Namely, the rural locations.

The economics of telecoms infrastructure have always created a headache for the industry, especially in countries like the U.S.

Simply put, there aren’t enough people living in an area to justify the vast expense of network rollout. If the telecoms operators could have made money, you can bet there wouldn’t be not spots.

This is where a ‘non-telecoms’ solution fits nicely, and there is precedent of new market entrants completely disrupted highly established markets.

Netflix turned entertainment on its head. AirBnB reshaped the hotel industry without ever building a hotel. Uber led to an entirely different way of moving from A to B, and not just taxis anymore.

In the rural states, there are consumer mobile and broadband services to potentially serve, IOT networks to be connected, agriculture and other primary industries to power, not to mention public services just as there are in the cities.

There is a market, just not a financially attractive one. Telecoms operators would have to spend eye-watering amount to cover vast and challenging terrains. It simply doesn’t add up.

Traditional approaches won’t work, so something new has to be considered.

But how much momentum could this give Starlink?

Starlink boss Elon Musk said in a recent interview he had no intention of putting the telecoms operators out of business, but that has not put all the questions to bed.

Mann continued, “Is rural enough for Starlink? It is a niche service, and Musk has big ambitions. Some analysts are talking about a takeover of a telecoms operators, which is a long shot, but a MVNO proposition could certainly be realistic.”

This is the precarious position telecoms operator find themselves in. Partnerships with satellite companies are attractive, but then it comes with risk.

In the UK, Vodafone and AST SpaceMobile are partnering to bring direct-to-device (D2D) space-based mobile broadband to standard 4G/5G smartphones. This looks attractive to the telecoms operator.

Then you have the middling risk. When Rocket Labs announced plans to acquire Iridium, the deal also included the company’s globally harmonized L-band spectrum and low Earth orbit (LEO) satellite network. This makes Rocket Labs a direct competitor in the rural regions.

Starlink is in a similar position to Rocket Labs, just a couple of factors greater. It has the same operational capabilities and spectrum to deliver consumer services at scale, but it also has cash. Whereas Rocket Labs could be a niche competitor few would worry about, Starlink could be a major disruption.

Everyone is focused on the disruption to the telecoms industry, but ultimately this might be a by-product of the bigger plan. What if all this routes back to Telsa?

The obvious answer is telecoms. Starlink wants to sell connectivity, operators sell connectivity, so the threat is easy to understand.

But that might also be too narrow.

Musk’s businesses rarely sit neatly inside one industry. SpaceX launches rockets. Starlink sells broadband. Tesla sells vehicles. xAI builds artificial intelligence. X provides the distribution layer. On paper, these are separate businesses. In practice, they increasingly point towards the same idea.

Controlled infrastructure, controlled data, controlled customer relationships.

That is where Tesla becomes interesting.

A company trying to make autonomous vehicles work at scale does not just need batteries, sensors and software. It needs reliable connectivity in as many places as possible. Robotaxis, logistics, remote monitoring, over-the-air updates, live mapping and future machine-to-machine services all become more valuable if the connectivity layer is controlled rather than rented.

This does not mean Starlink is simply a Tesla project in disguise, that would be too neat, and probably too generous. But it does show why telecoms operators should be careful about judging Starlink by traditional telecoms logic.

They are asking whether Starlink can steal enough broadband or mobile customers to matter.

Musk may be asking a different question entirely. What becomes possible when connectivity, devices, data, AI and mobility sit inside the same orbit?

That is the uncomfortable bit for operators. They still see connectivity as the product. Starlink may increasingly see it as the operating layer.

Or perhaps there is no grand plan at all. This is still the man who bought Twitter for $44 billion. But telecoms operators cannot afford to treat that as reassurance. Sometimes the chaos is the strategy.

We've tested over 50 portable power stations.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The UK’s newly announced £2 billion quantum investment package signals something more consequential than the years-long research race that I’ve patiently been following.

Quantum is increasingly being treated as a key part of the future of national infrastructure.

While the size is important, the clearest sign of that shift is the structure behind it.

Through the newly announced “ProQure: Scaling UK Quantum Computing” initiative, the UK has become the first country to commit to an advanced procurement program for large-scale quantum computers by the early 2030s.

That matters because governments do not produce technologies at scale unless they believe those technologies are strategically important to operational capability, economic resilience and long-term competitiveness.

In many ways, this marks quantum computing’s transition from the research era into the infrastructure era.

