There’s something off in the audiophile world right now, and it’s not just coming from Denmark. Between audiophile media excess that feels increasingly detached from reality, a long overdue Qobuz CarPlay update that finally fixes a daily annoyance, and a reminder from Wes Montgomery that timeless music outlasts every format war, this week’s news cuts in a few different directions. Add in the Marantz M1 earning an Editors’ Choice nod for doing the sensible thing exceptionally well, and the picture gets clearer: good engineering and good music still matter more than hype cycles, press junkets, or how many zeros are on the invoice.

This coming weekend marks the beginning of the silly season I mentioned last week. The calendar fills quickly with hi-fi shows that will get covered whether anyone really needs another one or not. FLAX arrives next weekend in Tampa, and the press will enjoy the warmth while it lasts. The Olympics are still underway, which means no Tampa Bay Lightning NHL games, still the best show in town. Shows are work, not vacations, and covering them costs money. Airfare, taxis, meals, and the quiet expenses nobody lists on a receipt add up fast.

It is also worth being clear with readers about how this works. Some shows cover hotel costs for media because without coverage there is no visibility, no buzz, and no record of what actually happened. Transparency matters. The media business is under real pressure right now. Publications are shrinking, budgets are tight, and layoffs have been widespread over the past year. Ask the people at the Washington Post, Tech Radar, Digital Trends, Sound & Vision, and others. We have been fortunate to add experienced talent because of that reality, but nobody should assume that publications are rolling in money. Even the biggest names are watching every dollar.

When it comes to press junkets, not everyone gets invited. These trips are usually reserved for high profile journalists from mainstream outlets like Forbes, T3, Wall Street Journal, and the New York Times, along with editors from specialist publications. We are not excluded from that group, which likely reflects our growing influence. I have been invited on overseas trips for product launches, factory tours, listening sessions, luxury car drives, and early looks at new TV technology in Asia, but illness, family emergencies, or other obligations have always gotten in the way. I have never been able to go.

Domestically, the rules are simple. We pay our own way. That has always been policy at eCoustics, with reimbursement handled later. Overseas press junkets are where things start to feel off, when necessary access blurs into hospitality and the line between reporting and obligation gets harder to see. Audio Group Denmark’s recent introduction in Aalborg of its $1.1 million flagship loudspeakers and $115,000 mono block power amplifiers for a very select group of the press sharpened that concern and has become a topic online in recent days.

When you are flown overseas, wined and dined, there is an unspoken expectation that coverage will reflect the experience. They are hardly alone in this practice, and it says nothing about the quality of what was introduced. By every account I have heard from those who were there, the experience was out of body phenomenal. The harder truth is that entry into this level of audio now borders on the absurd. One might need to sell off body parts just to get in the door, and even that feels optimistic given the general condition of most of the audiophile press.

Audiophile Excess Runs Wild in Denmark

Back in October at T.H.E. Show New York, which was held in New Jersey despite the branding gymnastics, I had my first real exposure to Audio Group Denmark. Calling it New York clearly sounds better on a banner, even if the venue landed nowhere near the part of the Garden State where I actually live. Still, it was enough to make one thing clear: Danish high-end audio is having a moment, and it is not subtle.

That moment extends well beyond Audio Group Denmark. Denmark has been quietly exporting serious audio thinking for decades, with brands like Gryphon, Dynaudio, Buchardt, DALI, Bang & Olufsen, Audiovector, Lyngdorf, Ortofon, and Raidho all contributing to Denmark’s oversized footprint in the high end. Different philosophies, different price brackets, same national tendency to push engineering harder than the market sometimes expects.

Audio Group Denmark sits firmly in that conversation but plays its own game. Its core brands Ansuz, Børresen, and Aavik were out in force, supported by their North American team and HiFi Loft, their dealer with locations on West 44th Street in Manhattan and in Glens Falls, just north of Saratoga Springs and not far from Lake George. It is a part of upstate New York where the term summer home tends to mean something very specific and very expensive.

What stood out was not just the technical ambition on display, but the pricing ambition as well. Danish brands across the board are pushing boundaries right now, both in how far they are willing to go technologically and how unapologetic they are about cost. Audio Group Denmark, in particular, has no interest in playing it safe. My first real exposure to them will not be my last. That was clear before I left the room.

Anyone thinking about a system designed to stay under $30,000 should stop reading now. Even a modest configuration built around their stand mount speakers, an integrated amplifier with streaming, and the required cabling clears that threshold quickly, before analog sources or outboard stages even enter the conversation.

At T.H.E. Show New York 2025, the two Danish systems on display occupied a very different financial lane, landing between $90,000 and $360,000 USD. Those figures are real. From a listening standpoint, the lower cost $90,000 system was far more compelling to me, but both already lived well beyond what most listeners would consider attainable.

