The digital identity sector has welcomed the UK government’s decision to backtrack on plans to make its national digital ID scheme mandatory for right-to-work checks – in effect, removing any compulsory aspect of the proposed scheme.
Less than a week after the Cabinet Office promoted MP Josh Simons to become minister for digital government, in charge of the digital ID policy, it has now removed the most controversial aspect of the proposals announced by prime minister Keir Starmer in September last year.
Starmer launched the national digital ID scheme by pitching it as a means to control undocumented immigration, through making use of a government app mandatory when employers conduct right-to-work checks.
The plan brought an instant backlash from civil rights groups and privacy campaigners, as well as widespread criticism from the tech sector, where digital identity providers have been encouraged for many years to go through a rigorous compliance process to achieve accreditation on a government-approved register of digital verification services. Startups and investors feared the impact of an official government digital ID app on their prospects for developing and growing the market in the UK.
Since then, the government has gradually changed its rhetoric, moving away from the contentious focus on undocumented immigration and attempting to present its plans as a way to make digital public services more modern and efficient.
An online petition calling for government to halt plans for the online identity programme received more than three million signatures, prompting a debate in Parliament where MPs from all parties laid out their concerns.
As a result, the government has now removed the compulsory element of the proposed scheme in advance of a consultation on national digital ID, which is due to commence soon. Use of some form of digital proof of identity will still be mandatory for right-to-work checks, but the government digital ID app will only be one option for doing so.
A government spokesperson said: “We are committed to mandatory digital right-to-work checks. Currently right-to-work checks include a hodge-podge of paper-based systems with no record of checks ever taking place. This is open to fraud and abuse.
“We have always been clear that details on the digital ID scheme will be set out following a full public consultation which will launch shortly. Digital ID will make everyday life easier for people, ensuring public services are more personal, joined-up and effective, while also remaining inclusive.”
Private sector leaders in the digital ID sector hope the move will allow the wider market to flourish and bring more choice for citizens in how and where they use such technology.
“Sanity has prevailed. This is a necessary reset. Let’s hope we can now shift the narrative and focus on the social, economic and practical benefits that voluntary digital ID will bring to UK citizens and residents,” said Richard Oliphant, an independent legal consultant and expert on digital identity.
Robin Tombs, CEO of Yoti, which has more than seven million users of its digital ID app, said: “The mandatory messaging provoked a strong backlash from many opponents and has increasingly risked sabotaging the value of the upcoming public consultation.
“The government team now has the opportunity to engage in a more productive, less contentious discussion focusing on improving access to public services for citizens who want to use a government, or certified private sector, digital ID.”
David Crack, chair of the Association of Digital Verification Professionals, added: “A warm welcome for this news which was always going to come. Congratulations to the government in making this known early. Now we can get on to discussing what type of digital ID the country needs and how we can all gain control of our data.”
According to figures from the independent Office for Budget Responsibility, the government scheme was expected to cost £1.8bn over the next three years – although Emran Mian, permanent secretary at the Department for Science, Innovation and Technology, told MPs in December that was “not a figure [the government] recognises”.
Private sector concerns about the government proposals led to a showdown meeting in December between industry representatives and the prime minister’s chief secretary, Darren Jones, who was given overall policy responsibility for digital ID, in an attempt to ease fears. Jones stressed that no firm decisions had been made and that government wanted to use the consultation process to determine the best approach to take.
Further concerns over the scheme came from its reliance on the existing One Login single sign-on system used for logging in to many online public services. Computer Weekly last year revealed a series of serious security and data protection concerns around the One Login system.
Your washing machine and dryer are easily two of the most useful home appliances. Two-in-one washer/dryer combos are often said to be even better, as they can dry your clothes as soon as the washing cycle is over. But while these devices may be the unsung heroes of home chores, humans still have to fold the clothes, an activity many people dislike. That’s where an AI robot for the home could be very useful, as it could replace the human in this final step of the laundry cycle. In fact, LG’s new robot, CLOiD, does more than fold laundry. It can talk to other home appliances, especially devices made by LG, and perform more chores at home — including loading the washer, cleaning the dishes, and preparing food.
Unveiled at CES 2026, LG CLOiD is part of the company’s “Zero Labor Home” vision, which involves having intelligent machines and robots handle the repetitive home chores that humans have to perform. The new robot won’t just fold clothes at home once it’s available commercially. It’ll be able to load the washing machine, as seen in a demo LG offered at CES. LG also launched a new two-in-one washer/dryer product that could finish the laundry cycle while the humans are away. CLOiD would then fold the laundry and handle other aspects of the smart home. A video that LG showed has the new robot interacting with the AC unit and the lighting system to manage the temperature and lights at home. The robot can also interact with a smart fridge to see what ingredients are stored and heat food via a smart oven.