So why does a procurement program matter more than the headline funding figure? The significance extends beyond public investment totals. Procurement creates market confidence. It helps establish deployment pathways, support supply chains, attracts private capital and gives emerging technologies a route from laboratory experimentation into practical use. In deep technology markets, the challenge is often building the ecosystem required to scale and operationalize it.

The challenge is becoming increasingly important as the broader quantum market matures. QuEra’s recent Quantum Readiness Report found that organizations are becoming far more disciplined in how they approach quantum investment. While 44% of respondents still expect their quantum budgets to increase in 2026, nearly half expect spending to remain flat, signaling a shift away from hype-driven expansion toward proof-driven decision-making.

More advanced adopters of quantum technology are increasingly motivated by what the industry refers to as the “classical wall”. This is the point at which conventional computing approaches struggle to solve increasingly complex computational problems. That shift from speculative experimentation toward problem-driven investment is an important sign of market maturation.

The changing investment landscape also helps explain why procurement and national strategy are becoming more central to the conversation.

One of the most striking findings in our research was that 62% of organizations now actively factor sovereignty into procurement decisions, while only 5% say sovereignty is not a consideration at all.

However, the highly globalized sourcing model that characterized the industry’s early years is giving way to a more regionally strategic one, shaped by concerns around resilience, supply chains, infrastructure control and long-term access to advanced computing capability. This makes the UK’s strategy particularly timely.

The government’s emphasis on sovereign capability, domestic ecosystem development, and long-term infrastructure investment reflects broader shifts already underway across the market.

Quantum computing is increasingly being viewed not only as a commercial technology opportunity, but also as a strategic national capability alongside AI, semiconductors, and advanced communications infrastructure.

Scaling quantum computing requires much more than building processors. It requires coordinated investment across software, networking, manufacturing, talent, infrastructure, and integration with classical systems.

The UK’s decision to support not only computing, but also sensing, networking, software labs, infrastructure, and skills programs reflects an understanding that leadership in quantum will be determined by ecosystem depth as much as scientific breakthroughs.

The talent challenge illustrates this clearly. Our report identified specialist workforce shortages as one of the most significant barriers to adoption, particularly in highly technical areas such as quantum error correction.

Developing scalable quantum systems requires expertise that remains globally scarce, making long-term investment in education, research institutions, and industry collaboration essential.

The UK enters this next phase from a position of genuine strength. Over the past decade, the National Quantum Technologies Programme has helped establish one of the world’s most advanced quantum ecosystems, supporting research hubs, startups, national infrastructure, and industry partnerships.

QuEra’s own presence at Harwell reflects the depth of scientific expertise and collaborative capability that already exists across the UK quantum landscape.

This investment builds on that foundation while accelerating the UK’s ability to translate quantum promise into economic opportunity. QuEra is one of several companies building next-generation quantum computers at Harwell, drawing on the country’s world-class talent and research base.

The next challenge is translating scientific leadership into scalable deployment.

Quantum computing remains a technically demanding, pre-commercial industry in many respects, but the direction of travel is becoming clearer. To be clear, quantum computers still cannot deliver broad commercial advantage on their own in most industries; that case is still being proven sector by sector.

But the procurement and infrastructure being built now is what will let that advantage arrive faster once it does. Governments, enterprises and investors are increasingly focusing less on theoretical promise and more on deployment readiness, operational capability, and long-term strategic value.

Public sector leadership will continue to play an essential role in that process. Our research found that government funding remains the leading driver of quantum investment globally, cited by 28% of respondents as the primary catalyst for budget increases. At the same time, government and defense organizations are expected to lead near-term commercialization efforts, highlighting the central role public investment continues to play in scaling the industry.

The countries that lead the next phase of quantum computing are likely to be the ones that successfully combine research excellence with procurement, infrastructure, talent development, industrial coordination, and commercial scale.

For enterprises and other governments watching from the sidelines, the practical lesson is straightforward. Building procurement pipelines, workforce training, and proof-of-concept programs now, even at modest scale, is what prepares an organization to scale up once quantum systems clear the classical wall in a given domain. Those that engage early will be better positioned than those that wait for certainty that may not arrive on a convenient schedule.

The UK’s latest announcement suggests it understands that distinction and that may ultimately matter more than the £2 billion headline itself.

We've tested, reviewed, and ranked the best business computers.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The conversation about AI and shopping is happening in the wrong place.

Everyone wants to know whether agents will browse for us, compare prices for us and, eventually, buy for us without asking. It's a fair question. But it’s not the one that matters right now.

The more consequential shift is happening somewhere far less visible: inside payment networks, authorization protocols and the systems that decide whether software is allowed to spend money on our behalf.