What was introduced last week, however, makes those show systems look almost entry-level. When you factor in the Børresen M8 Gold Signature loudspeakers at roughly $1.15 million per pair and the Aavik M-880 monoblock amplifiers at $115,000 each, the scale shifts entirely. These are not conceptual exercises or dressed up prototypes.

The Aavik M-880 uses a reworked Class A amplification stage that maintains its bias 0.63 volts above the required current level at all times. The goal is continuous Class A operation regardless of load or signal conditions, while keeping operating temperatures lower than traditional Class A designs to improve long term stability and reliability; which is a good plan when you consider the “rated” power output and size of these amplifiers.

Power delivery is equally unapologetic. Each M-880 is rated at 400 watts into 8 ohms, 800 watts into 4 ohms, and approximately 1,300 watts into 2 ohms. Add sources, cabling, and the supporting ecosystem that inevitably comes with systems at this level, and it is very likely that the total system cost is approaching $2 million at its peak.

The Aavik M-880 mono amplifier measures 794.02 mm high, 342.00 mm wide, and 509.68 mm deep, which translates to 31.26 inches in height, 13.46 inches in width, and 20.07 inches in depth. Each amplifier weighs 70.0 kilograms, or 154.3 pounds.

The Gold Standard?

At the heart of the Børresen M8 Gold Signature is a folded dipole bass architecture that defines both its scale and its intent. Each loudspeaker uses two dedicated bass modules populated by twelve 8-inch drivers, firing forward and backward in opposing polarity. The idea is not brute force but control, managing low frequency energy before the room gets a chance to do what rooms usually do.

Every pair is built and calibrated in Denmark, with final measurements and listening sessions completed before the speakers leave the factory. The look is unapologetically serious: black high gloss lacquer, carbon accents, and zero attempt to disguise the mass.

That mass is substantial. Each speaker stands just over 87 inches tall, spans roughly 25 inches in width, and reaches more than 32 inches deep. At 325 kilograms per cabinet, or about 716.5 pounds, placement is a commitment, not a casual decision. The specified frequency range stretches from 20 Hz to 50 kHz, with a sensitivity rating of 87 dB.

The system is effectively tri sectional. Bass impedance is rated at 5 ohms, while the midrange and treble sections sit at 8 ohms, with each section requiring more than 100 watts of amplification.

The crossover between mid bass and tweeter is set at 2,400 Hz, while bass integration is handled externally via an active crossover that is not included. High frequencies are delivered by Børresen’s RP94 Gold Signature ribbon planar tweeter, supported by two IronFree5 Gold Signature drivers for midrange and upper bass duties, while twelve IronFree8 Gold Signature drivers handle the low end.

This is not a loudspeaker designed to coexist quietly in a room. The fact that it was demonstrated in an auditorium sized performance hall, elevated on a stage, says a lot about the assumptions baked into the design. Context matters here. These are loudspeakers that expect space, structural support, and a listening environment that can accommodate their scale and output without compromise.

We shall miss the children.

Craft Recordings Revives Wes Montgomery’s Full House for the OJC Series

This Craft Recordings OJC pressing of Full House ($38.98 at Amazon) is all analog from the original tapes, cut by Kevin Gray at Cohearent Audio and pressed on 180 gram vinyl at RTI. A 24-bit/192kHz high resolution digital edition is available for those who want it. Recorded live on June 25, 1962 at Tsubo in Berkeley, the album captures Wes Montgomery at a point where restraint and intensity exist side by side. He can sound smooth and measured one moment, then suddenly lean in hard enough to make you sit up and pay attention.

Johnny Griffin is on tenor sax, backed by the Wynton Kelly Trio with Wynton Kelly, Paul Chambers, and Jimmy Cobb, all fresh from their time with Miles Davis and fully locked in. The pressing itself is clean and well executed, with excellent clarity through the guitar and horns and a sense of presence that feels natural rather than hyped. It is the kind of record that makes you wish you had been in the room that night, even if only for a set.

An audiophile once told me, back in my twenties, that Wes Montgomery was mostly hype and not all that impressive. This came from the same guy who shushed me so we could sit through yet another Eagles demo on speakers neither of us could afford. I left the show, walked into Sam the Record Man, bought two Wes Montgomery records, and learned something useful very quickly. Some audiophiles know as little about jazz guitar as I know about the inner workings of nuclear propulsion, which is saying something considering my college roommate went on to become a USN captain running submarines and carriers.

Wes Montgomery was not hype. He was about feel, timing, touch, and control, with the ability to shift from calm to confrontation without losing the thread. Records like Full House make that obvious within minutes. Call it whatever you want, but the playing still holds up, and it still exposes bad takes just as efficiently as it did back then.