Unfortunately, LG has not announced availability details for CLOiD. The new robot may impress buyers, but there’s no release date or price as yet.
How LG CLOiD folds laundry
The new robot isn’t exactly a humanoid robot, as it lacks human-like legs. It features a head, a torso, and two hands each with five fingers. The head is where the new robot’s brain resides. It features the computer (running AI processes), a display that shows facial expressions, a speaker for voice interaction with humans, and cameras and sensors so the robot can see its surroundings. To perform chores at home, CLOiD runs a complex AI system that combines a Vision Language Model (VLM) and a Vision Language Action (VLA). VLM is a technology that allows the AI to understand images and video. The VLA is what translates visual and verbal inputs into actions.
CLOiD’s torso can tilt to adjust its height, allowing the robot to pick up objects from knee level. Each arm has seven degrees of freedom, with LG saying this feature matches the mobility of a human. “The shoulder, elbow, and wrist allow forward, backward, rotational, and lateral motion, while each hand includes five independently actuated fingers for fine manipulation,” LG explains. This arm-hand configuration is what will let CLOiD perform the fine movements required to fold laundry.
The torso sits on a wheeled base featuring autonomous driving technology based on LG’s robot vacuums. LG explains in a press release that it selected this form for stability, safety, and cost-effectiveness. A robot on a wheeled base has a lower center of gravity and won’t tip over as easily if a child or pet pushes it. The downside of using a wheeled base instead of legs, as other concept robots shown at CES 2026 adopted, is that the robot can’t handle stairs. That may be a problem for households where the washing machine and dryer are stored in a basement or loft.
The secret ingredient behind CLOiD
The AI component is key for making a device like LG CLOiD work. The Korean company says it has trained the VLM and VLA models on tens of thousands of hours of household tasks. This process taught CLOiD to interpret user intent and execute actions. Also, CLOiD can interact with LG’s appliances that support the ThinQ AI Home platform and the ThinQ On smart home hub. This is what lets the new robot orchestrate actions as seen in one of the demo videos LG shared at CES 2026.
But in addition to the AI brain that CLOiD will use to manage a smart home and perform repetitive chores like folding laundry, LG unveiled a key hardware component that robot builders may not always detail in announcements. This is a new Axium actuator, seen in the LG CES 2026 presentation video above. The actuator sits in the new robot’s joints. It includes a motor that powers the force required to move and rotate the arms, and a reducer that handles speed and torque.
The CES demos show CLOiD performing actions at slower speeds than a human, but they appear to be precise nonetheless. Instead of using a brake to stop the movement, the actuator uses a regenerative slowing process that turns the excess motion into energy to recharge the battery. This sounds similar to the regenerative braking technology that some electric vehicles may feature to increase the car’s range by recharging the battery as the car slows down.
That said, LG did not offer battery life estimates for CLOiD. However, the new robot would likely handle chores while humans aren’t at home and pause for recharging, but that’s just speculation.
Microsoft is calling on more technology companies to “pay their own way” when covering the electricity costs associated with running artificial intelligence (AI) infrastructure and datacentres, rather than expect consumers to foot the bill.
The software giant’s vice-president and chair, Brad Smith, said Microsoft has vowed to “pay our way” to ensure its datacentres do not increase the electricity bills of everyday consumers, and that other tech firms should follow suit.
Expanding on this point, he said the US has an ageing electricity transmission infrastructure that is already straining under the weight of the demands put on it, with upgrades hampered by supply chain constraints on transformers and high-voltage equipment.
The impact of the latter is making it difficult to boost the transmission capacity of existing electricity networks, and building new ones can take up to a decade due to “permitting and siting delays”, said Smith, in a blog post.
“Some have suggested that AI will be so beneficial that the public should help pay for the added electricity the country needs for it … but we disagree with this approach,” he continued.
“Especially when tech companies are so profitable, we believe that it’s both unfair and politically unrealistic for our industry to ask the public to shoulder added electricity costs for AI. Instead, we believe the long-term success of AI infrastructure requires that tech companies pay their own way for the electricity costs they create.”
According to Smith, this approach will be needed to ensure the US and its AI infrastructure can tap into a “rapidly growing supply of electricity” and retain its leading position in the field of AI.