The technical capability for an AI agent to complete a purchase has existed for some time. What has been missing is the infrastructure that allows businesses and consumers to trust those transactions: confirming an agent had permission to act, ensuring it stayed within agreed boundaries, and creating a verifiable record if something goes wrong.

Over the past few months, that infrastructure has started to emerge. The biggest moves in this space haven't come from AI labs building increasingly showy assistants. They've come from the companies that move money.

At its annual conference, for example, Stripe used the opportunity to introduce wallets that agents can spend from, alongside a partnership letting businesses sell directly inside AI search and chat environments.

Visa introduced an early protocol to help merchants tell a legitimate AI agent apart from a bot. Mastercard announced its own framework for registering and authenticating agents before a transaction is allowed to proceed. Agents are already showing up at volume. What's been missing is the infrastructure to handle what happens next.

Adobe research from early this year shows that AI traffic to US retail sites grew 393% YoY in the first quarter, with AI-referred visitors converting 42% better than non-AI traffic. Just a year before, that same traffic was converting 38% below standard channels.

That’s why the language from the payments industry has shifted so noticeably toward authorization, intent and trust. Mastercard's own leadership has made things explicit: how do you distinguish a legitimate agent from a malicious one? How do you confirm the consumer really authorized the purchase? How do you verify that the agent did what it was told?

Those aren't AI questions. They're identity and trust questions that just happen to involve AI.

What's notable about the recent announcements is how consistent the underlying logic is across companies that otherwise compete directly. Visa's protocol and Mastercard's framework work the same way in principle: register the agent, verify intent and pass credentials only once trust is established.

Stripe's approach, issuing single-use payment credentials per task rather than exposing a customer's card details to an agent, solves the same problem from a different angle.

The protocols may look like back-office plumbing. Functionally, they're architecture for autonomous commerce. Every time a payments provider decides what counts as a verified agent, what data has to be passed to prove intent or how disputes get resolved, it's writing rules that shape who gets to participate and on what terms.

A consumer can be persuaded by an AI assistant to want something, but if the transaction can only execute through a specific track, with specific verification steps, then influence over the purchase decision and control over the purchase itself become two separate things. The company that captures the second captures the economics, regardless of who captured the first.

For merchants, this creates a concrete operational problem. If a transaction gets disputed, can the business prove what was authorized, by whom and under what constraints? These are the exact questions Visa, Mastercard and Stripe are quickly trying to answer; whoever answers them first effectively writes the rulebook everyone else follows.

Autonomous shopping is imminent, and in some categories it’ll come faster than in others. Routine reordering and B2B procurement are likely to happen first. Other more considered, emotionally fueled purchases will take longer; trust has to be earned. Payment readiness is only part of it, however.

The commerce stack underneath, covering product data, pricing, inventory and checkout, needs to be structured and accessible enough for agents to actually operate it. Businesses that haven't addressed that layer will find the payments question is the least of their problems.

The infrastructure being built right now by payment providers won't just process agent-driven transactions. It will determine which businesses are visible to agents in the first place, which transactions are trusted by default and which get pushed into friction and manual review. The real battle isn’t over autonomy. It’s over who owns the infrastructure of trust.

We've featured the best ecommerce platform.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

The Nintendo Switch was nothing short of a phenomenon. It won the hearts of millions, possesses one of the greatest games libraries in console history, and became one of the highest selling systems of all time. It brought the best parts of its failing predecessor, the Wii U, forward, assuming an ultra-convenient hybrid form, and during the COVID 19 pandemic, it felt like everyone and their dog were building their perfect island on Animal Crossing: New Horizons, or racing to the finish line in Mario Kart 8 Deluxe.

All of that is to say, that whatever came next would have huge shoes to fill. And so, in came the Nintendo Switch 2 — a system that appeared to be less of a revolution, and more of an evolution. It took what made the Switch great and ran with it, offering higher resolution and refresh rates, a larger screen for handheld mode, and even a mouse mode.

And in some respects, the Switch 2 has hit the ground running. It’s the second-fastest selling games console in recorded US history, and at just over a year old it's already seen multiple exclusives (such as Donkey Kong Bananza and Pokémon Pokopia) release to critical acclaim. But at the same time, some have been disappointed over the lack of a 3D Mario or truly new Zelda title on the horizon, while issues around battery life and pricing have also proved to be a turn off for others.

So, this got me thinking. Would gamers consider the Switch 2 to be more of a hit, or a flop? Sure, its sales figures have been impressive, but are users satisfied with the hardware upgrades? What are their thoughts on its game catalogue?

