Feed aggregator

Welcome to the Cisco Live! 2025 live blog.

Here, we will be bringing you all the latest news and releases from Cisco Live! 2025 in Amsterdam.

We're looking forward to everything kicking off tomorrow at 9am with some exciting keynotes from Cisco's leadership.

So stay tuned, and we'll see you bright and early tomorrow morning for all the news from Cisco Live! 2025 as they happen.

• Marvel isn't actively looking to find another actor to play T'Challa in Black Panther 3
• Industry insiders have suggested the comic giant is planning to recast the character
• However, no discussions have taken place yet, according to one of its producers

Marvel isn't actively looking for another actor to replace the late Chadwick Boseman as T'Challa in its cinematic universe.

That's according to Nate Moore, who told TechRadar that the comic book titan hadn't discussed the prospect of recasting the character for Black Panther 3. Moore, who's among the film's many producers, said fans should ignore recent rumors that Marvel was finally ready to cast another actor as T'Challa and, in the process, hand the Black Panther mantle to another star for the third time.

Boseman passed away in August 2020 after privately battling bowel cancer. In the years since the beloved actor's death, Marvel has batted away questions that it would recast T'Challa in the Marvel Cinematic Universe (MCU). Indeed, speculation was rife about who would replace Boseman as T'Challa in the months prior to principal photography beginning on Black Panther: Wakanda Forever, the 2022 sequel to 2018's multi-billion dollar-making and 2019 Best Picture Oscar nominee Black Panther. As it turned out, Marvel decided not to recast T'Challa and, as part of Wakanda Forever's plot, passed the African nation's protector-in-chief role onto Letitia Wright's Shuri, aka T'Challa's brother.

Shuri adopted the Black Panther superhero mantle in Wakanda Forever (Image credit: Marvel Studios)

With Marvel now confirming Black Panther 3 is in early development – studio head Kevin Feige admitted it was in the wake of Denzel Washington accidentally stating he'd have a role in it – the recasting rumor mill has restarted. In January, MCU leaker Jeff Sneider claimed Marvel had offered the role to an unnamed actor, but they turned it down due to their concerns with filling Boseman's iconic shoes. MyTimeToShineHello, another insider with a questionable track record for leaking information about Marvel projects, has also suggested that Marvel wants to find an actor to play T'Challa's son (also named T'Challa), who viewers met in Wakanda Forever's mid-credits scene.

Chadwick Boseman appeared as T'Challa/Black Panther three times before his unexpected death (Image credit: Marvel Studios)

Moore, though, is clearly baffled by the hearsay surrounding any casting news about the studio's next Black Panther film. When I asked if he could comment on the rumors, Moore smiled and said: "There's no truth to those rumors.

"To be honest, Ryan [Coogler, director of the Black Panther franchise], is knee-deep on [his next movie] Sinners, so he hasn't really started cracking the story with us. So, it's funny to read all those reports online of the villains, casting rumors, and this and that. It's news to all of us.

"What I will say is that it speaks to how much the fans love that property," he continued. "We should be so lucky for fans to be interested in it and want to know what's happening. I can honestly say we haven't done anything, so any reports are false, but I do look forward to getting back into it with Ryan once he's ready."

It's funny to read all those reports online of the villains, casting rumors, and this and that. It's news to all of us

Nate Moore, Black Panther 3 producer

It'll be a long time before we see Black Panther 3, but it's possible that T'Challa's aged-up son could feature in one of the next two Avengers movies if Marvel wants to include him. Filming is reportedly expected to start on the first of those films – Avengers: Doomsday – in March, so maybe we'll learn more then.

Regardless of whether T'Challa's son appears in either or both films, there will be a new Black Panther production to enjoy later this year, with Marvel's Eyes of Wakanda animated spin-off show set to debut on Disney Plus. Per a Marvel trailer that unveiled its TV series line-up for 2025 last October, the Marvel Phase 5 program will be released on the streaming platform on August 6.

You might also like

• Find out how to watch the Marvel movies in order, including the first two Black Panther films
• 'Our goal is to make the best movie possible': Captain America: Brave New World director and producer address the Marvel film's extensive reshoots and poor test screening rumors
• Marvel debuts new Thunderbolts* trailer during Super Bowl LIX, and it reveals the MCU movie's main villain who you probably expected to be a good guy

• Researchers have discovered a new host material for blue OLED pixels
• Tests showed a big jump in power efficiency
• The breakthrough could mean cheaper production

Good news for anyone hankering for a cheaper OLED TV: a new scientific breakthrough could pave the way for cheaper, more efficient OLED displays in our smart TVs, phones, tablets, and pretty much everywhere else – though the impact could be largest on TVs.

Researchers at the University of Cambridge and University of Manchester have reportedly discovered a new ‘oxadiazine’ host material for creating high-efficiency, lower-cost, blue OLED pixels (via OLED-info).

Blue OLED pixels are the white (blue?) whale of TV panel development, proving much more difficult to produce with the same light-emitting efficiency as their red and green counterparts, with more risk of instability and a short shelf-life. As the abstract from the scientific paper says: “Energy-efficient and deep-blue organic light-emitting diode (OLED) with long operating stability remains a key challenge to enable a disruptive change in OLED display and lighting technology.”

However, it looks like headway is being made. The paper shows the molecular design of the new material, with 21% improvements in power efficiency and excellent “thermal stability” for use in blue OLED pixels.

And while all OLED devices use blue pixels, the reason that's especially good for TVs is that even the latest and greatest TV tech requires multiple layers of blue pixels, so that's where the biggest effects may be felt.

The Panasonic Z95B features LG Display's new four-stack OLED panel. (Image credit: Future) The future of OLED

There’s been a big push in recent years towards developing 'phosphor OLED' panels with greater power efficiency and higher brightness

This means that a TV screen can deploy fewer layers of blue pixels for the same brightness output as existing screens.

LG's latest 'four-stack' Primary RGB Tandem OLED panel (used in the LG G5 and Panasonic Z95B), for example, uses more blue layers than red or green. QD-OLED, as used in the Samsung S95F, relies heavily on blue light coming through a layer of quantum dots, so also requires several layers of blue pixels. So any method for reducing the complexity or cost of blue pixels is going to be very welcome.

While phosphor OLED has proved difficult to manufacture so far, these kinds of breakthroughs show that headway is being made in blue options overall.

A laboratory breakthrough like this will take time to impact mass panel production, of course, but it does bode well for the development of OLED in the coming years, and the hope for cheaper, more accessible OLED TVs that offer excellent picture quality for longer.

While OLED TV prices have fallen in recent years, they’ve largely stalled around the $1,000 mark, and manufacturers and consumers alike have been waiting for something to break the deadlock. Hopefully, that future isn’t too far off.

You might also like

• Sonos' streaming box sounds like a dream product, apart from the 2 horrendous flaws
• Is this designer OLED TV worth its explosive price tag?
• Philips' new elite OLED TVs let you choose between extra image quality or better built-in sound

Hamas said delayed the release of the next group of Israeli hostages, accusing Israel of preventing aid and Palestinians from returning home in Gaza. Israel said Hamas is violating the ceasefire.

(Image credit: Alexi J. Rosenfeld)

If you're looking for a Super Bowl LIX game recap, you've tapped the wrong link. I want to talk about this year's Super Bowl ads, or more specifically a pair of tech-related ads, especially the one that had me reaching for the Kleenex and put the ongoing AI wars into a fresh perspective.

I'm old enough to remember a time when the Super Bowl ads were just about Pepsi, McDonald's, Wendy's, Doritos, beer, and the newest cars. Super Bowl LIX was another reminder that these days nothing drives ad dollars and, perhaps, consumer interest like technology.

Woven in between oddball AI slots like recreating football icon Jimmy Johnson for a simultaneously warm and creepy tribute were a whole lot of tech ads. T-Mobile ran an ad introducing its new Satellite cell service, and Square Space confusingly employed indie actor Barry Keoghan to pitch personal website building.

For me, though, nothing generated as much interest, engagement and emotion, along with some head-scratching, as a pair of AI-focused commercials, one from Google for the Pixel 9 phone running Gemini Live and the other from OpenAI, plugging ChatGPT.

Black and white

OpenAI's minute-long ChatGPT spot, which TechRadar's John-Anthony Disotto has also written about. provoked the head-scratching. The lack of a voice-over, or color (beyond black and white), and use of pointillism-style dots to create imagery, made it almost impossible at first to divine whose commercial I was watching.

Granted, the music is sort of catchy and the imagery of things like flames, horses racing, corn blooming, ships sailing the high seas, trains racing toward you, and light bulbs is sort of compelling in a, "What the hell am I watching?" sort of way. But the key to a good Super Bowl ad is to grab people within the first 10 seconds, and OpenAI failed in that regard. I appreciated that the imagery was growing more refined and clearer – the dots kept getting smaller to represent ever-more complex imagery – but I still didn't get it.

More than halfway through, there was some historical audio from a newscast, the Apollo 11 mission to the moon, and even the iconic modem handshake sound (ask your parents). Finally, with just 20 seconds left, we hear someone ask an AI chatbot to summarize... er... something. The request audio started overlapping so I couldn't tell who was asking what.

