Microplastics absolutely saturate the Earth’s environment, and that’s probably not a good thing unless you’re looking for a sediment marker for the Anthropocene period. On the other hand, environmental contamination only becomes a really big problem if it bioaccumulates– that is, builds up in the tissues of plants and animals. At least when it comes to worms, that’s not the case with microplastics, according to new research from the Canadian Light Source at the University of Saskatchewan.

The Canadian Light Source isn’t just some hoseheads in an igloo with a flashlight– it’s a 2.9 GeV Synchrotron tuned to produce high-energy photons. Back when Synchrotrons were used for particle physics, Synchrotron radiation was a very annoying energy sink, but nobody cares about 2.9 GeV electrons anymore. So rather than slam them into each other or a static target, the electrons just whip about endlessly, giving off both soft- and hard X-rays for material science studies– or, in this case, to observe the passage of polyethelyne microplastic particles through the guts of some very confused earth worms. To make them detectable by x-ray, the polyethylene was bonded to barium sulfate, an x-ray absorber. Equally opaque barium titanite glass microspheres were used with different worms, as a control.

Despite being fed plastic enriched with far more plastic than you’ll find outside of a 3D print farm, it seems the worm’s digestive system was able to reject the particles, even those as fine as 5 microns. That’s a good thing, because if the worms were absorbing plastic from the soil, it’s likely their predators would absorb it from the flesh of the worms, so and so forth up the food chain in the sort of cascade that made DDT a problem and makes mercury compounds so serious. If the worms are rejecting these compounds, there’s a chance other creatures can too– and at the very least, it means they aren’t building up on this bottom rung of the foot chain. If you’re looking for a more technical read, the full paper is available here.

It’s too early to say what this means for how microplastics get into humans and other animals, but it’s hopeful. Equally hopeful was the recent finding that studies that don’t rely on football-field sized X-ray machines might be picking up on microplastics from lab gloves, skewing results.

Header image: the digestive systems of earth worms as imaged by the Canadian Light Source. Credit Letwin, et al,
Environmental Toxicology and Chemistry, vgag072, https://doi.org/10.1093/etojnl/vgag072

China has more than 150 humanoid robot companies. It shipped roughly 90 per cent of the world’s humanoid robots in 2025. Its two largest makers, Unitree and AgiBot, are preparing initial public offerings that would value them at a combined 13 billion dollars. Morgan Stanley doubled its delivery forecast for the Chinese market this year to 28,000 units, a 133 per cent increase over 2025. And yet, when Morgan Stanley surveyed the companies that are supposed to buy these robots, only 23 per cent said they were satisfied with the products available. Battery life tops out at two to three hours per charge. Most deployments remain confined to exhibitions, showrooms, and Spring Festival galas where robots perform kung-fu routines for television cameras. The technology has arrived. The customers have not. China’s humanoid robot industry is the most capitalised, most productive, and most overpopulated robotics sector in the world, and it is heading for a reckoning that its government has already warned is coming.

The warning

In late 2025, China’s National Development and Reform Commission issued a rare public statement about the humanoid robot sector. Spokesperson Li Chao noted that the number of companies had climbed past 150 and was still growing, with more than half being startups or cross-industry entrants. The NDRC warned of redundant products, duplicated investment, and compressed space for genuine research and development. The language was measured. The implication was not. Beijing’s top economic planning agency was telling the market that it saw a bubble forming in the industry it had designated as one of ten priority sectors in the 15th Five-Year Plan, backed by a one-trillion-yuan state fund.

China’s smartphone supply chain has already begun pivoting to humanoid robot production, with companies like Lingyi iTech, a Foxconn supplier that assembles iPhones, targeting 500,000 humanoid units by 2030. The manufacturing infrastructure is real. The component ecosystem is deep. The problem is that the robots being produced are not yet generating the revenue their valuations imply. Unitree, which filed for a 608 million dollar IPO on Shanghai’s STAR Market, saw humanoid robot revenue surpass its quadruped robot business for the first time in 2025, but the company’s total scale remains modest relative to its targeted seven billion dollar valuation. AgiBot, which is aiming for a six billion dollar listing in Hong Kong, is in a similar position: significant technological capability, significant government backing, and a commercial market that has not yet materialised at the scale the IPO price demands.