Where to buy: $38.98 at Amazon

Marantz M1 Streaming Amplifier Is Hiding in Plain Sight

The Marantz M1 was released well over a year ago, but in a category that moves quickly, time can be useful. With so many network amplifiers competing on features alone, it is easy to miss products that take a more measured approach. The M1 does not try to dominate on paper. It focuses on stable performance, sensible design choices, and an emphasis on sound quality over spectacle.

The M1 is rated at 100 watts per channel with a specified distortion figure of 0.005 percent THD. It includes HDMI eARC for television integration and provides a dedicated subwoofer output with adjustable crossover points and a plus or minus 15 dB level trim. That allows for proper configuration of a 2.1 system rather than a fixed one size approach. The amplifier operates fully in the digital domain and supports hi resolution PCM up to 24-bit/192 kHz as well as DSD playback.

Streaming and connectivity are well covered. Bluetooth, Spotify Connect, Qobuz Connect, AirPlay 2, and HEOS are all supported, with HEOS also enabling multi room playback and integration with control systems such as Control4, URC, and Crestron. There is no built in phono stage, so analog playback requires an external solution.

A full review is coming next week, but early listening with the DALI Kupid, Q Acoustics 3020c, and Acoustic Energy AE100 MK2 was telling. Fireworks may be a strong word, but Bluesound and WiiM may not love what follows.

Where to buy: $1,000 at Crutchfield | Amazon

Qobuz Fixes CarPlay and Brings Siri Into the Loop

If you use Qobuz at home, great. If you use it in the car through Apple CarPlay, the experience until now has been less convincing. Scrolling through playlists while driving was awkward, the interface was not doing anyone any favors, and asking Siri to find a specific track or playlist went nowhere. That is the kind of thing that earns looks from the passenger seat that suggest you should keep both hands on the wheel.

For anyone who spends real time behind the wheel, those small frustrations add up. I average 30,000 to 40,000 miles a year, and there are only so many times you can give up and start jabbing at the dashboard while the NHL Network blares on SiriusXM before it becomes a pattern. The latest Qobuz CarPlay update tackles those pain points in a practical way, improving day to day usability and finally making Siri a functional part of the experience. It does not reinvent in car listening, but it makes Qobuz far more livable where many of us use it the most.

So what did Qobuz actually change, and why does it matter. The CarPlay experience has been rebuilt from the ground up, with a cleaner interface and features that users have been asking for since CarPlay support first arrived. The biggest day to day fix is simple but overdue: shuffle is now available directly from the player, exactly where it should have been all along.

Just as important, Siri finally works the way it should. You can now search, browse, and control playback entirely by voice without poking at the screen. That includes asking Siri to play a specific playlist, artist, or favorite track, turning shuffle or repeat on and off, adding the current song to a playlist or your library, and even asking what is currently playing. The full Discover experience is also available in CarPlay, including personalized playlists, Release Watch, and Radio, all accessible safely while driving.

It is also a cosmetic update, and that part matters more than it sounds. You can now actually see things you could not before, with a cleaner layout that makes sense at a glance. Scrolling through your own playlists or Qobuz’s curated ones no longer frustrates, and discovery is finally usable on a CarPlay screen. The interface is clearer, more logical, and far easier to navigate, unlike the backseat of my car, which remains a lost cause thanks to kids and a dog.

More importantly, this cleanup makes Qobuz’s strengths visible. Hi-res playlists and editorial content are no longer buried or awkward to access, which means the stuff audio dorks and editors actually care about is front and center where it belongs. It does not just look better. It makes the service easier to live with, especially if you spend serious time behind the wheel.

David Solomon can relax. The Facebook messages will stop. Qobuz finally fixed what needed fixing, and for those of us who live in the car as much as the listening room, that actually matters. Long live Qobuz.

Related Reading:

• Culture Wars At The Super Bowl, Meze STRADA’s Curious Tuning, Record Store Day 2026, Hi-Fi Show Overload?: Editor’s Round-Up
• Jazz Dispensary At 10, Integrators In The Keys, Wireless Headphones On The Road, Catherine O’Hara, And Sinners Leading The Pack: Editor’s Round-Up
• Steve Cropper, FiiO Air Link, Wharfedale Diamond 12.3 & Other War Stories — An Editor’s Round-Up From A Hospital Bed
• The Biggest Consumer Audio & Video Stories Of 2025: What Mattered, What Changed, And Why Content Still Rules

If you’ve only played racing games with a controller, a proper wheel is one of the most dramatic “oh, this is different” upgrades you can make. Braking feels more natural, steering gets more precise, and even older racing games become more fun because you’re actually driving instead of tapping sticks. Right now, the Logitech G29 Driving Force wheel and pedals are $199.99, down from $329.99 for 39% off. The important detail is the countdown: this discount is shown with a timer and is set to end soon.