One has to wonder if anyone at OpenAI even understands people.

The on-screen text says, 'All progress has a starting point.' Okay, sure. Whatever that means. Then you hear the more recognizable ChatGPT voice, and we see the first color of the entire ad, ChatGPT's blue voice chatbot icon, and we hear 'What do you want to create next?'

The spot ends with 'ChatGPT' in a large font, and then the OpenAI interwoven icon.

This is a commercial created by someone who thinks, "If they're smart enough, they'll get it. We're creating something new here. Let's be mysterious, charismatic, obtuse even. Everyone will love it and remember it."

I wonder if the only 'person' they showed this ad to was a version of ChatGPT, one of the models that can tell you about any photo or video you show it. Perhaps ChatGPT watched it and said "I see a series of dots and many moving images. This all represents ChatGPT (me)." Whoever created the ad likely fixated on the words "moving images" and assumed they'd struck a chord. And they did – one of confusion.

One has to wonder if anyone at OpenAI even understands people. Have they spent so much time in the bowels of ChatGPT's various powerful models that they've forgotten what it's like to be human, or even to talk to a human?

Moving images

The best Super Bowl LIX ads told a story: a horse rescuing a beer keg (trust me, it's better than it sounds); David Beckham discovering he has a secret twin (clever, funny) Matthew McConaughey showing how football might have been created just so we could eat more food. I'd argue that story-telling is a good rule of thumb for almost any TV commercial.

Google's Pixel ad, entitled 'Dream Job', was sneaky. It grabbed my attention from the start with a refrigerator featuring family shots and a guy clearly trying to find a job.

Google didn't try to hide the product or its intentions; the middle-aged guy is clearly holding a Pixel 9 phone and he's workshopping an interview pitch. From this point forward, though, Google cleverly blends product utility and emotion.

Google cleverly blends product utility and emotion.

The guy's pitch is dry and unemotional, but Google Gemini Live asks him a key question: "Tell me about the job that taught you the most." What follows is a walk back through the man's life as a stay-at-home dad, raising a little girl into a young woman. Each job experience he relates is actually about raising his daughter. When for example, we see an old video of him spoon-feeding his unhappy toddler daughter, he tells Gemini, "It was a role where I learned to take a lot of constructive criticism."

The whole two-minute spot is heartwarming.

Gemini eventually asks, "When it comes to work, what motivated you?" as the images flash back through his child's life right up to her high school graduation and his dropping her off for college. He answers, "I guess, knowing that people can depend on me."

Oh, but Google and Gemini are not done. Naturally, the daughter turns back and runs to the car to give her dad one more hug.

This is a commercial about AI, right? So why am I crying?! Through my tears, I see that the ad has circled back to showing Gemini complimenting the dad on his answer and telling him he's ready.

The ad ends with the words 'Google Pixel 9 with Gemini Live' and the dad beginning the real job interview.

Does Google get us?

Google managed to pull us along through a whole emotional journey while showing exactly how you might use its powerful AI chatbot. By focusing on our shared humanity, the ad made Gemini Live feel like something a human would use.

If you're a parent, the ad's theme probably hits you like a gut punch. Even if you're not, you'd be made of stone if it didn't move you

OpenAI had a big opportunity with its Super Bowl LIX commercial: it would reach at least 100 million people, and the ad's message would probably define how consumers thought about ChatGPT for at least the next few months, if not longer. The problem is, the ad OpenAI delivered gave us nothing to think about it. It was about as inhuman and dispassionate as HAL 9000.

Google, on the other hand, has been playing this game for longer (it ran its first Super Bowl ad 15 years ago) and understood the task at hand: connect with real people about something that is intrinsically artificial. Make them want to try it. Make them remember by imprinting an emotional note they won't soon forget.

Based on these ads, Google gets us, OpenAI does not. And that's why Google Gemini may ultimately win this AI race.

You might also like

• Google Gemini is racing to win the AI crown in 2025
• Google wants Gemini AI to be the star of every part of your life
• Gemini Live is here, free, and feeling chatty

• Microsoft could let you turn the Copilot key back into the old ‘Menu’ key
• That key brings up a context menu (like right-clicking with your mouse)
• This is useful in some scenarios, but many will want a fuller range of customization options for the key

Microsoft could soon give people whose keyboard has a Copilot key the ability to redefine it so a press of the key opens a context menu instead.

This is according to a well-known source of Windows-related rumors and happenings, PhantomOfEarth on X (formerly Twitter).

Future builds will add the option to remap the Copilot key to open a context menu.February 9, 2025

Remember, this is just a claim that Microsoft might do this in future test builds of Windows 11 – presumably based on clues found by the leaker digging around in current preview builds – and it still may not happen.

If none of the above makes much sense, and you’re scratching your head as to what’s potentially going on with the keyboard here, let us rewind a bit and explain.

What’s actually happening (or could be) is Microsoft is allowing the choice to revert the Copilot key back to what it was previously (on many laptop keyboards, anyway).

This old key that the Copilot button replaced was known as the ‘Menu’ key and it typically carries an icon with three horizontal lines (perhaps with a pointer, too), indicating that it’s used to summon the aforementioned context menu.

That context menu is the same one you invoke by right-clicking with your mouse, to give you options which are common actions in any given context (with files for example, you can click to see properties, or rename, and so forth).

Analysis: More choice is good, but…

(Image credit: Future)

Why would you want this old ‘Menu’ key that brings up the right-click context menu back, anyway? Well, it can be useful in situations where you don’t have a mouse (which is why this key is more often found on laptops) and so can’t necessarily right-click to bring up said menu. Microsoft used to have this key on its Surface devices, for example, between the ‘Alt’ key on the right and the arrow keys – but now it’s the Copilot key.

Not if you change it back by remapping, though – and if you find the context menu shortcut more useful than the Copilot key, well, apparently you might have your wish granted later this year. Albeit with the above caveats about this not even being in testing just yet.

Microsoft introduced the ability to remap the Copilot key to launch an app in preview (late last year). This move has since arrived in Windows 11, and you can switch the key to invoke a search, too, but remapping to an app comes with a notable (and annoying) catch that said software must be an MSIX-packaged application (not many apps are). This has been implemented that way for security reasons, in case you were wondering.

At any rate, a context menu option would at least be something, but I’m hoping that Microsoft will give us a lot more freedom eventually to redefine the Copilot key to do, well, anything we want (and act as a shortcut for any app at all, not just a limited selection). At least these all appear to be steps in the right direction for better customization, if only small strides.

Microsoft certainly appears to be giving up the notion that the Copilot key represented the most important introduction to the keyboard on Windows PCs since the Windows key itself.

Via XDA Developers

You may also like...

• Fed up with Copilot in Windows 11? Microsoft pushing a new keyboard shortcut for the AI isn’t likely to help, and it could mess with your desktop mojo
• Are you unable to get security updates for Windows 11 24H2? Here’s the likely reason why, and the fix to get your PC safe and secure again
• A Chrome-killing browser will be OpenAI's next big shot at Google, according to a new report

Sony blamed an operational issue for the 20-hour service outage, which affected console gamers on the PlayStation Network over the weekend.

Roborock Saros 10R: two-minute review

The Roborock Saros 10R is one of Roborock's 2025 flagship models, and proof if we needed it that robot vacuum cleaner innovation is advancing at a truly blistering pace.

Yes, the Saros 10R is expensive ($1,599.99), but that's the price you'll pay for any premium robot vac, especially when it's from one of the best robot vacuum brands around; Roborock is well known for consistently pushing the envelope of robot vac design, innovation and reliability.

This new flagship model can be considered a direct successor to the popular S8 MaxV Ultra and is available in two variants, the Saros 10R I'm reviewing here, which is equipped with a new type of Solid-State LiDAR that doesn't require a turret, and the cheaper Saros 10 which uses a retractable turret.

All you really need to know is that the Saros 10R is squat enough to zip under furnishings just 8cm in height while sporting a gorgeous looking mirrored glass dock, oodles of suction power, revolutionary StarSight LiDAR navigation and mapping, plus a fully integrated mopping system with full self-cleaning facilities.

If you have a cluttered home with a complex layout, lots of furniture and a surfeit of messy kids and pets in tow, this is the model – this autonomous household helper does pretty much everything but make the tea. That's the short version; read on for my full Roborock Saros 10R review.

(Image credit: Future) Roborock Saros 10R review: price & availability

• List price: $1,599.99
• Launched: February 2025
• Available: US

Although robot vac prices have generally dropped in price while offering far more tech for your money, the very latest premium models will always cost in excess of a grand, at least for the first few months of their shelf lives. That's certainly true of the new Roborock Saros 10R which retails at an eye-watering $1,599.99.

Yes, there are other high-end bots out there that offer generally similar performance to the Saros 10R – the Dreame L40 Ultra for instance – but I haven't come across another model that does quite as much as this one, including its ability to negotiate the deeper recesses of the home.

The Saros 10R launched in February 2025 and is available in the US only. As of writing, it's not expected to make it to UK shores – it does, expect it to arrive around March 2025. Assuming it's roughly the same price in pounds as it is in dollars, it would be around £1,290.