Stated goal
As detailed in the blog post, Smith went on to share examples of how Microsoft is already working with utility providers and public bodies to ensure the energy consumption habits of its datacentres do not financially burden local communities. “Our goal is straightforward: to ensure that the electricity cost of serving our datacentres is not passed on to our residential customers,” he said.
To guard against this, Smith said the company is asking utility providers to set higher energy rates for the company to cover the electricity costs of the datacentres it builds, owns and operates.
“In some areas, communities are already starting to benefit from this approach,” he said. “As part of our datacentre investment in Wisconsin, we are supporting a new rate structure that would charge ‘very large customers’, including datacentres, the cost of the electricity required to serve them.
“This protects residents by preventing those costs from being passed on, [but we] recognise the need to ensure that datacentre communities benefit everywhere,” said Smith. “We believe this approach can and should be a model for other states.”
The company is also committing to working with local utility providers, and paying for electricity capacity and support for grid infrastructure upgrades for its datacentre expansion plans. “We’ll [also] pursue innovation to make our datacentres more efficient … [by] using AI to reduce energy use and improve the performance of our software and hardware in the design and management of our datacentres,” he said.
“And we are collaborating closely with utilities to leverage tools like AI to improve planning, get more electricity from existing lines and equipment, improve system resilience and durability, and speed the development of new infrastructure, including nuclear energy technologies,” added Smith.
“By embedding these innovations into datacentres and by collaborating directly with local utilities, communities gain access to systems that are more efficient, more reliable and better prepared to support growth without increasing costs for households.”
National energy security and supply
The impact the growing demand for power-hungry AI datacentres is having on national energy security and supply is also a top-of-mind concern for the UK, and has been for some time, with the National Grid rolling out a series of initiatives to address the issue.
In line with Microsoft’s proposals, Computer Weekly has also covered cases of datacentre operators committing to financing electricity grid and substation upgrades in areas where they want to build facilities as part of their planning applications.
In the US, though, Microsoft’s blog emerged within days of a social media post being published by US president Donald Trump that stated his administration is working with “major American technology companies” to ensure US citizens “never” have to pay higher electricity bills because of datacentres.
In the post, he said his team has already been working closely with Microsoft on this matter to “ensure that Americans don’t ‘pick up the tab’” for its datacentre energy consumption habits. “The big technology companies who build [datacentres] must pay their own way,” added Trump.
One of the places in the UK capital where mobile connectivity is strongest is up to 30m underground, as more stations and tunnelled sections of the Tube network across London now have 4G and 5G mobile coverage as part of the ongoing development project between Transport for London (TfL) and Boldyn Networks.
TfL carries up to four million passengers a day on the London Underground network, and renewing and preparing the communications network for the future has long been regarded as essential to maintain and improve critical infrastructure. Before it carried out its first expansion work, TfL noted that legacy systems can slowly become unreliable and disrupt services, leading to delays and an overall negative impact on passenger journeys.
All of the three mobile network operators – VodafoneThree, EE and Virgin Media O2 – are taking part in the roll-out, as part of the Mayor of London’s and TfL’s stated commitment to bring mobile connectivity to the whole of London’s transport network.
The expanding coverage will host the new Emergency Services Network (ESN) which, when fully operational, is designed to give first responders immediate access to life-saving data, images and information in live situations and emergencies on the frontline.
Boldyn Networks was awarded a 20-year concession by TfL in June 2021 to deliver high-speed mobile connectivity across the entire London Underground network, creating a backbone of mobile and digital connectivity across the network to all ticket halls, platforms and tunnels on the Tube network, with total network coverage targeted for some time in 2026.
Around 400 engineers are regularly working on the project overnight, delivering the project during the limited overnight engineering hours on the Tube network. Boldyn is also committed to installing a fibre backbone across the capital to improve connectivity both above and below ground.
TfL and Boldyn are working to introduce high-speed 4G and 5G mobile coverage across the whole Tube, Docklands Light Railway (DLR), Elizabeth line network, and the Windrush line between Highbury & Islington and New Cross. Following full interoperability testing being completed across all Tube lines in 2025, design and initial testing work is now underway on the Windrush line and DLR, ahead of the tunnelled sections and stations getting coverage later in 2026.