What you’re getting

The G29 bundle includes a steering wheel and floor pedals with real force feedback, which is the feature that adds weight and road feel instead of just vibration. You also get stainless steel paddle shifters and a leather steering wheel cover, which gives the wheel a more “real gear” feel than the plasticky entry-level stuff.

Compatibility is a big part of the appeal here, too: it’s designed for PS5, PS4, PC, and Mac, so it can move with you across setups.

Why it’s worth it

The main reason to buy a wheel is immersion, but the second reason is control. Once you get used to it, you can be smoother and more consistent with steering and throttle, which helps in everything from casual cruising to competitive lap times.

This is also a great purchase for anyone who rotates between racing genres. It works for arcade-style racers, but it’s especially satisfying for sim-leaning games where throttle control and braking really matter. And because it’s a wheel + pedal kit, you’re not piecing together an expensive setup one part at a time just to get started.

The bottom line

At $199.99, the Logitech G29 is a solid deal for anyone who wants to make racing games feel more physical and more precise on PS5 or PC. If you’ve been waiting for a price drop, the timer matters here, because once the deal ends, this is the kind of accessory that tends to bounce back up quickly.

The news that we (pretty much) told you last year is now done and dusted. HBO Max is coming to the UK, and it’ll be wrapped up into a nice shiny bow for Sky customers as part of its Ultimate TV subscription when it goes live in March.

This is a boon for Sky customers (new and existing), as it keeps hold of the Warner Bros. content that’s not just legendary but become synonymous with Sky for decades now.

Its partnership with Warners brought the likes of Game of Thrones, Boardwalk Empire, Entourage (remember that?) and Girls to the UK, and this new deal makes it easier for Sky customers to access that content in one place.

But, and there always is a but with me, I do have an issue.

Why can’t we have it all in 4K as standard?

Advertisement

Ultimate TV subscription in name only

I expect, at the very least, that when you call a subscription ‘Ultimate’, you’re getting all the bells and whistles that come with it. The (new) Ultimate TV sub, will have Netflix, Disney+, HBO Max, and Sky’s own content all included as standard, but read the small print, and the versions of these services and apps that you get come with ads, and, they’re only available in HD.

I can’t help feel a little disappointed that watching on my 4K TV, it’s actually having to upscale this legion of content because 4K and HDR (and Atmos) are an optional extra.

Advertisement

This is not strictly Sky’s fault, though it does play into the issue by offering an optional UHD/Atmos pack that you have to buy to unlock access to 4K HDR content. Streaming services themselves have secreted away all their 4K goodness to their premium tiers, in the hopes that you’ll pay more money for higher quality.

Except it doesn’t feel like people are doing that.

After an initial explosion of 4K live broadcasts, it’s mostly been watered down to HD HDR. The last Euros were shown in just HDR, and I honestly can’t remember the last time the BBC showed the FA Cup Final in anything but HD – it feels like it was aeons ago now.

This is partly down to the expense of broadcasting in 4K HDR (especially live), but also that viewers didn’t seem to be watching in 4K either, preferring to catch the action on their mobile devices instead.

Advertisement

Advertisement

To get HBO Max, Netflix, and Disney+ in 4K HDR; you’ll have to pay more, which makes the Ultimate TV subscription, which initially looks like a steal, a little less appetising if you want the best quality in your home.

I’ve always advocated for streamers and broadcasters to bring 4K to more people by making it the standard. Instead, we’re still locked into HD, which has been around for 20 years now. I think it’s time to give up the ghost, stop trying to entice people to pay extra for 4K and give them the best experience from the outset.

I doubt that Sky would love my thoughts on it, but the word Ultimate implies an experience that you’re getting the best. You are in a way, but it comes with a caveat. I do wish that barrier to 4K was taken down, not just with Sky but with every streaming service. Though many are embracing the streaming future, it still feels like we’re locked in the past.

Advertisement

A spool of filament rests calmly on a shelf, looking exactly like the usual orange Prusament roll found in numerous 3D printers, yet it hides a little secret. Prusa wanted a one-of-a-kind gift and asked Matt Denton to transform a regular 2kg spool of filament into an out-of-the-ordinary remote-controlled robot dubbed SpoolBot, which you’d be hard-pressed to tell is actually a robot going for a little roll on its own power.

Denton started from scratch with a genuine Prusament spool and simply retained the outward appearance the same, which means it still has the orange filament wrapped around a center drum in a sloppy, but highly realistic pattern. One side of the spool is fixed, but the other side detaches with magnets to allow you to go in and make changes. Those black plastic ends are actually the pieces that move the spool; they are the driving wheels. Inside, everything is held together by an internal frame that prevents the entire structure from looking out of place.