If you live in the US, you can purchase the Saros 10R direct from Roborock, or via Amazon. Best Buy also sells Roborock robovacs, so it may also become available there.

• Value for money score: 4 out of 5

Roborock Saros 10R specs Roborock Saros 10R review: design

• Impressive vacuum and mopping features
• Extremely stylish mirrored dock
• Expert navigation, mapping and obstacle avoidance

Most robovacs these days are equipped with a LiDAR turret that protrudes about half-an-inch above the body of the bot. This not only makes the robot look a bit ugly but, more than that, it prevents the bot from passing under low-slung furniture like cabinets and some sofas and beds.

By stark contrast, this stunning, black-only hybrid model from Roborock eschews a turreted LiDAR in favour of a flush, newfangled solid-state LiDAR navigation system (as opposed to rotating LiDAR, where the laser beam rotated 360-degrees). This not only takes navigation performance to a whole new level, but also allows the robot to venture beneath furniture with just an 8cm gap to the floor.

In fact, at just 3.14" (7.98cm) in height, the Saros 10R is the first robot vacuum I've ever tested to make it under my TV cabinet with just one millimetre of clearance. What's more, having a robot vac with no LiDAR turret means no more issues of a spinning LiDAR module sticking on a loose strand of pet hair – it happens. And of course, it also gives the unit a much more sleek and stylish appearance.

Navigation features

Aside from raw suction power and overall cleaning performance, a robot vac is only as good as its navigation system. As alluded to above, the majority of modern robots use two-dimensional LDS (Laser Distance Sensor) LiDAR and it serves them well. However, Roborock has thrown that book out of the window and has equipped this model with StarSight Autonomous System 2.0, a Dual Transmitter Solid-State LiDAR system that offers '3D Time-of-Flight scanning of the home for expert navigation around and under low-lying furniture, and enhanced mapping by capturing detailed height and depth information'. I'll take that, whatever it means.

The Saros 10R also features front-mounted RGB and infrared cameras for high-resolution visual recognition, side-mounted VertiBeam Lateral Obstacle Avoidance for moving around irregular-shaped furniture, walls and unfurled cables, and a front-mounted LED headlight for improved navigation in dark spaces. There is also another infrared camera and transmitter on the rear, plus the usual array of downward-facing drop sensors to stop it from falling down a flight of stairs, no matter which way it's facing. (Here's more on how robot vacuums navigate.)

To give you some idea of how far ahead of the game this robot's LiDAR system is, Roborock says that its 3D StarSight LiDAR system produces over 21,600 sensor points as opposed to standard two-dimensional LDS LiDAR, which produces just one. Consequently, this potpourri of navigation systems makes the Saros 10R ideal for cluttered homes with irregular layouts, a surfeit of furnishings and a lot of obstructions like toys, shoes and other household items. As of writing, the Saros 10R's obstacle avoidance system can recognise up to 108 household objects.

Bot design

That's Roborock's pioneering navigation system out of the way, so let's take a look at the mechanical stuff starting with this model's suction power. According to Roborock, this model sports 20,000 Pascals of 'HyperForce' suction power and that's about as strong as it currently gets (the upcoming Eureka J15 Max Ultra will apparently offer 22,000Pa, but that's the highest we've seen). It means its suction will reach deep down into medium and deep-pile carpet, removing a lot of obstreperous dust particles in the process. This model also includes a feature called DirTec that automatically increases vacuum power whenever heavy soiling is detected.

I should point out that raw suction power alone isn't the be-all and end-all of a vacuum cleaner's performance, since much of the work is also performed by the brush roller. And I think Roborock has nailed it with its DuoDivide anti-tangle main brush – a rubber two-piece corkscrew design that shuffles pet hair towards the centre where it's instantly sucked into the robot's slightly larger-than-average 270ml internal bin. This brush head will also raise by a few centimeters when heading to and from the dock to avoid any unnecessary contact with the floor.

Along with the main brush bar, the Saros 10R is also fitted with an anti-tangle side sweeping brush that pops out of the side by a couple of inches so it can reach into corners and around furniture legs. This side brush also lifts when not required, like when on carpet.

(Image credit: Future)

Like the Roborock Qrevo Curv I reviewed a few weeks ago, the Saros 10R can negotiate thresholds, small steps and floor-mounted tubed furniture legs up to 0.39 inches (1cm) in height. However, the Qrevo Curv's chassis can raise much higher so consider that machine instead if you have tall thresholds.

I'm personally not a big fan of robot vacs that also wet mop the floor, but that seems to be the 'in' thing these days, especially when it comes to the mid-to-upper range of robot vacs. Sure enough, this one comes with a full bells-and-whistles mopping system.

Like most modern hybrid robot vacs, the Saros 10R uses twin spinning mop pads for cleaning hard floors and, like the Qrevo Curv and some Dreame bots, the right-hand side mop articulates outwards for edge cleaning along skirting boards and around furniture legs. Similarly, the mops lift by 10mm when traversing carpets. However, since the mops are attached magnetically, they can also be left in the dock whenever the robot is assigned to a vacuum-only task so that there is zero chance of a damp mop running over even the shaggiest of rugs and carpets.

You get three levels of water flow with this model's mopping system – low, medium and high – but I would suggest sticking to low if used on wooden flooring. Furthermore, I can't vouch for the safety of using this model – or any other hybrid bot for that matter – if your floors are made from delicate laminate since thin laminated flooring is easily warped if water seeps into the joins. (Here's more on how to clean hardwood floors.)

Like the vast majority of Roborock robot vacs, this one also offers its own built-in voice control, if that is your wont. What's more, the voice control continues to work even if your internet system is offline. Alternatively, you can program it to respond to Google Home and Amazon Alexa prompts. Likewise, the robot itself has a voice and it will often pipe up to let you know what it's doing before and after a task, including announcing when it's about to empty its bin into the dock's larger dust bag.

All robot vacs I've tested have either used Chinese or American-accented voices, which is fine. However, with this model – indeed all Roborock models – there's also the option to download British-accented voices (male or female) which makes ownership in the UK a little more personalized (if this model ever arrives in the UK…).

(Image credit: Future)

As if you haven't already absorbed enough details for one day, the Saros 10R also provides some cool tech for your pets. For instance, you can remotely drive the robot around like an RC car or select Pet Search and it will look for your dog, cat or gerbil and take a picture of it. You can also use its built-in microphone and speaker to chat with your furry friends, but I would practice this first while you're at home in case they freak out the first time you use it remotely.

Dock design

I recently wrote a feature on robot vac innovations I'd like to see and chief amongst these was a request for better looking robot docks that you wouldn't mind having in the sitting room if space elsewhere was limited. By bizarre coincidence, this model arrived a week later sporting a cleaning dock embellished with a tempered glass front fascia replete with a stunning black mirror finish that reflects its surroundings – especially wooden floors – so well that you hardly notice it sitting there. Thank you Roborock.

At 18.7" (47.5cm) in height, 15" (38cm) in width and 19" (48.8cm) in depth (if you include the ramp), the Saros 10R's charging and self-cleaning dock isn't especially small, so you will need to find an adequate spot to locate it with several inches of space on either side so the bot can dock without difficulty.

(Image credit: Future)

Like most modern hybrid robovacs, the Saros 10R's dock will automatically empty the contents of the robot's small 270ml bin into its much larger two-liter dust bag every vacuuming session and sometimes mid task if requested to do so in the app.

For the mopping system, it comes with two water containers – 3 liters for clean water and 3 litres for the dirty stuff – which you will have to attend to from time to time by simply emptying the dirty water into a sink or toilet and topping up the other tank with fresh water. This model also comes with a detergent dispenser that can be filled with your own choice of hard floor detergent, though Roborock, like any floor cleaning product manufacturer, will naturally advise against this so you're persuaded to buy their own brand.

(Image credit: Future)

There's no shortage of sci-fi tech in this dock when it comes to cleaning the mops. On returning to the dock after a mopping run, the dock's scrubbing mechanism washes both mops using hot water at a very effective 176F (80C). If its intelligent dirt detection system senses excessive dirt in its cleaning tray, it will rewash the mops to ensure near perfect cleanliness for the next session. It will then proceed to dry the mops using wafts of warm air at 131F (55C). When it's time for its next mopping mission, the dock will automatically fill the robot's small water reservoir with warm water for improved stain removal.

Some docks I've seen also come with a self-cleaning feature that injects water into the mop tray. However, in most instances you need to get down on your knees with a brush to help move the dirty water towards a suction tube. By contrast, the Saros 10R's self-cleaning system does it all for you by scrubbing the tray in hot 176F (80C) water before porting the dirty stuff up a small pipe positioned at the rear. To activate the self-cleaning function, simply go to Dock Settings in the Roborock app and select 'Self-Cleaning for Cleaning Tray'. I should add that it's worth checking the cleaning tray from time to time to remove any excessively large pieces of debris that are sometimes collected by the mops and the robot's brush rollers. The two-part tray is easily cleaned with a brush and some hot tap water.

If you have a smart modern kitchen and you like your appliances to be fully integrated, you can also request this model as a svelte plumbed-in version that automatically fills the fresh water tank while emptying the contents of the dirty water container into a drainage outlet.