The gap

The Morgan Stanley survey, led by China industrials analyst Sheng Zhong, found that 62 per cent of Chinese companies said they were likely to adopt humanoid robots within three years. That willingness, however, collided with a set of practical constraints that the industry has not resolved. The 23 per cent satisfaction rate reflected shortcomings in dexterity, functionality, and pricing. Ninety-two per cent of respondents said robots needed to fall below 200,000 renminbi, roughly 28,000 dollars, before mass adoption became viable. Only about 10 per cent of companies surveyed were currently evaluating or running pilot projects. The demand exists in theory. In practice, the robots are too expensive, too limited in capability, and too short on battery life to justify the investment for most industrial applications.

UBTech, one of the sector’s largest players, offered 18 million dollars to recruit a chief AI scientist, a salary that reflects both the intensity of the talent war and the recognition that the engineering challenges remaining are substantial. The Walker S2, UBTech’s latest industrial humanoid, entered mass production in early 2026 with orders exceeding 800 million yuan, and the company is building a factory in Beijing targeting 10,000 units per year by the end of 2026. But production capacity and commercial demand are different things. Morgan Stanley’s Zhong described 2026 as “a critical year as humanoid integrators strive to reach commercialisation and build up their ecosystems,” and warned of an impending shake-out. Production, he noted, is likely to be materially larger than sales, because major players are manufacturing robots internally for training and verification rather than shipping them to paying customers.

The spectacle

In April, a humanoid robot called Lightning, developed by Chinese smartphone maker Honor, won the Beijing E-Town Half-Marathon in 50 minutes and 26 seconds, beating the human world record by nearly seven minutes. More than a hundred robots competed. The event was covered globally. An engineer on the winning team said the achievement enabled technology transfer into structural reliability and cooling that would eventually benefit industrial applications. Robotics experts were less certain. The skills displayed during a half-marathon, sustained bipedal locomotion on a flat surface, do not translate to the manual dexterity, real-world perception, and adaptive problem-solving required for factory work, logistics, or the service applications that the industry’s business plans depend on.

The gap between spectacle and substance defines China’s humanoid robot moment. The Spring Festival Gala performances, the marathon records, and the viral videos of robots doing backflips generate the attention that attracts capital. The capital funds the next round of development. The development produces more impressive demonstrations. But the cycle does not produce revenue at the scale needed to justify the valuations being assigned. China’s industrial model has historically excelled at commercialising technology faster and cheaper than any Western economy, turning solar panels, electric vehicles, and batteries into globally dominant export industries within a decade. The question is whether humanoid robots follow that pattern or whether they represent a category where the gap between demonstration and deployment is structurally wider than the manufacturing advantage can close.

The competition

China’s dominance in humanoid robot shipments has not gone unnoticed. Boston Dynamics began commercial production of its electric Atlas robot in January 2026 and announced plans to deploy tens of thousands of units at Hyundai Motor Group factories, with a manufacturing facility near Savannah, Georgia, targeting 30,000 units per year by 2028. Figure AI, the leading American humanoid startup, holds a 39 billion dollar private valuation after its September 2025 fundraise, despite shipping a fraction of the volume Chinese companies manage. Tesla’s Optimus is performing basic tasks in its own factories, with Elon Musk projecting mass production and a price point of 20,000 to 30,000 dollars, though the robot is, by Musk’s own admission, “not in usage in a material way.” The Pentagon has awarded 24 million dollars in contracts to Foundation Future Industries for humanoid robot soldiers tested in Ukraine, opening a military market that Chinese companies cannot access but that validates the strategic importance governments are placing on the technology.

The pricing dynamics favour China. Unitree’s H2 is positioned below 30,000 dollars. Kepler, another Chinese maker, is targeting the same range. At CES 2026, the sheer number of Chinese humanoid robots on display, and their aggressive pricing, made clear that the supply-side economics are already competitive. The question is whether demand at those price points exists in sufficient volume to sustain an industry with 150 companies competing for it.