Bambu Lab A1 3D Printer, Support Multi-Color 3D Printing, High Speed & Precision, Full-Auto Calibration…

• High-Speed Precision: Experience unparalleled speed and precision with the Bambu Lab A1 3D Printer. With an impressive acceleration of 10,000 mm/s…
• Multi-Color Printing with AMS lite: Unlock your creativity with vibrant and multi-colored 3D prints. The Bambu Lab A1 3D printers make multi-color…
• Full-Auto Calibration: Say goodbye to manual calibration hassles. The A1 3D printer takes care of all the calibration processes automatically…

The whole device is powered by two geared DC motors from Pololu, each equipped with an encoder that tells the bot exactly what it’s doing and causes the drive wheels at the spool’s edges to turn. Batteries are put low to function as a counterbalance, ensuring that the entire assembly remains upright when the spool spins around its center. A DFRobot Romeo Mini ESP32-C3 board handles all of the control, and it works in tandem with a BNO085 IMU sensor to keep an eye on things and ensure the spool remains upright and stable. An RC receiver links to a simple handheld controller.

Movement is accomplished by a technique known as differential drive; in order to travel in a straight path, both motors must be moving at the same speed; however, varying the speeds results in pleasant smooth bends. The IMU monitors how far up and down and side to side it moves, and if it becomes unsteady, it slows down the motors to keep the whole thing from toppling over. Then there’s the gyro feedback, which effectively keeps the spool on track even when it’s on an uneven surface like a carpet. Several operating modes are available, allowing you to choose how the bot behaves. For example, you can keep it sitting in one location, have it follow a trail back to where it began, cruise along at a given speed, or even perform some beautiful spins at full or half speed.

The entire assembly required a lot of meticulous planning and engineering to get right; as you can see, the outer shell must be able to spin freely on the center hub, so it’s all about getting the bearings perfect so it travels smoothly. The insides of the spool are made up of 3D-printed elements in black PETG and orange PLA, which are held in place by a variety of ingenious components such as heat-set inlays and precision bolts. The wiring is neatly tucked away so that it does not interfere with the movement of the various components. It took around five weeks to complete the project, from designing it in CAD to printing the pieces and writing the code.

Matt Denton chose to be generous and share all of the build files and code with the world. You can get them on Printables and on GitHub, and there is even a video guide that shows you how to create one yourself, from installing the bearings to wrapping the filament around the central drum and calibrating the controller. Of course, like any decent robot, SpoolBot includes a couple googly eyes and a small indication to give it some personality as it rolls around the floor.
[Source]

The hardest choice to make for building your next MacBook might be selecting a color. According to Bloomberg‘s Mark Gurman, Apple has tested colors including light yellow, light green, blue and pink for its next entry-level MacBook that’s aimed at students and enterprise users.

Beyond the more vibrant colors, Gurman said that Apple has also trialed its classic silver and dark gray colorways for its cheaper laptop. Gurman added that not all of these six colors will make it to the final product, but Apple has recently shown it’s not afraid to dip into flashier options. Apple refreshed the iMac in 2024 with a total of seven colors and swapped out the space gray option for sky blue for the latest MacBook Air.

Color choices aside, the latest rumors point to the upcoming MacBook having a price tag that’s anywhere between $699 and $799. To achieve that lower price point, Apple is expected to port over its chips designed for iPhones, like the A18 Pro that we first saw with the iPhone 16 Pro Max. We’re also anticipating Apple will compromise on specs, ports, or even the display, but Gurman reported that the company won’t be skimping when it comes to the shell. According to Gurman, Apple will employ a new manufacturing process to craft aluminum shells for the affordable MacBook, instead of opting for a cheaper material like plastic to cut costs. We may not have to wait long to see the official colors of the budget MacBook, as Gurman reported that it will be announced during an event in March.

Welcome back to TechCrunch Mobility — your central hub for news and insights on the future of transportation. To get this in your inbox, sign up here for free — just click TechCrunch Mobility!

We are in the midst of one of my four favorite times of year — earnings season. And it’s not just that I like numbers. These required filings cut through a lot of the marketing noise presented by companies the rest of the year. They also help me assess the short- and long-term stakes the companies face.

Rivian’s fourth-quarter and full-year earnings did precisely that. My takeaway: Software, and specifically its technology joint venture with Volkswagen Group, was the company’s savior in 2025. It will also buoy the company into 2026 (another $2 billion is expected from VW Group) as Rivian launches its most important product to date: the lower-cost R2 SUV. 

The company’s earnings also provided a progress report on its bid to lower the cost of goods sold per unit. The TL;DR is that the cogs per unit for its current portfolio is still high but dropping, meaning it’s losing less on each vehicle it sells. According to Rivian, the company’s automotive cogs per unit delivered was $100,900 in 2025, down from $110,400 in 2024. 