• Design score: 5 out of 5

Roborock Saros 10R review: performance

• Industry-leading navigation
• Great vacuum performance
• Excellent dock functionality

There's a lot to get into on the performance front here, so I've broken this section up into sub-sections. I'll start with that pioneering navigation, then look at obstacle avoidance, vacuuming performance (including the results f our suction tests), mopping performance, and finally dock performance.

Navigation performance

As mentioned above, the Roborock Saros 10R uses StarSight to navigate, an innovative system comprising a turret-free LiDAR, front-facing RGB and infrared cameras, and all sorts of other sensors to help it maneuver through even the most complex of floorspace layouts.

I've had it running for the past few weeks and at no point has it failed to find its dock or simply stopped mid-task for no reason. Mind, that's been the case with most Roborock models I've reviewed. Its mapping is exemplary and you can watch it all unfold in real time on the app's map screen.

Like a whirling dervish, it trundles off on its mapping run – an advised process when using it for the first time – and completes the task with impressive speed. In fact it took just five minutes for it to map my open-plan living space comprising five different areas. Yes, it merged two areas together as one room but even that was easy to rectify, simply by going into the app and dividing them into two separate rooms.

One of the most common anomalies with LiDAR-equipped robot vacs is their refusal to punch through a bed or sofa valance. This is because LiDAR treats a fabric valance as a solid boundary. Well I'm pleased to say that the Saros 10R is the first LiDAR-equipped robot vac I've ever tested that's capable of penetrating the valance on my lounge sofa.

Rather confusingly, it doesn't do it every time it's vacuuming the lounge so perhaps it's something to do with its angle of approach at the time. That's the thing with robots; they sometimes behave in mysterious ways. Either way, I'm very happy with this discovery.

(Image credit: Future)

I also tried to replicate a threshold using books but the Saros 10R isn't designed for anything much higher than 0.39 inches (10mm). Unlike the Roborock Qrevo Curv, which scaled the same book experiment like an off-road tank, the Saros 10R made an attempt to ride over the books but then thought better of it and retreated. Hence, if you have tall thresholds up to 1.5" (4cm) in height, Roborock's Qrevo Curv model would be a better bet.

Obstacle avoidance

Most robot vacs I've tested haven't been that great at small obstacle avoidance, especially when the objects have been less than an inch and a half in height. Sure enough, the Saros 10R at first refused to recognize one specific obstacle – a one inch-high nylon dog bone that's been too low in height for most robot vac's cameras and obstacle sensors to detect. Like the majority of bots, it basically rode over the bone and then struggled to extricate itself by moving hither and thither while one of its wheels raised the whole unit up in the air at a cock-eyed angle.

(Image credit: Future)

I checked the Roborock app to see if I'd missed a setting and, lo and behold, I discovered that I hadn't activated the 'Pet' option that 'recognizes and avoids pets and their waste with precision'. I remember seeing this option when I first set it all up but the 'avoids pets' explanation sounded unnecessary to me, since most pets are larger in size than any small obstacle. So I originally left the option off. In hindsight, I think this function in the app needs to be amended to just say 'avoids pet waste' to circumvent any confusion, because once I'd activated it the Saros 10R became a different beast altogether.

(Image credit: Future)

In fact, it consistently avoided said bone every time I dropped it in its path. Sometimes it would avoid it by about six inches and other times come perilously close but deftly work its way around it. I should add, however, that it may not clean as closely to the object when the pet option is selected. Swings and roundabouts spring to mind. Would I rely on this model to reliably avoid all pet waste all the time? No way, José. I've been down that road already and don't wish to go there again.

Vacuum performance

The Saros 10R vacuums like a trooper and comes with five power bands – from Quiet to Max+. I have my test model set between Max and Max+ since I see no reason to select anything less, given that the bot has enough battery to complete an entire sweep of my home in either of these modes. And besides, even if it were to run out of juice it will simply return for a full recharge – which takes about two and a half hours – before heading out again to complete the task.

Unlike non-LiDAR robots like those from iRobot, this model vacuums in a systematic pattern that makes a lot of sense, especially when you're watching it work. I also like the way the side brush pops out of the casing whenever the robot approaches a corner. It does a fine job of flinging debris into the path of the roller brush. This brush also lifts off the ground when the robot is vacuuming rugs.

Aside from its generally excellent suction performance, a quick shout out to the developers for coming up with such an effective two-piece corkscrew-shaped brush roller because at no time has any pet tangled on it. Instead, any hair is cleverly coaxed towards a centre gap between the rollers and into the mouth of the suction area. In this regard, I can wholeheartedly recommend this model as one of the best robot vacuums for collecting pet hair.

For my hard floor vacuum test I scattered raw oats, flour and half a crushed Hobnob biscuit and set it to clean a small zone in the kitchen using the Max setting and two passes for good measure. As expected, the robot disengaged its mops in the dock and headed straight to the specific area I drew on the map. It duly sucked up all the larger pieces of detritus but did leave traces of flour in the floor seams.

Image 1 of 3

(Image credit: Future)Image 2 of 3

(Image credit: Future)Image 3 of 3

(Image credit: Future)

So I reprogrammed the test, this time using Max+ mode and it collected a bit more flour than before but still left some evidence in the cracks. I personally don't consider this a fail because no robot I've tested has ever completely removed flour from the joins in my wooden flooring. And anyway, in the real world one wouldn't rely on a robot vac to perform a one-off sweep of a specific area when you should, by rights, have a cordless stick vac to hand.

Image 1 of 2

(Image credit: Future)Image 2 of 2

(Image credit: Future)

Test two involved the same ingredients on the medium pile rug I have in front of the dogs' beds, with the Saros 10R set to Max+ mode and two passes. Aside from some light evidence of flour on the rug that was easily finished off by the Dreame R20 stick vac I recently reviewed, I still rate this test a success. I was especially impressed with the way it moved both vertically and laterally across the rug to ensure full coverage.

Image 1 of 2

(Image credit: Future)Image 2 of 2

(Image credit: Future) Mopping performance

As I've mentioned above, I'm not a fan of mopping robots, mostly because I'm lucky enough to have a cordless wet-and-dry hard floor cleaner which I find is more than enough for occasional mopping duties. Nevertheless, the industry seems to have decided that mopping is what people want so, along with observing its bi-weekly mopping runs, I also put this model through a specific mopping test in the kitchen.

Using a hotchpotch of wet ingredients – soy sauce, syrup and a teaspoon of milk – and a pre-set zone routine in the app using the low water setting, the Saros 10R's dock filled the unit's small onboard water container with a mix of warm water and detergent, and primed its two spinning mops. It then trundled off to its predetermined zone.

Image 1 of 2

(Image credit: Future)Image 2 of 2

(Image credit: Future)

Well blow me down with a feather if it didn't clear the floor of all evidence in a single sweep. This really impressed me because I at least expected to see a smear of syrup. That's a fine result in my book.

Dock performance

I've only had one minor issue with the dock and it occurred directly after a family Christmas with five dogs and three cats in attendance. Once everyone had gone home, I unboxed the Saros 10R and set it to work in the lounge. This room hadn't been swept for three days and the Christmas tree had dropped a lot of pine needles.

Cutting to the chase, the unit duly swept up as well as anyone could reasonably expect from a high-performance model of this nature but when it proceeded to dump the contents of the robot's bin into the dock's large 2-litre dust bag, the hair clogged up the works. A quick check of the robot's bin showed an enormous conglomeration of pet hair and pine needles protruding from the bin's outlet port and another clump that had blocked the inlet tube leading to the dock's dust bag.

I should add that every robot I've tested has had at least one issue like this – and even my Dyson Gen5detect has jammed up from time to time – so I put it down to rare circumstances and a combination of matted pet hair and pine needles. I'm personally not sure if the Saros 10R's dock suction is as strong as the Qrevo Curv's because it sounds quieter, which is a good thing. Nevertheless, I'm very pleased to report that the dock has experienced no further problems during the past four weeks I've been using it.

(Image credit: Future)

In terms of self cleaning the mops, it did them well enough in my opinion given the amount of muck on the floor. However, I would always advise removing them and giving them a good scrub under a hot tap with some washing-up liquid for good measure. By the way, I especially liked the way the dock cleaned the mops' tray using water and a revolving mechanism to sweep the dirty stuff towards a small rubber suction tube.

(Image credit: Future)

Although Roborock advertises up to seven week's worth of floor matter collection before its dust bag requires binning, I am having to change the bag every three weeks – but then I do have a large menagerie of furry creatures. Hence, I'd like to see some bigger dust bags going forward.

In conclusion, I have been really impressed with this robot vac's overall performance and I find it hard to give it anything less than five stars in this category. It's just so reliable and efficient at the tasks it was designed to do.

• Performance score: 5 out of 5

Roborock Saros 10R review: app

• User friendly
• Superb customization
• Impeccably clear maps

Judging by the Reddit forums, I'm not the only one to think that the Roborock app is the best on the market. This app is just so customizable, intuitive and informative to use. Firstly, it produces a highly detailed map of the home with evidence of every nook and cranny. And you can edit the map to your heart's content, whether that's dividing or merging rooms, or adding keep out zones and virtual barriers. Aside from being able to create an endless range of timed schedules, you can also assign specific zones you want cleaned on a frequent basis, and save them as routines that appear on the home screen for convenience.