The reckoning

Zhong’s prediction of a shake-out is not a minority view. The NDRC’s warning, the Morgan Stanley satisfaction data, the IPO inspection of Unitree just twelve days after its STAR Market application was accepted, and the simple arithmetic of 150 companies chasing a market that delivered roughly 14,000 units in China in 2025 all point in the same direction. The companies that survive will be those that solve the commercialisation problem: identifying repeatable, scalable use cases where the economics of a humanoid robot are superior to the alternatives, whether those alternatives are purpose-built industrial arms, wheeled platforms, or human workers. The companies that do not will have burned through their funding producing impressive machines that no one outside a trade show needed to buy.

China’s humanoid robot industry has the manufacturing base, the component supply chain, the government support, and the engineering talent to lead the world. What it does not yet have is the market. The one-trillion-yuan state fund and the 15th Five-Year Plan designation ensure that capital will continue to flow. The NDRC warning ensures that Beijing is watching how it flows. Somewhere between the billion-dollar IPOs and the 23 per cent satisfaction rate, between the marathon records and the two-hour battery life, is the answer to whether China’s humanoid robot boom produces the next great Chinese export industry or the most expensive collection of trade show demonstrations the technology sector has ever funded. The robots can run a half-marathon faster than any human alive. They cannot yet work an eight-hour shift.

Mouse P.I. for Hire has caused quite the stir since it was revealed by Polish developer Fumi Games a couple of years back. Its rubberhose animation style, Doom-inspired boomer shooter gameplay, and Mickey Mouse-esque cast of characters helped it to stand out in the indie scene, and it’s already enjoyed a healthy dosage of positive reception from critics and players alike.

So, when I got the chance to try it out on Nintendo Switch 2, it’s safe to say I was pretty excited. Playing detective in a noir, rodent-filled world sounds pretty enticing, right? And given that I had some long-haul flights up ahead, going with the Switch 2 edition to mouse around on the go felt like a no-brainer.

The project falls under the US-Ireland R&D Partnership, a tri-jurisdictional initiative founded in 2006 with the aim of supporting collaborative research projects involving partners in Ireland, Northern Ireland and the US.

A new all-island research programme will support Irish-US collaboration between researchers, innovators and industry partners through a $20m investment.

The ‘research translation and commercialisation initiative’ is a trilateral project funded by Research Ireland, Northern Ireland’s Department for the Economy and the US National Science Foundation (NSF) Directorate for Technology, Innovation and Partnerships (TIP).

The project falls under the US-Ireland R&D Partnership, a tri-jurisdictional initiative founded in 2006 with the aim of supporting collaborative research projects involving partners in Ireland, Northern Ireland and the US by bringing together government departments, funding agencies, academic institutions and industry to address shared scientific, economic and societal challenges.

The initiative, which will be open to current and past tripartite partnership awardee teams who have received their US support from NSF, is also part-funded through the Irish Government’s Shared Island Fund and supported by InterTradeIreland, the cross-border trade and business development body for all-island economic collaboration.

Taoiseach Micheál Martin, TD said: “The US-Ireland R&D Partnership is a powerful example of how sustained international cooperation delivers real benefits for our people, our economy and our research community.

“This new investment builds on 20 years of success and will help ensure that cutting-edge research developed across the island of Ireland and the United States can be translated into real-world solutions and high-value jobs.”

The new initiative plans to identify research under the themes of cybersecurity, energy and sustainability, telecommunications, sensors and sensor networks, and nanoscale science and engineering, and was established as an expansion activity to support the translation of research outputs from the US-Ireland R&D Partnership into market-ready products, services and solutions.

First minister of Northern Ireland Michelle O’Neill said: “This new transatlantic initiative represents a significant opportunity to turn excellent research into real benefits for our economy and our communities, while strengthening the strong relationships we have built with partners in the US and across this island.”

The collaboration is also targeting the development of bespoke training programmes for affiliated researchers to help them to upskill in advancing their work along the translation and commercialisation path, with further funding opportunities available to selected participating teams to kickstart the creation of research-related start-ups.