Alongside the whole of the Elizabeth line, which was completed in December 2024, 62 out of 121 Tube stations which are located “underground” now have begun receiving mobile coverage in their ticket halls, corridors and platforms. Key stations including Euston Square, Cannon Street and Battersea Power Station have recently gone live with 4G and 5G mobile coverage, with more – including King’s Cross St Pancras, Gloucester Road, Warwick Avenue and Vauxhall – set to go live in the next few months.
TfL also revealed that work to extend coverage in the tunnels along Tube lines was continuing to “make good progress”, with the first sections of the Circle and District line between Blackfriars and Cannon Street, and between Notting Hill Gate and Bayswater, now live. TfL and Boldyn are working to introduce more sections as quickly as possible during 2026 and they expect “the vast majority” of the Northern and Metropolitan lines to have coverage in the tunnels by end of summer 2026.
Work to deliver mobile coverage across the whole Tube network will continue throughout 2026, with work focusing on sections of the Circle and District line, where a number of stations already have limited mobile coverage due to being closer to the surface, as well as along the Victoria, Jubilee, Bakerloo and Piccadilly lines. Installation work will also continue along sections outside of Central London, and where smaller tunnelled sections need to be treated individually.
Installation work is already taking place alongside the planned escalator works at Cutty Sark Docklands Light Railway station to help ensure that customers can benefit from mobile coverage at this station as quickly as possible.
Alongside installing mobile coverage across the Tube, TfL and Boldyn are also working to install small-cell mobile technology on TfL assets, such as lighting columns, to enhance mobile connectivity in high-footfall urban areas. Some of the capital’s busiest areas such as King’s Cross, Waterloo, London Bridge, Old Street, The Shard and Hyde Park Corner already have such connectivity.
Commenting on the network and the aims of the project as a whole, Isabel Coman, director of engineering and asset strategy at TfL, said: “It’s great to see further progress in our goal to introduce high-speed mobile coverage across our Tube network…We are on a clear path towards having 4G and 5G mobile coverage across the whole network by the end of 2026. Engineers are working hard overnight during the limited engineering hours to deliver this programme.”
Boldyn Networks UK & Ireland chief operating officer Nick Hudson added: “Our long-term partnership with TfL to extend reliable 4G and 5G mobile coverage across the London Underground is grounded in improving everyday journeys for millions of people. A project of this scale demands extraordinary engineering effort and close-knit collaboration with TfL, with work often carried out overnight in one of the world’s most complex transport networks.
“We’re immensely proud of what’s been achieved so far, and each section completed brings us closer to our goal of creating a more connected London for those who visit the city and those who call it home.”
Like all sports teams these days, Bath Rugby has found that providing state-of-the-art connectivity is part and parcel of the modern experience, and looking to elevate matchday mobile for its fans, the club has announced the successful deployment of an Antevia Networks 5G Shift end-to-end private network at its Recreation Ground stadium.
The current Premiership Rugby Champions, Bath first began life as a rugby club in 1865, and played its first game at the historic stadium in 1894. The iconic and idiosyncratically designed stadium holds as many as 14,500 spectators for each home game. It is set deep in the heart of the city that is famous for its Georgian architecture and Roman baths, almost 100 miles west of London.
From a connectivity perspective, Bath Rugby had previously experienced severe congestion on its legacy public Wi-Fi and macro cellular networks, which would cause point-of-sale terminals to fail, slowing service and impacting revenue. What began as a project to stabilise this retail connectivity was expanded to support a growing set of use cases throughout the matchday experience, leading to increased return on investment.
Antevia Networks says that through its technology it can change the economics of 5G private networks. The company believes their roll-out has been held back by high-cost and slow, complex installation cycles typically involving major suppliers and telecoms domain expert integrators.
Addressing these concerns, the 5G private network at the Recreation Ground is said to be able to deliver predictable, low-latency connectivity across the entire stadium and surrounding fan zones on match days. In addition, it has been built to support a range of operational and commercial use cases, including point-of-sale transactions, mission-critical staff communications, CCTV and body-worn cameras, media connectivity, and digital signage.
Designed to overcome traditional limitations and provide better performance than legacy Wi-Fi and congested public mobile networks, 5G Shift is described as being able to provide “unbroken” coverage using a fraction of the access points, “dramatically” reducing the cost and complexity traditionally associated with cellular deployments.
Powered by patented multiplexing and shared cell technologies, the network requires no RF planning, can be deployed rapidly and is managed like enterprise Wi-Fi by the club’s existing IT team. 5G Shift uses a cloud-based virtualised RAN architecture built on O-RAN standards and COTS hardware. The patented multiplexing technology enables all radios to appear as a single shared 5G cell to eliminate handover issues, and reduces the number of access points required. In deployments to date, Antevia says it has required less than 10% of the access points required for Wi-Fi.