I spoke with industry experts, fellow journalists — and most importantly, you, our wonderful readers, and it proved to be a fascinating experience. I heard a lot of varying viewpoints, and now I’m here to report back.

To begin with, I spoke with Mat Piscatella, Executive Director of Games at the Market Research firm Circana, who had some excellent insight into the commercial performance of the Nintendo Switch 2.

“Well, over its first 12 months in the market, Nintendo Switch 2 hit 5.9M units sold in the US,” Piscatella told me.

Become a TechRadar Insider by simply clicking 'Join Now' at the top of this page. Have a question? Please email membership@techradar.com

This, he noted, ensured it became the second fastest-selling video game hardware in the US since Circana’s records began in 1995 — finishing behind the Game Boy Advance only. “If that's not considered a hit, I'm not sure what would be,” he remarked.

“For the 2026 year-to-date period ending May, US hardware spending in total was up 37% compared to the same period a year ago,” Piscatella said. He stated that this was “entirely driven by Nintendo Switch 2 as double-digit percentage sales declines were experienced across the other in market platforms”.

He also highlighted high sales of software, such as Donkey Kong Bananza, Metroid Prime 4: Beyond, Pokémon Legends: Z-A, Pokémon: Pokopia, and Mario Kart World, a game which was largely obtained via a bundle with the system.

With one year in the books, it would be hard to ask for much more when it comes to the sales performance of the Nintendo Switch 2

Mat Piscatella, Executive Director of Games at Circana

So far so good, then, for Piscatella, who also highlighted the benefit of the Switch 2’s backwards compatibility — giving users access to thousands of unique titles from the original system.

He did flag that June date may indicate “tough year-on-year comp for the system,” largely due to sales being compared to the launch month. “After that we should see Switch 2 continue to perform well into its announced September price increase. Then we'll have to see what happens."

The early signs have been promising, commercially speaking, but our Managing Editor of Social & Engagement (and Nintendo enthusiast), Josephine Watson, says that “the Switch 2 needed more than sales to break Nintendo’s generational curse of yo-yo-ing console success. It needed sticking power”. So, one year on, has it stuck the landing?

She argued that the Switch 2’s weaker holiday sales and subsequent production cuts suggest that “the console has yet to prove its mettle in the same way the first-generation [Switch] did, even with such high sales figures in its first year.” Despite this, she did acknowledge the mid-lifecycle boom that the OG Switch experienced, largely due to the release of Animal Crossing: New Horizons during the pandemic.

In terms of games, Watson had a somewhat mixed take, arguing that the Switch 2 had “a concerningly hollow launch title line-up carried mostly by Mario Kart World,” though she did suggest that the system gained momentum moving into the new year, especially with titles like Pokémon Pokopia and strong third-party additions like Resident Evil Requiem.

Is the Switch 2 a success one year on for Watson? “No — but yes, kinda,” she said. She lamented the lackluster battery life and pointed to how Nintendo could’ve done more to eradicate stick drift from the Joy-Con 2 controllers, but did express satisfaction with the improved processing power and refined eShop. And even though her views on the game library haven’t been entirely positive, the Legend of Zelda: Ocarina of Time remake, scheduled for release in 2026, “is the game-changer,” she said.

TechRadar Senior Writer, Hamish Hector, admitted that he was “a little skeptical” of the Nintendo Switch 2 — both a little before and a little after its launch. But since grabbing one in the latter stages of 2025, those worries seem to have largely dissipated.

“It has cemented itself as my go-to gaming handheld,” Hector told me. “While it doesn’t have the raw power or library breadth of my Asus ROG Xbox Ally X — a handheld gaming PC with access to Steam’s vast suite of games — the Switch 2 does still boast a solid selection of AAA ports, incredible first-party hits and superb indie darlings”.

He noted how some AAA third-party ports have varied in quality, although hit the mark more often than they miss, and most issues manifest themselves as simple “graphical performance hitches” rather than “anything majorly frustrating”. He even hailed the Switch 2’s value for money, emphasizing that “At $449.99 / £395.99 it’s much more affordable than any worthwhile PC handheld, and even after its upcoming price hike to $499.99 it’ll still be a solid pickup”.

We’ve heard a range of views so far, albeit generally positive ones. It feels as if the general consensus is that the Nintendo Switch 2 has been an overall success, with a strong start in terms of sales and a commendable library to match. But I was most excited to hear from were you, our audience. So, I asked, and you answered: here’s whether you considered the Switch 2 to be a hit or a flop.