“For nearly 20 years, the US-Ireland R&D Partnership has not only jointly funded numerous trilateral science and engineering research projects, it has also served as a model of how to successfully facilitate cross-border research and development,” said Brian Stone, the NSF’s chief of staff.

“Today’s announcement from NSF TIP, the Government of Ireland and Department for the Economy marks a natural next step in our transatlantic partnership, expanding our collaboration to accelerate the translation of projects into businesses and solutions, delivering significant scientific, economic and real-world benefits.”

The US-Ireland R&D Partnership has supported 107 collaborative research projects to date through $196m in combined government funding for research projects across an array of sectors. Examples include: research on next-generation communications and 6G networks conducted by University College Dublin, Queen’s University Belfast and Purdue University; work on sustainable animal health solutions by University of Tennessee, University College Cork and Queen’s University; and colorectal cancer research carried out by GE Global Research, Queen’s University and Royal College of Surgeons Ireland.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

When it comes to flagship phones, the word “Ultra” has started to lose meaning. Every brand throws it around, but very few actually deliver something that feels… ultra. Take the Samsung Galaxy S26 Ultra, for instance. It’s a solid phone, sure, but exciting? Not quite. And that’s the bigger issue with the US market right now. Some of the most interesting Android flagships simply don’t make it here.

Meanwhile, brands like Vivo, Oppo, and Honor are quietly pushing smartphone cameras into territory that feels closer to dedicated cameras than ever before. And then there’s the Xiaomi 17 Ultra. After using it for a couple of weeks, one thing is clear: this isn’t just a phone with a great camera. It’s a camera that happens to be a phone. And honestly, it kind of feels like a modern-day revival of the Samsung Galaxy Camera.

If this thing officially launched in the US, it would shake things up in a big way.

Spec Sheet Flex, But Make It Real

The Xiaomi 17 Ultra doesn’t just show up with a spec sheet — it shows off. You’re looking at a Leica-tuned triple-camera setup led by a 50MP 1-inch Light Fusion 1050L sensor with an f/1.67 aperture and LOFIC HDR, which is basically a fancy way of saying it handles highlights and shadows like a champ. Then there’s the real party trick: a 200MP periscope telephoto (Samsung HP9, 1/1.4″) with a slick continuous optical zoom from 75mm to 100mm (around 3.2x to 4.3x), stretching all the way to a wild 400mm equivalent via in-sensor crop.

Rounding things out is a 50MP ultrawide with a 115° field of view and macro support, plus a surprisingly serious 50MP autofocus selfie camera up front. And yes, it shoots 8K at 30fps and 4K at 120fps with Dolby Vision and ACES Log, which is as close as a phone gets to saying, “Yeah, I can do cinema too.” Additionally, there’s Leica optics and color tuning across all lenses. In fact, that Leica partnership isn’t just branding either. It shows up in how the photos look, feel, and behave.

Daylight Drama, Minus the Drama

Let’s start with daylight shots, because this is where most phones already do well. The Xiaomi 17 Ultra does better. Images are sharp, detailed, and rich without looking artificially processed. You get two primary profiles: Leica Authentic and Leica Vibrant. I found myself leaning toward Vibrant more often, and here’s the thing: it doesn’t go overboard.

Colors pop, but they don’t scream. Greens look lively without turning neon, blues stay controlled, and overall contrast feels more… intentional.

Honestly, it’s a refreshing break from the oversharpened, overprocessed look that some flagships lean into. Furthermore, the HDR performance is another highlight. Even in tricky lighting, the phone balances highlights and shadows beautifully, without flattening the scene.

Zoom Game That Actually Feels Like a Camera

This is where things start getting really fun. The combination of multiple lenses and a continuous optical zoom system means you’re not just jumping between fixed focal lengths. You’re actually working with something that feels closer to a real camera lens. From 1x to 2x, 3.2x, and even beyond, the results stay impressively sharp. Colors remain consistent across zoom levels, which is something many phones still struggle with.