(Image credit: Future)

The app also offers something called SmartPlan as an option when cleaning. This feature supposedly weighs up the tasks it's previously performed and adjusts its cleaning method accordingly. Hence, if, say, the robot mopped the kitchen yesterday it might avoid doing it today. I personally leave it off since I don't want it doing anything I don't want it to do.

This app also features a full smorgasbord of custom features, including settings for floor cleaning, the dock, the robot's reactive obstacle avoidance, maintenance, remote viewing, robot settings, the list goes on. Furthermore, each menu you select brings up a range of individual tweaks with explanations that you could spend many happy hours fiddling with.

• App score: 5 out of 5

Should you buy the Roborock Saros 10R? Buy it if...

You want unbridled efficiency

This bot performs all tasks with consummate skill.

You want hands-free operation

The Saros 10R's self-cleaning features are a key selling point – the dock will not only empty the dust cup and wash and dry the mop pads, but clean itself too.

You need versatility

This bot is designed for complex environments – layouts with a mix of floor types and lots of obstacles.

Don't buy it if...

You're on a budget

The Saros 10R is expensive with a capital E. And while the premium features and tech help justify that asking price, if you don't need the very latest tech then there are plenty of options that'll do a very good cleaning job for a lower price.

You have very tall room thresholds

While the Saros 10R is pretty good at getting over raised thresholds, it's not as good as the Roborock Qrevo Curv.

You don't require a mopping function

Mopping (and the related mop dock functions) bump up the cost of this model considerably, so if you're not fussed about that then look for a vacuum-only model and save some cash.

How I tested the Roborock Saros 10R

I received my test model of the Saros 10R just before Christmas and set it up before the New Year 2025. For the following month I ran different daily schedules in both vacuum and mopping modes. It has been one of those robot vacs that just gets on with the job without making a song and dance about it.

However, for purposes of this review, I also put it through three specific TechRadar tests involving both vacuuming and mopping just to make sure it was as efficient as it seemed. As you will have read, it sailed through everything without any major hiccups. I should add, however, that that was the least I expected from a brand which has a reputation for producing great gear that just works.

Read more about how we test robot vacuum cleaners.

• First reviewed February 2025

• Hospital Sisters Health System files new report with the Maine Attorney General
• It confirmed more than 800,000 affected in an August 2023 breach
• Compromised people are getting a year's worth of free identity theft monitoring

Hospital Sisters Health System (HSHS), a nonprofit, Catholic healthcare system, suffered a cyberattack one and a half years ago, which resulted in the theft of sensitive patient data.

The firm has now filed a report with the Maine Office of the Attorney General, in which it detailed the attack, noting it discovered an “unauthorized third party” gaining temporary access to its network, on August 27, 2023.

“Upon learning of the situation, we immediately took steps to contain and remediate the incident and launched an internal investigation,” HSHS said in the filing.

Stealing sensitive data

The investigation determined that the unnamed attackers dwelled on HSHS’ network between August 16 and August 27, and during that time exfiltrated sensitive information belonging to exactly 882,782 people.

“We have since been reviewing those files and notifying individuals whose information was found in the files on a rolling basis as our review has continued,” the organization said.

While the type of information stolen varied from person to person, in general it included full names, postal addresses, birth dates, medical record numbers, limited treatment information, health insurance information, Social Security numbers (SSN), and driver’s license numbers.

This is more than enough to engage in highly personalized phishing, identity theft, or even wire fraud. However, HSHS says that at this time it has “no reason to believe” the data has been misused.

Healthcare information is highly sought on the black market because it contains sensitive personal, financial, and medical data that can be exploited for various types of fraud and cybercrimes. Unlike credit card data, which can be quickly canceled, stolen medical records provide long-term value as they include Social Security numbers, insurance details, and medical histories that can be used for identity theft, fraudulent billing, prescription fraud, and even blackmail. Additionally, the resale price of medical records is significantly higher than financial data due to their completeness and difficulty in detection.

That being said, even though there is no evidence of misuse, “out of an abundance of caution”, HSHS offered affected individuals a year’s worth of credit and identity theft monitoring through Equifax.

Via BleepingComputer

You might also like

• United Healthcare data breach may have affected 190 million Americans
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

Nintendo has done what it does best when it comes to the newly-announced Nintendo Switch 2; take an excellent piece of hardware and tweak it until it peaks, throwing in a few feature gambits along the way.

Of course, the strategy relies on the hardware being great enough to warrant a new entry (see the Virtual Boy) to begin with, and they don't always stick the landing (see the Wii U and the Nintendo 64DD).

It always surprises me, however, to see how risk-averse gamers are when it comes to Nintendo's hardware. Yes, there have been some blunders, but broadly speaking Nintendo has always been among the most creative hardware manufacturers when it comes to creating consoles that inspire new games and new ways to play.

We've had our first glimpse at the Nintendo Switch 2, and while many of the new features are shrouded in mystery, and will remain so until the Nintendo Direct: Nintendo Switch 2 event on April 2, I think people have already grossly misunderstood what will make the Nintendo Switch sequel an excellent follow-up to one of my all-time favorite consoles: its rumored new mouse mode.

From a new matte coating to an additional USB-C port, a lot of long-rumored features were confirmed by Nintendo's Switch 2 announcement trailer. Among the most exciting, however, were the mysterious new system button, larger size, magnetic Joy-Con, and an improved kickstand that could finally make tabletop play more viable for the console.

The standout for me, however, was what appeared to be confirmation that the new Joy-Con will offer mouse-like control, opening new possibilities for PC game ports as well as new ways to play upcoming titles on the Switch 2.

Among the first games I purchased for the original Switch was an all-time favorite of mine; Sid Meier's Civilization VI. My excitement was short-lived; a game with this many fiddly interfaces on such a small screen just didn't gel with me, and I found myself crawling back to my gaming PC for the more enjoyable mouse input.

Of course, questions still remain; how comfortable and ergonomic can a Joy-Con be to use in this orientation? How smoothly will it work on different surfaces? Will the new and (hopefully) improved Joy-Con be robust enough for regular mouse mode use, or will we see degradation like the first-generation Nintendo Switch Joy-Con drift?

We won't know the answers to these questions until we've had hands-on time with the console, but one thing that was never a question in my mind was whether or not users would be a fan of the new feature.

(Image credit: Nintendo) Put some respect on mouse mode

Turns out, not everyone is as excited as I am for the new mouse mode, and I can't really fathom why.

We surveyed our WhatsApp followers to find out which new rumored features they were most excited about, and I was pretty surprised by the results.

The vast majority, understandably, said they were reserving their excitement until pricing has been confirmed. The most we know so far on that comes from Nintendo president Shuntaro Furukawa, who shared in a recent Q&A that the Switch 2 would be in line with the "affordable prices customers expect". TechRadar Gaming predicts that the cost of the Switch 2 will be around $399.99 / £349.99 / AU$699.95, or lower, which is excellent news, if true.

The next-most popular answers were a tie, as of writing; unsurprisingly, the larger screen (and Joy-Con) ranks highly among prospective Switch 2 owners, but I was floored to see the mysterious new right Joy-Con system button take the lead over mouse mode.

(Image credit: Future / Meta WhatsApp) Pushing imaginary buttons

For one thing, we have no confirmation whatsoever as to what this button does. Nintendo gave us some solid visual cues for many of the new additions, including mouse mode, but we saw nary a click of the new button.

Rumors have been whirling for months since Joy-Con prototype images leaked online about what this button may be for, especially given earlier iterations featured a 'C' label which birthed all kinds of rumors. Does the 'C' stand for capture, or chat, perhaps? Or was it simply a label for prototype units?

Its placement suggests that it's more likely a system button than one used in gameplay, and given there's still a capture button for screenshots and short gameplay snippets, I'd hedge my bets that it's likely either for in-game text chat or quick control for voice chat.

And sure, that's fun; I used the Nintendo Switch Online-exclusive voice chat feature a lot during the pandemic when I was in the throes of my Animal Crossing: New Horizons grind, and shared frustrations with many Nintendo fans that online gaming features have been significantly lacking in the current-generation console.

Still, I fail to see how this could possibly inspire excitement more than Joy-Con mouse control.

(Image credit: Nintendo) Keeping it casual

In the weeks since the announcement, I've seen countless comments from readers and viewers seething with unfounded rage that the hardware won't be on par with other handheld gaming devices, comparing the Switch 2 to the likes of the Lenovo Legion Go, the Steam Deck, and the Asus ROG Ally X; which is about as fruitful of a comparison as apples to oranges.

Even when it comes to mouse control, it's a false equivalency; the Steam Deck trackpads offer a different form factor and experience, as does the Legion Go's FPS mode. The simple fact of the matter is we won't know until we've tried it, but I for one am incredibly excited to see how Nintendo might use the feature to allow for more whacky, whimsical, and wonderful games on the Switch 2.

Consider the success of the Wii and its motion controls. Were they new to the market? No; at the time, motion control was all the rage in gaming. What Nintendo did was make it more accessible, opening the market out to families and casual gamers with a piece of moderately priced hardware.