The upcoming R2, which is supposed to be considerably cheaper (both in production cost and price tag) than its flagship R1T truck and R1S SUV, will be the next big test. We’ll get some insight into the results of that later this year.

The R2 is expected to go into production in the first half of the year (we’re hearing June), and based on its guidance for 2026, Rivian is confident it has the demand and the ability to ramp production. The company expects to deliver between 62,000 and 67,000 vehicles in 2026 — which could provide up to a 59% bump from last year. Rivian delivered 42,247 vehicles in 2025, which includes its two R1 consumer vehicles and the electric delivery van (EDV).

The market loved that guidance, btw. Rivian stock shot up 27% in the day after it reported earnings.

A little bird

Over the past 18 months, I’ve noticed a divergence in how Uber and Lyft are approaching AVs. Uber is locking up AV partnerships with every player it can. Lyft is trailing behind. Turns out, I am not alone in this observation. 

Insiders have shared their puzzlement about why Lyft hasn’t been more aggressive on this front. They noted that Lyft is sitting on about $1.8 billion in cash, cash equivalents, and restricted cash, and recently announced a new $1 billion share repurchase program that represents about 15% of its market cap, per CNBC. That has some wondering why Lyft did not invest in parts of the AV value chain like Uber is doing versus buying shares back.

Meanwhile, these little birds also pointed to a few top executives who have departed over the past year. Aurélien Nolf left his position as VP of financial planning and analysis and investor relations to become CFO of Navan. Audrey Liu, who was an executive VP and head of rider and community safety, is now at Adobe. Ameena Gill, who was VP of safety and customer care just took a job at rival Uber.

Got a tip for us? Email Kirsten Korosec at kirsten.korosec@techcrunch.com or my Signal at kkorosec.07, or email Sean O’Kane at sean.okane@techcrunch.com. 

Deals!

Close followers of the mobility-crazed years, between 2015 and 2019, might recall how many lidar companies popped up during that time. Many of the dominant and buzziest ones have since shuttered, while some of the smallest players have hung on and expanded. 

Take Ouster, for instance. I remember way back when Ouster had this tiny little booth in the jam-packed startups area (Eureka Park) at CES. Today, the company is much bigger — thanks to scale, its 2022 merger with rival Velodyne, and its acquisition of Sense Photonics in 2021. And it doesn’t appear to be finished. 

The company most recently acquired Stereolabs, a company that makes vision-based perception systems for robotics and industrial applications, for a combination of $35 million and 1.8 million shares.

As TechCrunch senior reporter Sean O’Kane notes in his article, the deal is the latest in a march toward consolidation among perception sensor suppliers. (Just last month, MicroVision bought the lidar assets of the buzzy-but-now-bankrupt Luminar for $33 million.)

So why all the activity? It’s complicated, as they say. From my POV, the frenzy around “physical AI” has reignited interest and investment in sensor technologies, particularly cameras.

Other deals that got my attention …

Ever, the EV-only marketplace, raised $31 million in a Series A funding round led by Eclipse. Other backers include Ibex Investors, Lifeline Ventures, and JIMCO — the investment arm of the Saudi Arabian Jameel family (an early investor in Rivian).

Natilus, the San Diego-based startup developing blended-wing aircraft, raised $28 million in a Series A funding round led by Draper Associates. Other investors include Type One Ventures, The Veteran Fund, and Flexport, as well as new backers New Vista Capital, Soma Capital, Liquid 2 VC, VU Venture Partners, and Wave FX.

Notable reads and other tidbits

Aurora shared in its Q4 and full-year earnings report that its self-driving trucks can now travel nonstop on a 1,000-mile route between Fort Worth and Phoenix — exceeding what a human driver can legally accomplish. The company shared a number of other tidbits, and financials, which you can read about here. 

The U.S. Securities and Exchange Commission closed its investigation into Fisker last year, TechCrunch was able to learn, thanks to a Freedom of Information Act request. 

Lyft has launched teen accounts, a product that allows minors as young as 13 to hail a ride without an adult in 200 U.S. cities, including Atlanta, Boston, Chicago, and New York.

A fresh batch of videos gives us the best look at how Rivian has changed the rear door manual release on its upcoming R2 SUV. This seemingly minor design detail has life-or-death stakes and comes as the EV industry, and particularly Tesla, is getting pressure to change concealed, electronic door handles. 

The Trump administration officially repealed the EPA’s 2009 “endangerment finding,” which found that greenhouse gases such as carbon dioxide and methane were a threat to human health and welfare. This change would only affect tailpipe emissions for cars and trucks — if the EPA makes it through the lengthy process of repealing the law, which will certainly include numerous lawsuits aimed at stopping it.