And here’s the surprising part. I ended up using the camera at around 3.2x most of the time. It just hits that sweet spot for composition, perspective, and background separation.

Portraits That Don’t Try Too Hard

Portrait photography is another strong suit here, and it benefits massively from that telephoto hardware. You can shoot portraits using the tele lens for natural depth, or switch to portrait mode for additional processing. Either way, the results are excellent.

Edge detection is clean, subject separation looks natural, and the background blur doesn’t feel fake or overdone. In many cases, it genuinely holds its own against a decent DSLR setup.

What I really liked, though, is that you don’t always need portrait mode. Just using the telephoto lens gives you that natural compression and bokeh, especially for subjects like pets or candid shots.

Low Light, No Panic

While daylight photography is great, it’s great on a lot of other phones too. However, low-light photography is where this phone really flexes. That 1-inch sensor combined with the wide f/1.67 aperture allows it to pull in a ton of light. And the results show.

Even in challenging conditions with minimal lighting, the Xiaomi 17 Ultra manages to retain detail, control noise, and preserve the overall mood of the scene. Importantly, it doesn’t try to turn night into day. You still get that nighttime feel, just with better clarity and detail. Highlights are controlled, lens flare is minimal, and textures don’t get smudged into oblivion.

Ultrawide, But Actually Useful

The ultrawide camera here isn’t an afterthought. At 14mm, it captures a seriously wide field of view, which is great for landscapes, architecture, and group shots. Even better, image quality holds up surprisingly well, including in lower light.

That said, there’s one small annoyance. The placement of the ultrawide lens near the edge of the camera module means it’s very easy to accidentally get a finger in the frame. It’s not a dealbreaker, but definitely something to be mindful of.

The Photography Kit Pro

Speaking of the camera array, one of the best things Xiaomi did with this phone was to introduce the Photography Kit Pro, and the second best thing they did was to supply me with the kit, too. You get better ergonomics, physical controls for shooting, and an overall experience that makes you want to take more photos. It bridges that gap between smartphone photography and traditional cameras in a really satisfying way. The grip also doubles as a battery pack, which is incredibly useful during long shooting sessions.

There’s even a USB-C passthrough, so it’s easy to charge both the phone and grip simultaneously. That said, I wish Xiaomi added data passthrough as well, enabling one to connect an external SSD while the grip is attached. Maybe in future iterations, they could also add a microSD card slot to this, or better yet, a full-sized SD card slot to appeal to the photographers out there.

Selfie Cam… Exists

Now, all isn’t perfect here, and that brings me to the selfies. It’s… fine. Just fine.

HDR can be a bit inconsistent, colors often lean a little too punchy, and while there’s an attempt to smooth out skin textures, the result feels a bit off.

Of course, photography is subjective, but personally, this is one area where I’d still pick a Google Pixel any day. Even the iPhone does a solid job if you prefer softer-looking images, as you can see in the comparison shot above.

The Best Camera You Can’t (Officially) Buy?

So… is this the best camera phone right now? If photography is your priority, it’s honestly very hard to argue against it. The Xiaomi 17 Ultra brings together industry-leading hardware, genuinely thoughtful image processing, full RAW support for those who like to tweak every pixel, and smart AI tools that actually feel useful instead of gimmicky. And the best part? It’s not just a one-trick pony. Beyond the camera, you’re still getting a proper flagship experience with a top-tier chipset, a gorgeous display, and battery life that comfortably goes the distance.

But here’s the frustrating bit: you can’t officially buy it in the US. And that’s a real shame. Because if a phone like this were widely available, it would force the likes of Apple and Samsung to push their camera systems further, faster. The Xiaomi 17 Ultra isn’t trying to be the most balanced smartphone out there. Instead, it’s aiming to be the best camera you can carry in your pocket. And after spending time with it, it’s hard not to feel like the US market is seriously missing out.

Apple is rumored to be giving users the option to run various AI features in iOS 27 with third-party models as an alternative to Apple Intelligence.

Apple has been trying to catch up to the rest of the AI market, but it may not have to worry about doing so for iOS 27. If a report is true, Apple will be making it easier to use third-party alternates throughout the operating system.