As a result of deployment, Bath Rugby uses the network to connect mission-critical and high-value applications including point-of-sale terminals for food, beverage and club-shop transactions; push-to-talk communications for medical teams and match-day operations; CCTV and body-worn cameras for crowd safety and control; and digital signage across fan areas delivering real-time interviews, replays and information. It can also see use in providing temporary connectivity for VIP hospitality areas and high-speed media access for journalists filing live reports.
“Bath Rugby’s Recreation Ground is a perfect example of how simple, scalable and cost-effective private 5G can deliver real-world value,” said Antevia Networks CEO Simon Cosgrove. “5G Shift gives organisations comprehensive coverage with ultra-low latency at a low price point, without the complexities and heavy engineering normally associated with the deployment of private cellular.
“We’re proud to be supporting Bath Rugby with a network that stands up, even when the stadium is at full capacity, and we’re excited to keep expanding the use cases that can help deliver operational and commercial benefits for the club.”
For the deployment, Bath Rugby and Antevia worked with local enterprise resilient wireless services firmSpry Fox Networks.
In its first ever major legal action outside the United States, Microsoft’s Digital Crimes Unit (DCU) has disrupted cyber crime-as-a-service network RedVDS – whose subscribers have cheated their victims out of millions of pounds – after obtaining separate court orders in the UK and Florida.
The DCU turned to the British legal system because the malicious infrastructure used to run RedVDS was hosted by a UK-based provider. A great number of victims of RedVDS users, well over 7,500, are also located in the UK, it said.
“Cyber crime today is powered by shared infrastructure, which means disrupting individual attackers is not enough. Through this coordinated action, Microsoft has disrupted RedVDS’s operations, including seizing two domains that host the RedVDS marketplace and customer portal, while also laying the groundwork to identify the individuals behind them,” said Microsoft DCU assistant general counsel, Stephen Masada.
The takedown operation drew Europol’s European Cybercrime Centre (EC3), with further support provided by the German authorities through the Central Office for Combating Internet Crime (ZIT) at the Public Prosecutor’s Office in the city of Frankfurt-am-Main, and the Criminal Police Office for the state of Brandenburg.
At the time of writing, the RedVDS website states that its domain has been seized by Microsoft.
Industrialised fraud
The RedDVS cyber criminal service charged as little as $24 (£18) per month to provide digital fraudsters with access to disposable virtual computers used to scale fraud operations cheaply and securely.
The DCU believes RedVDS users have compromised more than 191,000 organisations worldwide since September 2025 and netted over $40m in the US alone, with prominent victims including Alabama-based H2-Pharma, a supplier of allergy, cancer and mental health medications, which lost $7.3m; and Florida-based Gatehouse Dock Condominium Association, which was tricked out of $500,000 it had set aside for repairs to its members’ homes.
The service was used for a wide range of cyber criminal activity, including running phishing campaigns, hosting malicious infrastructure and facilitating fraud. It was often used alongside generative AI (GenAI) tools to help identify more targets quicker, generate more convincing lures, and in some cases to manipulate video footage or clone voices.
However, where RedVDS appeared to excel was in supporting business email compromise (BEC) where cyber criminals impersonate trusted individuals to send payments to accounts they control.
In particular, its users targeted the real estate sector, compromising the accounts of estate agents, escrow agents or title companies. The DCU believes that as many as 9,000 customers in the real estate industry, most in Australia and Canada, were affected by this activity to some degree.
Masada said the DCU’s latest action built on ongoing efforts to disrupt fraud and scam infrastructure via both legal and technical actions, and through global collaboration.
“It marks the 35th civil action targeting cyber crime infrastructure by Microsoft’s Digital Crimes Unit, underscoring a sustained strategy to go beyond individual takedowns and dismantle the services that criminals rely on to operate and scale,” he said.
“As services like RedVDS continue to emerge, Microsoft will keep working with partners across sectors and borders to identify and disrupt the infrastructure behind cyber-enabled fraud, making it harder for criminals to profit and easier for people and organisations to stay safe online.”
The annual CES trade fair in Las Vegas serves us a glimpse of what’s to come in consumer technology in 2026 and beyond. BGR saw plenty of cool and exciting gadgets at the show. Many are available to buy right now, whereas others are still in the concept phase, being developed into actual products that may grace store shelves soon.