I asked our audience about their feelings on the Switch 2 across multiple platforms. With our poll here on the TechRadar website and over on WhatsApp, users were given four choices: 'It’s a hit'; 'it’s pretty good'; 'it’s a little disappointing'; and 'it’s a flop'. We got 892 votes, and the breakdown was as follows.

Response

Votes (raw numbers)

Votes (as % of total)

It’s a hit

224

25.1%

It’s pretty good

197

22.1%

It’s a little disappointing

378

42.4%

It’s a flop

93

10.4%

Over on YouTube, we wanted to try a more direct approach and gave our audience only two options: hit or flop. We received 446 votes (at the time of writing), with 61% selecting ‘hit’, and 39% going with ‘flop’.

But what do these results actually tell us? At a glance, the picture seems pretty divided. Across all of our polls on our website, WhatsApp, and YouTube, we received 1,338 responses. Of those, 693 (51.8%) were generally positive on the Switch 2 (respondents called it ‘a hit’ or ‘pretty good’), whereas 645 (48.2%) were generally more negative (respondents either selected ‘it’s a little disappointing’ or ‘it’s a flop’).

Over on our Facebook page, the reasons for such division became clear, with a wide range of differing opinions floating around.

Of course, the very meaning of ‘hit or flop’ in itself can be interpreted in a number of ways. One commenter, Andreas Carmblad, stated that “hit or flop is reflected in the sales figures,” suggesting that Nintendo’s new console has been a success so far.

Others were more focused on the quality of the system’s software lineup, with most talking up the first-party line-up — though one commenter described the current set of exclusives as “woeful,” adding “hopefully a few big releases can turn it around [like] OoT and a new 3D Mario game”.

Elsewhere, commenters expressed disappointment over the console’s battery life in handheld mode — understandable given that it can run flat in around 2 hours with more demanding games.

However, many considered the new hardware changes to be a hit, with Tim McDougall saying “It's an improvement in almost every way [over the original Switch] and it's now much closer in performance to the current gen home consoles. It's 99% backwards compatible with the first Switch, with many performance benefits. I'd say it's a success”.

I was slightly surprised to see so many members of our poll select ‘slightly’ disappointed in our poll, and I expect that this comes down to a sentiment that the console is still a few games short of being a true must-have for all gamers. However, with Zelda and Pokémon games releasing soon, a lot of opinions could be swayed — and I suspect a blockbuster Mario title would only seal the deal.

What I was pleased to see, however, is that very few members of our audience (10.4% in our main poll) consider the Switch 2 to be an outright flop. After all, the sales speak for themselves, and there have been a lot of very well received exclusives early in the system’s lifecycle.

This brings me onto my personal thoughts. I pre-ordered the Nintendo Switch 2, and my expectations were high for the new system. Although I loved the original Switch’s library and appreciated its flexibility, the truth is, I actually didn’t love it all that much. For me, quality often felt overly compromised in handheld mode, with a lot of intensive titles struggling to reach up to 60fps. But the Switch 2 changed everything.

Now, I can enjoy titles in full HD at 60fps (and sometimes up to 120fps) in handheld mode, and I feel like the gap in quality when playing on the go — rather than in front of the TV — has shortened. What’s more, the Switch 2’s Handheld Mode Boost means that I can enjoy my original Switch games as if I was playing them with my system docked — a truly fantastic feature.

Then there’s the games. Controversially, I consider Mario Kart World to be the series’ best entry since the Wii generation, with a return to the true chaos and zaniness that made the games great in the first place — with knockout mode and the bizarre cast of characters being highlights for me. I’ve also had a blast with DK and Pokémon Pokopia, have spent countless hours racing in Kirby Air Riders, and exploring in Yoshi and the Mysterious Book.

The Switch 2 has also managed to do what the original model never could for me, in that I now often choose Nintendo for third party experiences. The brilliant combination of portability and higher fidelity visuals has made grinding in Dragon Quest I & II HD-2D Remake or playing detective in Raidou Remastered: The Mystery of the Soulless Army feel free of compromise.

Pair that with impressive sales performance, more convenient magnetic Joy-Cons, and even some pretty impressive virtual surround sound processing, and the Switch 2, in my eyes, is a sure-fire hit.

Yes, I share many users’ battery life gripes, and I personally believe that some of the first party games are far too pricey — I mean, $79.99 / £74.99 / AU$119.95 for a physical copy of Mario Kart World? Really? But even with incoming price hikes, I’d consider Nintendo’s latest hardware to offer plenty of bang for your buck, and although I’d love to get Super Mario Odyssey 2 within the next year or so, I’m perfectly satisfied with the current, and upcoming crop of titles at this stage.