According to sources of Bloomberg on Tuesday, users will be able to select from multiple third-party AI models, which can be used for various tasks in the operating system. It’s a change arriving in iOS 27, iPadOS 27, and macOS 27.

While users can already use ChatGPT for some actions on their iPhone already, the new version will work with other models as well. These integrations have apparently included models from Anthropic and Google, the sources claim.

Those models will be tasked with answering queries, editing and generating text, and image generation. This is a lot like the existing capabilities of ChatGPT in iOS 26.

Extensions and the App Store

The choice will be available as part of “Extensions,” which will let users access the generative AI capabilities from installed apps, via Apple Intelligence. This includes Siri, Writing Tools, and Image Playground, a message in a test build apparently said.

For Siri, users will be able to select a different voice for conversations that use external models. This is to make it easier for users to quickly understand which AI source is handling the query.

As usual, Apple intends to warn users that it isn’t responsible for content generated by any of the selected third-party models.

While it will require users to install apps from their selected provider beforehand, Apple will also be making it easier for users to get onboard. There’s word of a specific App Store section that will list compatible AI apps that users can download.

The connection to the App Store is something that has been brought up long in the past. Back in March 2024, there were murmors of an AI App Store, which the new report is similar to in concept.

Rumors of Siri supporting other third-party AI tools have also surfaced, including one March report mentioning the use of installed apps.

However, there’s also the question of whether users will actually take advantage of this capability in the first place.

While Apple has been behind in the AI race, it did move to catch up in January thanks to a multi-year deal with Google. Under it, Apple would use Google’s Gemini models and cloud technology to help flesh out Apple’s Foundational Models.

With WWDC 2026 on the horizon in June, we don’t have long to wait to see what Apple’s AI strategy will actually be.

One thing is certain about The Devil Wears Prada 2: The ambitious undertaking of making a sequel of a cult status film after 20 years has succeeded, at least as far as box office figures are concerned. The numbers speak for themselves, with $77 million generated in US theaters and another $157 million in the rest of the world since its April 29 release.

In the face of such a box office smash, this installment has inspired heated debates for days about its quality and comparisons to the original. In Italy, those arguments even extend to the dubbing of the film.

The controversy stems from the choice of voice actors in the Italian version of The Devil Wears Prada 2, who are themselves a nod to continuity; it’s the same cast as the original. Connie Bismuto is back to voice Anne Hathaway as Andy, Francesca Manicone dubs Emily Blunt as Emily, Gabriele Lavia is once again Stanley Tucci’s Nigel, and above all, Maria Pia Di Meo, the actress who has been the familiar and expressive voice of Meryl Streep in practically all the Italian adaptations of recent years—including the fearsome Miranda Priestly—returned for the sequel.

While many fans were happy to revisit these familiar voices, other viewers noticed some idiosyncrasies, largely due to the advanced age of the voice actors themselves, especially Di Meo and Lavia.

Di Meo, born in 1939, is undoubtedly a master of Italian dubbing, and her performances, linked to such great Hollywood actresses as Jane Fonda, Julie Andrews, Mia Farrow, Barbra Streisand, and Streep, have made her one of the most recognizable and expressive voices of cinema in that country’s theaters.

Yet some say her performance now reveals too much of the passage of time and that there’s a disconnect between her 87-year-old voice and that of a character as energetic and sharp as Miranda (played, in the original, by a 76-year-old Streep). Could this nine-year gap be too great to bridge? The same has been said of Lavia, who dubs Stanley Tucci with a result that often sounds a bit forced.

But more than a question of age, perhaps there’s a broader discussion to be had about dubbing in general and its effectiveness in an era in which downloads first and then streaming platforms have accustomed us to seeing more and more content in the original language.

Even just listening to the trailers released online for The Devil Wears Prada 2, a native Italian speaker will notice not only that the voices that have aged into varying degrees of mismatch but also that the speed of the lines makes them hard to follow. And what about the adaptation of the dialog? “I’m a features editor at Runway,” Anne Hathaway’s Andy says proudly, but how many of those who live outside newsrooms know what a features editor is? And again, when Miranda’s second assistant says, “I have to pee, I drank a venti,” how many people outside of the US understand on the fly that she’s referring to a Starbucks drink?