We’ve already covered the most exciting tech at CES 2026 and the best gadgets unveiled at the show, but that’s just scratching the surface of this massive international exhibition. There were countless other devices and technologies being demoed around the show floor and behind closed doors at exclusive events, so here are four more of the coolest tech innovations that BGR spotted at CES that undoubtedly deserve your attention. These innovations range from exciting new gadgets that you’ll be able to buy soon to interesting new technologies that will level-up your esports gaming.
TCL X11L SQD-Mini LED TV
Although the RGB Mini-LED or RGB LED technology was expected to be the big thing at this year’s CES, TCL truly wowed with its X11L SQD-Mini LED TV. It’s a massive improvement over the current crop of Mini-LED TVs. Instead of using red, green, and blue LEDs like standard RGB Mini-LED TVs, it keeps the same blue LEDs, which are found in many traditional Mini-LED TVs, but employs better quantum dots for more accurate colors and a new color filter. These improvements help the TV provide full coverage of the BT.2020 color space, meaning it can provide exceptional color fidelity. Because it uses single-color blue LEDs, it also avoids color crosstalk — when colors from one color LED bleed into another LED’s colors — a problem that can plague RGB Mini-LED TVs.
Another advantage of the newly showcased TCL TV is its 10,000 nits peak brightness, which is among the highest you can get on any TV. This high peak brightness enables the TV to provide better HDR performance and counter glare and reflections more effectively. It’s also the first TV to claim support for Dolby Vision 2, a new HDR format, which will arrive via a software update. The 85-inch and 98-inch models of the TV are already on preorder, with a 75-inch coming down the line. However, it’s not a cheap TV by any stretch, with the 98-inch costing $10,000.
Lenovo Legion Pro Rollable
Lenovo’s Legion Pro Rollable was one of the most exciting laptops to be shown at this year’s CES. Although it’s still very much a concept, it offers a glimpse into what’s possible in such a form factor. As the name suggests, it’s a laptop that uses motors to offer a horizontally expandable screen. Lenovo has shown off laptops with similar rollable displays in the past; however, this is the first one that’s not featuring a vertically expanding display. The company is targeting esports athletes who often need a wide screen to practice, which is hard to get in a portable form factor when traveling.
The laptop uses an OLED screen and has a 16-inch display by default, which can expand to 21.5-inch or 24-inch wide. It has a dual-motor design to expand and contract the screen. Right now, it has the same internals as Legion Pro 7i, which means you get a top-spec Intel Core Ultra CPU and an Nvidia GeForce RTX 5090 GPU. That said, the consumer version, if there is ever one, will likely feature different specifications.
Jackery Solar Mars Bot
Jackery showed off a pretty cool power station, solar panel, and robot combo called the Solar Mars Bot that can move around to get the most efficient solar charging. Its built-in wheels allow it to adjust its position outside your home or campsite to track the sun and pursue the best sunlight irradiation angle. The company mentioned this design during CES 2024 as a concept, but it has since come much closer to becoming a reality. However, there is still no word on its release date or pricing.
There are retractable solar panels on the top of the device, which can fold and unfold as necessary, and are capable of producing up to 600W of power. More importantly, the platform on which solar panels are fixed can change its angle up to 60 degrees to get the most sun and increase its efficiency. The rest of the Solar Mars Bot is similar to other Jackery power stations and includes multiple ports and sockets to power your devices, including multiple USB ports and AC outlets. The company has also included 5G connectivity in the Solar Mars Bot, allowing you to check the battery and robot status even when it’s out of Bluetooth range.
Nvidia DLSS 4.5
PJ McDonnell/Shutterstock
DLSS on Nvidia graphics cards (which means Deep Learning Super Sampling) has quickly emerged as an excellent tool to get more out of your GPU. It uses machine learning to render graphics at a lower resolution and then upscales them with the help of AI to fill in any missing details. This allows mid-range graphics cards to push out significantly higher frame rates than they would be able to achieve with native rendering. DLSS 4, which was introduced at CES 2025, was already pretty powerful. However, Nvidia is taking things to the next level with DLSS 4.5. The company showed off the upgrade at CES 2026, and it’s already available for all RTX graphics cards. One of its highlights is the updated upscaling technology, which enables significantly improved visual clarity with fine detail.