That is where Nintendo thrives. Performance and power are never the focus, and it boggles my mind when I so frequently see gamers misunderstand and misrepresent the very clear – and successful – price point strategy Nintendo opts for.

However, they do need to make games cheaper; I'll give the haters that one.

You might also like...

• The best Nintendo Switch games to play in 2025
• Nintendo is already preparing for Switch 2 scalpers ahead of the console's release
• An all-digital PS6 doesn't seem possible, according to ex-PlayStation boss Shawn Layden: 'I think it would be hard for them to go fully disc-less'

Moore’s law, named after Gordon E. Moore, the co-founder of Intel, states that the number of components on a single chip doubles every two years at minimal cost. It’s not exactly hard science, but his simple observation of how rapidly chip technology was progressing back in 1965, has proved to be pretty useful for predicting where technology will be in the years to come.

In his latest blog post, Sam Altman, CEO of OpenAI, makers of ChatGPT, just stated that ‘ Moore’s law changed the world at 2x every 18 months; this is unbelievably stronger. ‘ Altman was referring to the massive price drop we've seen for AI usage. For developers, this is measured in token cost, and the token cost to use AI has been falling about 10x every 12 months. Between early 2023 to mid-2024, the price per token for using ChatGPT has dropped about 150x.

AGI is coming

Technically, Altman got it wrong by writing “18 months” instead of “about two years”. The 18 months figure is a common misquote since there was a separate prediction by Moore’s colleague, Intel Executive David House that Moore’s Law would mean that chip performance would double every 18 months, with no increase in power consumption. So, a forgivable slip.

However, Altman’s point is that the falling cost of using AI is another indicator that artificial general intelligence (AGI) is “coming into view”. AGI is an artificial intelligence that is on a par with, or smarter than, human intelligence, and developing it is the whole reason that companies like OpenAI exist because the benefits of AGI could be truly world-changing, despite the dangers that we keep getting warned about.

It’s worth reading the whole of Sam Altman's blog post because it very much reads like a warning that the future is coming sooner than we think and that we really start to need to prepare for how the world is going to change. Altman’s predictions are certainly mind-blowing. His three key predictions for the immediate future are:

• Scientific progress will likely be much faster than it is today
• The price of many goods will eventually fall dramatically
• The price of luxury goods and a few inherently limited resources like land may rise even more dramatically

Altman doesn’t predict these changes happening in 2025, but since he’s already gone on record saying that humanity will achieve AGI this year, the changes would appear to be right around the corner.

His message is also that we have some tough choices coming up regarding AGI and how freely it should be used, because one possible future, “we can see is AI being used by authoritarian governments to control their population through mass surveillance and loss of autonomy.”

In Altman’s vision, ensuring that the benefits of AGI are broadly distributed is critical to prevent this from happening.

He ends with a quite astonishing prediction: “Anyone in 2035 should be able to marshall the intellectual capacity equivalent to everyone in 2025; everyone should have access to unlimited genius to direct however they can imagine.”

It’s hard to know how to process the breadth of Altman’s vision right now. Companies like OpenAI are committed to developing AGI, but the dangers inherent in who controls this technology are something that Altman is urging us to start considering and talking about right now.

You may also like

• 'AI Godfather' sounds the alarm on autonomous AI
• Sam Altman predicts artificial superintelligence (AGI) will happen this year
• Google is bragging about Gemini 2.0 Flash, but how does it compare to ChatGPT for casual AI chatbot users?

Artificial intelligence (AI) has great potential to revolutionize supply chain operations by enhancing demand forecasting, optimizing inventory management and streamlining logistics, to create more resilient and agile operations. However, to gain a competitive advantage, many companies are implementing AI before ensuring their supply chain systems are secure.

Without the proper data safeguards in place, AI-enabled systems can lead to significant vulnerabilities, including data breaches, biased decision-making and increased cost. These vulnerabilities can disrupt critical operations, damage reputations, and erode customer trust, causing long-term harm to the business. According to Gartner (hyperlink first link in link section), “Attacks on the software supply chain, including both proprietary and commercial code, pose significant security, regulatory and operational risks to organizations. A source estimates costs from these attacks will rise from $46 billion in 2023 to $138 billion by 2031.”

Organizations that are not already doing so should consider using cybersecurity as criteria for third-party transactions. This is especially important as items like order details, payment information and other confidential details move throughout a global supply chain – reaching third party suppliers and logistics partners.

To protect data integrity and maintain customer trust, organizations should approach AI with a security-first mindset. These three tactics can help.

1. Conduct a risk assessment to understand data security

A risk assessment will identify the systems and intellectual property that are most valuable to your company. Additionally, the risk assessment will identify what data is critical to your organization and where it’s located. Once defined, cybersecurity risks can be prioritized based on the likelihood of occurrence and the impact on your business. Then, your list of prioritized risks can be used to establish effective response protocols.

By understanding how cyber-attacks can impact your business, you will be able to recommend time-sensitive actions to be used in a multi-layer defense protocol. Think of it like breaking into Fort Knox: if a malicious actor breaks through the first line of security and no other safeguards are in place, they have free reign to cause unmitigated damage. However, with a multi-layer defense response in place, you can reduce the harm.

While considering data, it is important to understand how data flows, and where it is located. For example, items listed on Amazon’s marketplace could be sold and shipped by third parties. When this happens, important customer data is routed to the seller/supplier, who may have less secure data security, which could increase the threat of a breach or other attacks.

2. Implement security testing to develop immediate response protocols

Cyber-attacks are constant and unpredictable. As a result, constant monitoring and testing is necessary. When vulnerabilities are detected, automation can trigger predefined responses – such as isolating infected systems, notifying personnel and initiating remediation processes to ensure that future attacks are prevented.

Scanning tools can monitor transactions, communications, and activities in the supply chain to identify suspicious behavior faster. By scanning software, hardware and network infrastructure, companies can identify weaknesses, like unpatched systems that could be exploited. These efforts help build a more resilient and adaptive defense against malicious actors. Through proactively identifying and patching weaknesses before they are breached, many intrusions can be reduced or eliminated.

In addition to scanning, you can use penetration testing to locate vulnerabilities and red teaming to simulate offensive attacks, to see how your systems would respond to potential cyber-attacks. Together, these tests will uncover hidden weaknesses and assess the effectiveness of your current security.

3. Promote a security-first culture

To protect your company, your employees, and your customers, it is critical for companies to make security a cultural pillar. Companies should provide training to ensure that employees are aware of its importance and are actively engaged in maintaining high security standards.

In a supply chain where sensitive information travels between numerous parties, every employee and vendor should be considered as being a part of the security process. Efforts to foster a security-first environment include training on security threats, simulated phishing attempts with incentives for employees who demonstrate best practices, or the development of a security policy that includes IT, legal and procurement team members at every stage of the supply chain.

By promoting a security-first culture, you increase overall awareness and vigilance, making every individual a part of the security solution. A top-down commitment ensures that the right resources are allocated to cybersecurity, reducing the likelihood of lapses that can compromise the supply chain.

Putting it all together

With assessments that highlight areas of vulnerability, automated testing and monitoring, and championing a security-first culture, companies can cover nearly every phase of their operations and stay ahead of constantly evolving threats long into the future.

AI paves the way for new opportunities and technological advancements, revolutionizing industries and enhancing everyday life. By prioritizing robust security measures, your company can protect sensitive data and maintain customer trust, ensuring that both your business and customers reap the full benefits of AI safely and responsibly.

We've featured the best encryption software.

This article was produced as part of TechRadarPro's Expert Insights channel where we 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/news/submit-your-story-to-techradar-pro

• Trimble warns Cityworks is being abused in RCE attacks
• The company released a patch to address the issue
• CISA warns users to apply patch as soon as possible

Hackers are hijacking government software to access sensitive servers, experts have warned.

The warning comes from software vendor Trimble, whose product seems to have been used in the attack. In a letter sent to its customers and partners, Trimble said it observed cybercriminals abusing a deserialization vulnerability in its Cityworks product to engage in Remote Code Execution (RCE) and deploy Cobalt Strike beacons on Microsoft Internet Information Services (IIS) servers.

Trimble Cityworks is a Geographic Information System (GIS) asset management and permitting software designed to help local governments and utilities manage infrastructure, maintenance, and operations efficiently. It was found to have been vulnerable to CVE-2025-0994, a high-severity deserialization bug allowing for RCE, given a severity score of 8.6 (high).

Patching the flaw

“Following our investigations of reports of unauthorized attempts to gain access to specific customers’ Cityworks deployments, we have three updates to provide you,” the company said in the letter. To tackle the threat, Trimble updated Cityworks 15.x to version 15.8.9, and 23.x to 23.10. It also warned about discovering some on-prem deployments having overprivileged IIS identity permissions, and added that some deployments haid incorrect attachment directory configurations.

All of these should be addressed at the same time, to mitigate the threat and resume normal operations with Cityworks.

We don’t know how big the attack is, or if any organizations were compromised as a result, but the US Cybersecurity and Infrastructure Security Agency (CISA) has released a coordinated advisory, urging customers to apply the patches as soon as possible, BleepingComputer has found. “CISA reminds organizations to perform proper impact analysis and risk assessment prior to deploying defensive measures,” it was said in the advisory.