Perhaps, then, what hasn’t aged so well is not so much the voices of individual dubbers but a dubbing system that no longer keeps pace—in most cases—with the speed and specificity with which the content itself is produced. In the face of this consideration, however, one cannot ignore that, at least in a market like Italy, especially at the cinema, people overwhelmingly go to see dubbed versions of movies.

So these same online debates perhaps serve to keep attention focused on how many countries outside of the US experience these films. And one that deserves not only greater respect but also a quality that isn’t fully guaranteed with today’s frenetic pace.

This story originally appeared on WIRED Italia and has been translated from Italian.

A previously undocumented Linux implant named Quasar Linux (QLNX) is targeting developers’ systems with a mix of rootkit, backdoor, and credential-stealing capabilities.

The malware kit is deployed in development and DevOps environments in npm, PyPI, GitHub, AWS, Docker, and Kubernetes. This could enable supply-chain attacks where the threat actor publishes malicious packages on code distribution platforms.

Researchers at cybersecurity company Trend Micro analyzed the QLNX implant and found that “it dynamically compiles rootkit shared objects and PAM backdoor modules on the target host using gcc [GNU Compiler Collection].”

A report from the company this week notes that QLNX was designed for stealth and long-term persistence, as it runs in-memory, deletes the original binary from disk, wipes logs, spoofs process names, and clears forensic environment variables.

The malware uses seven distinct persistence mechanisms, including LD_PRELOAD, systemd, crontab, init.d scripts, XDG autostart, and ‘.bashrc’ injection, ensuring it loads into every dynamically linked process and respawns if killed.

QLNX features multiple functional blocks dedicated to specific activities, making it a complete attack tool. Its core components can be summarized as follows:

• RAT core — Central control component built around a 58-command framework that provides interactive shell access, file and process management, system control, and network operations, while maintaining persistent communication with the C2 over custom TCP/TLS or HTTP/S channels.
• Rootkit — Dual-layer stealth mechanism combining a userland LD_PRELOAD rootkit and a kernel-level eBPF component. The userland layer hooks libc functions to hide files, processes, and malware artifacts, while the eBPF layer conceals PIDs, file paths, and network ports at the kernel level. Both are deployed dynamically, with the userland rootkit compiled on the target system.
• Credential access layer — Combines credential harvesting (SSH keys, browsers, cloud and developer configs, /etc/shadow, clipboard) with PAM-based backdoors that intercept and log plaintext authentication data.
• Surveillance module — Keylogging, screenshot capture, and clipboard monitoring.
• Networking and lateral movement — TCP tunneling, SOCKS proxy, port scanning, SSH-based lateral movement, and peer-to-peer mesh networking.
• Execution and injection engine — Process injection (ptrace, /proc/pid/mem) and in-memory execution of payloads (shared objects, BOF/COFF).
• Filesystem monitoring — Real-time tracking of file activity via inotify.

After initial access, QLNX establishes a fileless foothold, deploys persistence and stealth mechanisms, and then harvests developer and cloud credentials.

By targeting developer workstations, attackers can bypass enterprise security controls and access the credentials that underpin software delivery pipelines.

This approach mirrors recent supply chain incidents in which stolen developer credentials were used to publish trojanized packages to public repositories.

Trend Micro has not provided details about specific attacks or any attribution for QLNX, so the deployment volume and specific activity levels of this new malware are unclear.

At the time of publication, the Quasar Linux implant is detected by only four security solutions, which flag its binary as malicious. Trend Micro has provided indicators of compromise (IoCs) to help defenders detect QLNX infections and protect against them.

AI chained four zero-days into one exploit that bypassed both renderer and OS sandboxes. A wave of new exploits is coming.

At the Autonomous Validation Summit (May 12 & 14), see how autonomous, context-rich validation finds what’s exploitable, proves controls hold, and closes the remediation loop.

Claim Your Spot