Uber has locked in a couple dozen AV partnerships, and we’re starting to see the results of those deals. China’s Baidu and Uber plan to launch robotaxis in Dubai in the next month, starting with select locations within the Jumeirah area. Meanwhile, Chinese robotaxi company WeRide and Uber announced a “major expansion of their strategic partnership” to deploy at least 1,200 robotaxis across the Middle East through 2027, according to the companies. As part of this, WeRide and Uber have launched a robotaxi service in downtown Abu Dhabi.

Waymo pulled the human safety driver from its autonomous test vehicles in Nashville as the Alphabet-owned company moves closer to launching a robotaxi service in the city. Meanwhile, this tech-forward company is wrestling with the analog problem of ensuring the doors of its robotaxis are properly shut. Its solution? Pay DoorDash gig workers to shut Waymo robotaxi doors. Waymo tells us this is a pilot program in Atlanta to enhance its AV fleet efficiency. 

One final Waymo item: The company is starting to roll out its sixth-generation “Waymo Driver,” which is integrated into the Zeekr RT (rebranded Ojai) and will eventually be in the Hyundai Ioniq 5. Waymo has started “fully autonomous operations” in the Ojai vehicle in San Francisco and Los Angeles and is giving access to employees. The public will have to wait for a bit.

One more thing …

Rivian has pitched its upcoming R2 SUV as a more affordable model. What does “more affordable” mean? The company has thrown around $45,000 and $50,000 as a base price. The company’s launch version of the R2, which will be a dual-mode and all-wheel-drive premium trim, will undoubtedly be more expensive. In our newsletter this week, we asked readers, “What’s your guess on the cost of the launch edition?”

Sign up for our newsletter to participate in our polls!

CTM360 reports that more than 4,000 malicious Google Groups and 3,500 Google-hosted URLs are being used in an active malware campaign targeting global organizations.

The attackers abuse Google’s trusted ecosystem to distribute credential-stealing malware and establish persistent access on compromised devices.

The activity is global, with attackers embedding organization names and industry-relevant keywords into posts to increase credibility and drive downloads.

Read the full report here: https://www.ctm360.com/reports/ninja-browser-lumma-infostealer

How the campaign works

The attack chain begins with social engineering inside Google Groups. Threat actors infiltrate industry-related forums and post technical discussions that appear legitimate, covering topics such as network issues, authentication errors, or software configurations

Within these threads, attackers embed download links disguised as: “Download {Organization_Name} for Windows 10”

To evade detection, they use URL shorteners or Google-hosted redirectors via Docs and Drive. The redirector is designed to detect the victim’s operating system and deliver different payloads depending on whether the target is using Windows or Linux

 

Windows Infection Flow: Lumma Info-Stealer

For Windows users, the campaign delivers a password-protected compressed archive hosted on a malicious file-sharing infrastructure

Oversized archive to evade detection

The decompressed archive size is approximately 950MB, though the actual malicious payload is only around 33MB. CTM360 researchers found that the executable was padded with null bytes — a technique designed to exceed antivirus file-size scanning thresholds and disrupt static analysis engines.

AutoIt-based reconstruction

Once executed, the malware:

• Reassembles segmented binary files.

• Launches an AutoIt-compiled executable.

Advertisement

• Decrypts and executes a memory-resident payload.

The behavior matches Lumma Stealer, a commercially sold infostealer frequently used in credential-harvesting campaigns

Observed behavior includes:

• Browser credential exfiltration.

• Session cookie harvesting.

• Shell-based command execution.

Advertisement

• HTTP POST requests to C2 infrastructure (e.g., healgeni[.]live).

• Use of multipart/form-data POST requests to mask exfiltrated content.

CTM360 identified multiple associated IP addresses and SHA-256 hashes linked to the Lumma-stealer payload.

Linux Infection Flow: Trojanized “Ninja Browser”

Linux users are redirected to download a trojanized Chromium-based browser branded as “Ninja Browser.”

The software presents itself as a privacy-focused browser with built-in anonymity features.

However, CTM360’s analysis reveals that it silently installs malicious extensions without user consent and implements hidden persistence mechanisms that enable future compromise by the threat actor.

Malicious extension behavior

A built-in extension named “NinjaBrowserMonetisation” was observed to:

• Track users via unique identifiers

• Inject scripts into web sessions

• Load remote content

• Manipulate browser tabs and cookies

Advertisement

• Store data externally

The extension contains heavily obfuscated JavaScript using XOR and Base56-like encoding

While not immediately activating all embedded domains, the infrastructure suggests future payload deployment capability.