“Organizations observing suspected malicious activity should follow established internal procedures and report findings to CISA for tracking and correlation against other incidents.”

Via BleepingComputer

You might also like

• Bad news — many of today's top passwords can be brute force cracked in less than an hour
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

• Captain America 4's creative team has responded to rumors about its development
• The Marvel movie is said to have received poor audience reactions during initial screeners and subsequently undergone various reshoots
• Its director and one of its producers say the speculation has been blown out of proportion

Captain America: Brave New World's creative team has responded to rumors surrounding the Marvel movie's protracted development.

The Marvel Phase 5 film is set to fly into theaters this Friday (February 14), aka Valentine's Day. However, there's no *ahem* love lost between those working on the superhero flick and industry insiders, with the latter claiming that Captain America 4's development cycle has been difficult. From reports of poor audience reactions during initial test screenings, to the rumors that it underwent significant reshoots across three different time periods, you'd be forgiven for thinking Brave New World's production was anything but troubled.

Speaking to TechRadar at the film's UK junket, though, director Julius Onah and producer Nate Moore played it off as nothing more than idle gossip. Indeed, the pair insisted that one of 2025's most exciting new movies only underwent one period of extensive reshoots in a bid to make it the best film possible.

No, Marvel didn't sneakily reshoot large portions of Brave New World on multiple occasions (Image credit: Marvel Studios)

"They're not as significant [the reshoots] as the reports say they are," Moore told me. "It's actually less than what we've done on other Marvel projects.

"Our goal is to make the best movie – always," he continued. "And our task is to keep turning over the stones until it's as good as we think we can make it. The blessing of working at a place like Marvel is we have the resources to do that, so why not take that opportunity?"

Our task is to keep turning over the stones until it's as good as we think we can make it

Nate Moore, Captain America 4 producer

When I put the same question to Onah in a separate interview, the director of Captain America's latest big-screen adventure, which stars Anthony Mackie's Sam Wilson as Steve Rogers' successor, said: "Every movie of this scale has additional photography baked into the creative process. There are things you're going to refine and the story is going to evolve.

"Like every single MCU [Marvel Cinematic Universe] movie, we had one period of reshoots, so whatever the reports might have been out there, that's not the reality of how this movie was made. When you're making something of this scale with action, visual effects, and different plot threads and characters, you have to take your time. You have to make sure you're creating the best experience for the audience. But, yes, we had one singular additional photography process and then utilized that to make this film the best it could be."

I’ve consistently heard CAPTAIN AMERICA: BRAVE NEW WORLD is OK-to-pretty good. Those expecting a disaster are gonna be disappointed.February 9, 2025

That answers speculation about Cap 4 requiring more than one spell of reshoots to get it right, but what about the conjecture that Brave New World's original cut was given the thumbs down by fans who were lucky enough to see it early? For that, we go back to Moore.

"The test screenings have been over-reported as well," he replied. "We're aware of what people are talking about a lot of the time, but there's no use in us raising our hands when those reports come to say 'well, that's not correct'. We really want the film to do the talking and, now that it finally can, I think people are going to be surprised and impressed. I can't wait for everyone to see it."

I'll be covering Brave New World in much more detail leading up and after the superhero movie's launch. Ahead of more exclusive coverage heading your way, find out more about the film in my Captain America: Brave New World guide and read my Captain America 4 cast and character piece to see who's joining Mackie on the screen for this one.

You might also like

• Here's how to watch the Marvel movies in order
• Get the lowdown on what we know about Daredevil: Born Again
• Marvel debuts new Thunderbolts* trailer during Super Bowl LIX, and it reveals the MCU movie's main villain who you probably expected to be a good guy

Data has exploded over the last decade. People generate more than 402.74 million terabytes of new unique data every day in 2024. Amounting to 147 zettabytes of data this year. This figure is even more remarkable if you consider that an estimated 90% of the world's data was generated in the last two years alone.

The growth of data is overwhelming enterprises, leading to challenges in processing, transferring, and storing vast amounts of information efficiently. This often results in increased costs, data management complexities, and the need for advanced infrastructure and analytics solutions to handle the influx.

In today's digital landscape, data processing can occur either at centralized locations like data centers and on-premise servers or at the edge, where data is generated and used, such as in the field or in factories. The choice between these two approaches is becoming increasingly critical due to the sheer volume of data being produced.

Processing data centrally can lead to significant transmission delays, impacting decision-making times. This latency makes edge computing, which processes data closer to its source, an attractive solution. By reducing the distance data needs to travel, edge computing can enhance real-time processing and responsiveness. However, despite its advantages, edge alone isn’t enough — security should be at the forefront of these technological advancements.

The benefits of edge computing

Edge computing effectively shortens the distance data has to travel, removing the need to send data to a central processing location, thereby alleviating a tremendous strain on network resources. It also represents a significant reduction in transactional and transmission delays, enabling devices and users to use data more efficiently and securely and make faster decisions.

The emergence of the hybrid work model has delivered benefits, but it has also come with inefficiencies, sometimes requiring employees to travel back and forth more frequently. Edge devices can, in some cases, reduce the need for a worker to make a repair call or check the status of a piece of equipment on a factory floor. By processing IoT data closer to devices, edge computing can help enhance bandwidth and deliver near-real-time data.

Thus, one of the UK ports that acts as a vital hub for global trade, has significantly enhanced its operational efficiency and safety by deploying a Private 5G network. By using already developed solutions from Verizon Business, the port’s network now supports advanced technologies such as IoT sensors, autonomous vehicles, AI-powered logistics, and augmented remote assistance, enabling real-time asset tracking and environmental monitoring. These innovations have helped position the port as a leader in port innovation, delivering substantial operational improvements and efficiencies across its ports and harbors.

Synergy with 5G

Edge computing shows tremendous potential to introduce compelling new use cases, but as applications become increasingly immersive and graphics-intensive, they also become increasingly data-intensive. In other words, for enterprises to take full advantage of these applications and to make use of emerging technologies, like artificial intelligence and machine learning, they’ll need a robust network infrastructure.

With more advanced solutions comes a need for more processing power. With more endpoints and devices at the edge, built-in network security is increasingly important. Among other things, 5G can enhance security for edge computing which can allow edge computing to be fully realised. 5G can deliver increased bandwidth that can scale on demand; the reliability needed to support the continuous availability of mission-critical applications; end-to-end data encryption and can power advanced cybersecurity tools; the ability to increase network capacity to support more devices; and a network agile enough to support rapid changes and demands.

However, with the promise of more connected devices, especially IoT devices, comes an increased risk. Securing these devices in a 5G-enabled world is not just a technical challenge – it’s a fundamental requirement for building trust and ensuring that this innovation is sustainable. Securing IoT devices requires a multilayer approach: strong authentication methods, data encryption both at rest and in transit, and constant updates to keep devices secure against new threats. But it’s not enough to secure the devices in isolation. IoT security must be integrated into the broader network, with tight access controls and continuous monitoring to detect and respond to threats. This is the promise of 5G: not just faster speeds or more devices, but a safer, more secure foundation for the next wave of innovation.

Enterprises need to adapt

With innovative technologies knocking on enterprises’ doors daily, the latter face constant mounting pressure and challenges to deploy the technologies to keep up with the pace of global development. However, organizations must ensure their current ecosystems, internal policies and tech stack are ready for the deployment of edge computing technologies. Simply, outdated connection hardware could cause havoc in the warehouse of the manufacturing department or old cybersecurity protocols could expose retailers’ vulnerabilities making it a sweet spot for hackers.

To avoid such a nightmare on a mundane Friday afternoon, businesses should follow a step-by-step guide to preparing their operations for the deployment of 5G edge computing technologies:

1. Upgrading the company’s tech stack: For enterprises, this could mean upgrading the WAN, while for smaller businesses, it may begin with migrating to a VoIP communication solution to prepare to streamline mobility, video conferencing and online meetings.
2. Making the company’s tech stack policies agile: These behind-the-scenes efforts help make IT more agile and efficient, such as adopting a software-defined WAN to better meet changing service demands via policies that help a network adapt on the fly.
3. Elevating existing tech solutions: technological transformation is critical to digital transformation. For example, a logistics company could update its supply chain software with next-generation fleet telematics that reduces blind spots and improves on-time delivery rates with new smart IoT sensors. 5G-powered computer vision can make these processes even more valuable.
4. Anticipating the forthcoming innovations: Innovation goes beyond operation, addressing value-added solutions that drive the business forward. These innovations help businesses get noticed in a crowded market—especially those that can anticipate the needs of their customers ahead of time.
5. Becoming the tech disruptor: The fifth and final state of transformation is being ready to disrupt. These organizations have begun to adopt tools like augmented and virtual reality and other new technologies to design revolutionary experiences, changing the way the business approaches.

Securing the Edge

However, the decentralized nature of edge computing introduces unique vulnerabilities such as increased exposure to cyber-attacks, unauthorized access, and tampering due to the diverse environments in which edge devices operate. Unlike centralized systems, edge devices operate across diverse environments, increasing exposure to security risks. As data is processed closer to the source, devices may become targets of cyber attacks, unauthorized access, or tampering. Implementing a layered security strategy is critical to address these challenges. This includes robust authentication measures, endpoint security solutions, and encryption protocols to protect data both in transit and at rest.