More importantly, the visual quality of the “Performance mode” in DLSS 4.5 is as good or better than the “Quality mode” in DLSS 4, which helps you get a higher number of frames from your hardware without any artifacts. For RTX 50-series GPUs, there is also the new Dynamic Multi-Frame Generation feature, which can generate up to 5 synthetic frames for each real frame and can dynamically adjust the artificial frames depending on the base frame rate to ensure smoothness at all times. Bottom line, DLSS 4.5 delivers dramatically improved visual performance with enhanced sharpness, HDR lighting, and lighting effects.
The Consumer Electronics Show is a yearly expo in Vegas for manufacturers of all kinds to wow both the media and public with weird and wonderful bleeding-edge technology; from rollable OLED screens that unfurl to several times their original size, to household robots that can barely fold laundry at the level of a two-year old. But away from the usual assortment of increasingly thin TVs and “smart” fridges that no one asked for lies an entirely stranger set of products. These products either wow you for their creativity, or utter stupidity. It’s a fine line, and I’ll let you be the ultimate judge of which side these products lie on.
In truth, much of the concepts in this list will never see the light of day. Some are an obvious ploy to garner exposure from media outlets (guilty!), some are prototypes of products that might eventually come to market with a vastly different feature set, while some are available for pre-order now. You don’t even have to wait to buy some of the best CES gadgets. But these are the strangest, in their own special way.
An Actual Toilet Cam: Throne One
The quanitifed-self movement of the early 2000s introduced the concept of measuring everything in your daily life, but early attempts were limited to self analysis of primitive sensor data. Nowadays you can track everything, from the number of times you’ve sipped a water bottle to diaper changes, and it’s not unusual to pull out your Apple Health records when you visit the GP. So it was only a matter of time before gut health got in on the game.
Equipped with a camera and microphone, the Throne One from Throne Science sits on the side of your bowl to analyze everything about your stool and urinary habits, and lets you know when something is amiss. While this might sound like a joke, if you can make it past the ick factor, gut health is perhaps one of the most overlooked early indicators for a wide range of health issues. While temporary color or consistency changes are usually down to a change in diet, consistent issues point toward something more serious. Available for pre-order and shipping in February, the Throne One will cost $400 as well as requiring a monthly subscription for the analysis software.
AI Companion Doll: Lovense Emily
Developed by interactive sex tech makers Lovense, Emily is a lifesize silicone companion doll that integrates a sultry voice-capable AI, a series of expressive facial motors, and embedded sensors throughout her body. She’s web-connected and features a choice of five different personalities and roleplay scenarios, while a virtual cloud copy of your companion means you’ll be able to talk to her outside of the home through the accompanying smartphone app — and she remembers previous conversations to shape future ones (ChatGPT offers a similar “memory” feature). If reading that didn’t already make this feel like an episode of Black Mirror, hold on: If she ever breaks, you can download her personality into a new body. Uniquely, Emily can also connect to and control all your other Lovense web-connected toys.
Potential buyers can put down their $200 deposit now for a sizeable discount off the final price, which Lovense has said will be somewhere in the range of $4,000 to $8,000. While they haven’t disclosed which LLM is powering the companion, it’s likely a custom solution. ChatGPT has announced adult mode for verified users was supposed to be coming at the end of last year, while X’s standalone Grok AI app is notoriously spicy.
Personalized Art, On Demand: SwitchBot AI Art Frame
Generative AI “art” is a lot like Marmite: you either love it with a slice of toast or think it devalues everything about the human experience and will be the downfall of society. And digital art frames are nothing new; I reviewed the Meural Canvas around a decade ago, a glorified big photo frame that came with premium subscription options to access exclusive streams of licensed artwork. But with the advent of generative AI, it was never going to be long before someone combined the two concepts into one product.
The SwitchBot AI art frame — powered by NanoBanana — lets you generate any subject matter in any style you like: instant, on-demand, personalized fake art displayed on a gorgeous and technically impressive EInk Spectra 6 screen. It’s weird, but also strangely compelling. The smaller 7.3″ size is available now for $150, with pre-orders open for the larger 13.3 and 31.5 inch versions. SwitchBot is better known for their button-pressing smart home gadget, which is surprisingly useful for turning dumb household appliances into smart ones.
Desktop Holographic Anime Companion: Razer’s Project Ava
Following an apparent trend of global loneliness (who would’ve thought staring at a screen for 18 hours a day would have that effect?), there are a lot of virtual companions making an appearance at this year’s CES, from full size household robots to tiny little anime girls that you can keep forever inside a cylindrical cage on your desk. Project Ava, from gaming giants Razer, is a Grok-powered gaming copilot; but it’s not the only tiny desktop companion at CES, being joined by the LePro Ami (“Your always on 3D soulmate”).