Silent persistence mechanism

CTM360 also identified scheduled tasks configured to:

• Poll attacker-controlled servers daily

• Silently install updates without user interaction

Advertisement

• Maintain long-term persistence

Additionally, researchers observed that the browser defaults to a Russian-based search engine named “X-Finder” and redirects to another suspicious AI-themed search page

The infrastructure appears tied to domains such as:

• ninja-browser[.]com

• nb-download[.]com

Advertisement

• nbdownload[.]space

Campaign Infrastructure & Indicators of Compromise

CTM360 linked the activity to infrastructure, including:

IPs:

• 152.42.139[.]18

• 89.111.170[.]100

C2 domain:

Multiple SHA-256 hashes and domains associated with credential harvesting and info-stealer distribution were identified and are available in the report.

Risks to organizations

Lumma Stealer risks:

Ninja Browser risks:

• Silent credential harvesting

• Remote command execution

• Backdoor-like persistence

Advertisement

• Automatic malicious updates without user consent

Because the campaign abuses Google-hosted services, the attack bypasses traditional trust-based filtering mechanisms and increases user confidence in malicious content.

Defensive recommendations

CTM360 advises organizations to:

• Inspect shortened URLs and Google Docs/Drive redirect chains.

• Block the IoCs at firewall and EDR levels.

Advertisement

• Educate users against downloading software from public forums/sources without verification.

• Monitor scheduled task creation on endpoints.

• Audit browser extension installations.

The campaign highlights a broader trend: attackers are increasingly weaponizing trusted SaaS platforms as delivery infrastructure to evade detection.

About the Research

The findings were published in CTM360’s February 2026 threat intelligence report, “Ninja Browser & Lumma Infostealer Delivered via Weaponized Google Services”

CTM360 continues to monitor this activity and track related infrastructure.

Read the full report here: https://www.ctm360.com/reports/ninja-browser-lumma-infostealer

Detect Cyber Threats 24/7 with CTM360

Monitor, analyze, and promptly mitigate risks across your external digital landscape with the CTM360.

Join our Community Edition

Sponsored and written by CTM360.

In the scramble for the interface, Glean is betting on something less visible: becoming the intelligence layer beneath it. 

Seven years ago, Glean set out to be the Google for enterprise — an AI-powered search tool designed to index and search across a company’s SaaS tool library, from Slack to Jira, Google Drive to Salesforce. Today, the company’s strategy has shifted from building a better enterprise chatbot to becoming the connective tissue between models and enterprise systems.

“The layer we built initially – a good search product – required us to deeply understand people and how they work and what their preferences are,” Jain told TechCrunch on last week’s episode of Equity, which we recorded at Web Summit Qatar. “All of that is now becoming foundational in terms of building high quality agents.”

He says that while large language models are powerful, they’re also generic. 

“The AI models themselves don’t really understand anything about your business,” Jain said. “They don’t know who the different people are, they don’t know what kind of work you do, what kind of products you build. So you have to connect the reasoning and generative power of the models with the context inside your company.”

Glean’s pitch is that it already maps that context and can sit between the model and the enterprise data. 

The Glean Assistant is often the entry point for customers — a familiar chat interface powered by a mix of leading proprietary (ie, ChatGPT, Gemini, Claude) and open-source models, grounded in the company’s internal data. But what keeps customers, Jain argues, is everything underneath it. 

First is model access. Rather than forcing companies to commit to a single LLM provider, Glean acts as the abstraction layer, allowing enterprises to switch between or combine models as capabilities evolve. That’s why Jain says he doesn’t see OpenAI, Anthropic, or Google as competition, but rather as partners. 

“Our product gets better because we’re able to leverage the innovation that they are making in the market,” Jain said. 

Second are the connectors. Glean integrates deeply with systems like Slack, Jira, Salesforce, and Google Drive to map how information flows across them and enable agents to act inside those tools. 

And third, and perhaps most important, is governance. 

“You need to build a permissions-aware governance layer and retrieval layer that is able to bring the right information, but knowing who’s asking that question so that it filters the information based on their access rights,” Jain said. 

In large organizations, that layer can be the difference between piloting AI solutions and deploying them at scale. Enterprises can’t simply load all their internal data into a model and create a wrapper to sort out the solutions later, says Jain. 

Also critical is ensuring the models don’t hallucinate. Jain says its system verifies model outputs against source documents, generates line-by-line citations, and ensures that responses respect existing access rights. 

The question is whether that middle layer survives as platform giants push deeper into the stack. Microsoft and Google already control much of the enterprise workflow surface area, and they’re hungry for more. If Copilot or Gemini can access the same internal systems with the same permissions, does a standalone intelligence layer still matter?

Jain argues enterprises don’t want to be locked into a single model or productivity suite and would rather opt for a neutral infrastructure layer rather than a vertically integrated assistant.

Investors have bought into that thesis. Glean raised a $150 million Series F in June 2025, nearly doubling its valuation to $7.2 billion. Unlike the frontier AI labs, Glean doesn’t need massive compute budgets.

“We have a very healthy, fast-growing business,” Jain said.