Storing and processing data at the edge also raises important privacy concerns, especially when housing sensitive information such as health and financial data. Organizations must design their infrastructures with privacy always being front of mind, adhering to regulations such as GDPR and deploying encryption and anonymization strategies. Regular audits and strong governance ensure data stays protected, proving edge computing can balance both innovation and compliance.

In a world where security is only as strong as its weakest link, zero-trust architecture and network segmentation aren't just options but real imperatives. Every device, every user, and every request must be verified. By isolating sensitive assets and controlling access, businesses can limit the damage of potential future data breaches. This interconnected approach ensures that security measures are comprehensive and cohesive, addressing the multifaceted challenges posed by edge computing.

The explosion of data over the last decade has overwhelmed enterprises, leading to challenges in processing, transferring, and storing vast amounts of information efficiently. Centralized data processing can cause significant transmission delays, impacting decision-making times. Edge computing, which processes data closer to its source, offers a solution by enhancing real-time processing and responsiveness. However, to fully leverage these technologies, enterprises must ensure their tech infrastructure and policies are compatible and that the necessary security measures are in place.

We've featured the best business VPN.

This article was produced as part of TechRadarPro's Expert Insights channel where we 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/news/submit-your-story-to-techradar-pro

The more speed, the better, right? Not necessarily. Here's what you need to know about a multi-gigabit internet service.

Whether you just want to stream Netflix or have serious privacy needs, NordVPN is a fast, user-friendly and full-featured virtual private network.

• Zotac has fired up a scheme via its US Discord channel
• Active community members can enter a raffle to win the chance to buy an RTX 5090 or 5080
• It looks like this plan is helping in the good fight to beat scalpers

The scramble for Nvidia’s new graphics cards has been an unpleasant affair (once again), especially with the RTX 5090, as stock of the flagship GPU has been even more scarce than the RTX 5080 – but Zotac has a plan to help genuine buyers (as opposed to scalpers) in the US.

VideoCardz reports that Zotac’s Discord channel in the US is running a “priority access campaign exclusively for our loyal community” to offer active members an opportunity to purchase an RTX 5090 or 5080 graphics card.

It’s a raffle system, with a chance to buy as the prize, and there are some conditions attached, as you might expect. Notably a clause that any Discord member must have been actively engaged in discussions, challenges, and the like, to be eligible. Anyone joining to spam a few three-word-messages in the hopes of getting a purchase won’t (or shouldn’t) be considered.

Any kind of “manipulation” or “cheating” won’t be tolerated either, Zotac says. Furthermore, the board maker will keep an eye on reselling activity, with serial numbers of GPUs that are bought being recorded, and a threat of publicly exposing any scalper who does resell their graphics card.

(Image credit: Zotac) Analysis: Zotac shows the way

Quite how Zotac might police the entire playing field for resellers isn’t clear, and probably not feasible, but still – I’ve got to admire the stance taken here against scalpers and bots.

These nefarious characters make any low-stock hardware launch much more of a nightmare than it should be, so efforts such as the scheme Zotac has concocted here must be applauded.

Wondering how Zotac plan has played out so far? According to the Reddit thread that VideoCardz picked up on this development through, Zotac’s Discord channel has witnessed a flood of folks signing up since the announcement.

So, it’s certainly going to drive more traffic for the channel, clearly, but if all a new member does is spam rubbish, they should be weeded out and be ineligible for the Blackwell purchase raffles.

Those who were active members before the announcement have apparently got a private channel with separate raffles to the main ones, with much greater chances of winning (like 25%, compared to less than 1% for the main raffles, we’re told, though take that with some seasoning).

It all sounds well enough organized, in short.

Why don’t all graphics card makers go down this route? Well, as VideoCardz points out, Zotac is one of the smaller outfits of the larger manufacturers (if you get what I mean), and it’s probably much more of a headache for some of the biggest operations in the GPU sphere to run a scheme such as this.

Still, shoring up defenses against scalping more broadly when it comes to popular launches like the RTX 5000 GPUs is well worth giving some more thought to, for both retailers and Nvidia’s board-making partners alike. Sadly, what we’ve seen elsewhere is official asking prices from graphics card makers themselves creeping up, or even jumping up, as a reaction to thin supply in the face of huge demand – so, what’d be a start is not doing this, please.

We’re trying to help as best we can, through our guides on where to buy an RTX 5080 today and also our RTX 5090 graphics card finder too. Best of luck out there, GPU hunters, but of course if you’re already on the Zotac Discord, you need to get yourself in the lottery there.

You might also like...

• Nvidia is investigating reports of crashes plaguing RTX 5090 and 5080 GPUs, with possible driver issues maybe hitting RTX 4000 models too
• Nvidia's RTX 5080 has dethroned AMD's RX 7900 XTX at the same price – but good luck finding one
• The end of the Nvidia RTX 4000 series is nigh – with the RTX 4070 expected to sell out in weeks

• Tech sector layoffs rose from 98,000 (3.9%) to 152,000 (5.7%) in just one month
• Software development and white-collar jobs could be most at-risk
• In-person and skilled roles remain competitive

New Wall Street Journal reporting has revealed an alarming trend within the tech sector – much like sceptics anticipated years ago, artificial intelligence looks to be displacing human workers and leading to higher unemployment rates.

The report found IT sector unemployment rates rose from 3.9% in December to 5.7% in January, or from 98,000 to 152,000 Janco Associates analysis of the US Department of Labor data.

More broadly, 143,000 new jobs were added in January 2025 to the US economy, albeit at a slower rate than optimal.

AI is costing IT jobs

In fact, white-collar and knowledge workers are seen as the most at-risk when it comes to AI-induced job displacement. Jaco Associates CEO Victor Janulaitis commented: “Jobs are being eliminated within the IT function which are routine and mundane, such as reporting, clerical administration.”

Companies are also reducing their reliance on programmers and system designers in the hope that artificial intelligence can deliver further cost savings. The number of software development job posts dropped 8.5% year-over-year in January 2025.

Although last year's pattern was considerably lower than 2023, when layoffs.fyi tracked 264,000 tech sector redundancies, as estimated 152,000 tech workers still lost their jobs in 2024 – nearly as many as the 165,000 workers who lost their jobs in 2022.

Recent notable job losses include Sonos (12% of its headcount), Meta (5%), Microsoft, Amazon and Google.

The report also suggests that further corporate investment in artificial intelligence could serve as an early sign that future job cuts could come – a trend described as “cost avoidance.”

However, while certain jobs may be at risk, others remain in high demand. The report reveals that certain in-person and skilled roles are in greater demand than many white-collar positions – the 'in-person' element of that trend is particularly interesting, given the widespread return-to-office mandates that have been enacted post-pandemic.

It's not all bad news though – Janulaitis revealed that January’s figures, which paint a negative picture for the year ahead, could indeed be artificially inflated by many companies who are looking to implement this year’s cost-cutting measures now, rather than later.

Regardless, with a further 10,800 job cuts actioned in the first five weeks of 2025 across the industry, many workers are faced with an uncertain future.

You might also like

• Worried about the news? Check out the best job sites and best recruitment platforms
• Prepare for changes with the best online learning platforms
• White-collar jobs are increasingly being exposed to AI

• Millions of devices, likely infected with malware, are being used in a hacking campaign
• Researchers spotted brute-force attacks against VPN and other internet-connected devices
• The majority of the IP addresses are located in Brazil

A wide range of Virtual Private Network (VPN) and other networking devices are currently under attack by threat actors trying to break in to wider networks, experts have warned.

Threat monitoring platform The Shadowserver Foundation warned about the ongoing attack on X, noting someone is currently using roughly 2.8 million different IP addresses to try and guess the passwords for VPNs and similar devices built by Palo Alto Networks, Ivanti, SonicWall, and others.

Besides VPNs, the threat actors are going for gateways, security appliances, and other edge devices connected to the public internet.

Brute force

To conduct the attack, the threat actors are using MikroTik, Huawei, Cisco, Boa, and ZTE routers and other internet-connected devices, likely compromised with malware, or broken into themselves, thanks to weak passwords.

Speaking to BleepingComputer, The Shadowserver Foundation said that the attack recently increased in intensity.

From those 2.8 million, the majority (1.1 million) are located in Brazil, with the rest split between Turkey, Russia, Argentina, Morocco, and Mexico.

This is a typical brute-force attack, in which threat actors try to log into a device by submitting an enormous amount of username/password combinations, until one succeeds. Brute-force attacks are usually successful against devices protected with poor passwords (those that don’t have a strong combination of uppercase and lowercase letters, numbers, and special symbols). The whole process is automated, making it possible on a grander scale.

The automation part is made possible through malware. Usually, the devices used in the attack are part of a botnet, or a residential proxy service. Residential proxies are IP addresses assigned to real devices by internet service providers (ISPs). They make it appear as though the user is browsing from a legitimate residential location rather than a data center, which makes them a major target for cybercriminals.

You might also like

• Bad news — many of today's top passwords can be brute force cracked in less than an hour
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app