The tech isn’t actually new — these seem to pop up every year, with the earliest I can find dating back to nine years ago from Gatebox in Japan — but AI has finally reached a level where these are more than glorified 3D desktop toys. With a connection to advanced AI voice models, you’ll finally be able to hold a normal conversation. With Razer’s gaming might behind it, they’ve also promised that it’ll react to your gaming achievements or you can ask it for help when you’re stuck.
How much of this will make it to the final product or if it’ll be manufactured beyond the initial influencer round of reviews is still to be seen. What we do is that like anything Razer does, it’ll have oodles of RGB.
It’s Like Music in Your Mouth — Literally : Lava’s Lollipop Star
You might have heard of bone conduction headphones, which produce audio by directly vibrating the bones around your ear. It’s useful tech for those who need to stay connected to the outside world (or just can’t stand earbuds), though the quality isn’t brilliant. Lollipop Star uses the same tech, but in your teeth. When you suck on it (or technically, bite down on it), the music plays through your jaw. By all reports, it’s both muffled and quiet, which are two things that seem somewhat important for playing music, but I’m no expert.
In a world where we’re moving toward banning wasteful single-use plastic straws and cutlery, I can really see single-use lollipop sticks with a built-in lithium-ion battery taking off. At $9 each, you’d have to be quite the sucker for this particular innovation when the exclusive “drops” arrive shortly. Alternatively, you can buy some of these AMOS Candy musical lollipops on Amazon.
Smart AI Hair Clipper: Glyde
AI hair clippers probably weren’t on your bingo card this year, but users can scan their head shape into the Glyde app and select a desired style, strap on a special mask to protect their eyes, and then have at it — zero skill needed. Just move the clippers around over your head, and the Glyde AI automatically adjusts cutting length to match the programmed style thanks to sensors that detect the speed, tilt, and angle of the blade in real time. While a simple buzzcut is hard to mess up, anything involving a fade is usually tricky. The Glyde makes it effortless with the “fade band” marking the start. You still need to physically move the clippers over your hair, but if Glyde detects sudden movements, incorrect starting position, or not cutting upwards in a straight line, it simply retracts the blade to avoid mistakes.
As weird as thing sounds, if they hit the target price of $150-200 then it’s not far off the cost of a decent pair of non-smart clippers.
Zero-G Mental Health Pod: Reconcept
Originating from France, the Reconcept mental health pod is an alien-esque egg-shaped thing you clamber into and have a massage while it plays soothing music and binaural beats in a guided relaxation session. It’s not really zero gravity of course, but the Neutral Body Posture with the head and legs reclined was developed by NASA to improve circulation and reduce respiratory and muscular strain in astronauts experiencing microgravity.
The Reconcept pod sounds like a lot of fun — unless you’re claustrophobic, in which case it probably won’t have a positive effect on your mental health. Thankfully there are plenty of adjustable beds to help you achieve a zero-g position in the luxury of your own home. Just add a Meta Quest VR headset with haptic vest and you’re all set for plenty of meditation experiences like Tripp.
Pure Nightmare Fuel: Dreame Cyber X Robot Vacuum Transporter
For all their convenience robot vacuums have thus far been unable to tackle the incredible complexity of stairs. Until now. Dreame’s Cyber X is a terrifying beast equipped with four legs that look like tiny chainsaws — as if the dog wasn’t scared of the vacuum enough. What’s really interesting is that the Cyber X isn’t actually a robot vacuum — it’s just a transporter or base station. The vacuum has to drive inside, be carried to another floor, then get out to do its business. So while the Cyber X can move your vacuum around, it still can’t actually clean the stairs.
The Roborock Saros also made an appearance at CES with a little pair of wheels on legs to help it climb, but it’s a vacuum as well. Unfortunately the Saros can’t get back down again (or at least, it wasn’t demonstrated doing that so we’ll have to assume it can’t). But that’s half the problem solved, anyway, and it means I can finally justify that indoor slide to my wife because “the robovac needs it.”
Breakfast In Bed (As Long As You Live on One Level And Breakfast is a Croissant): LG CLOiD
The LG CLOiD is a friendly humanoid robot designed to coordinate household tasks with other connected home appliances in LG’s ThinQ ecosystem. So you’ll not only have to spend an extraordinary amount on the robot, you’ll also need to smartify everything else in your home. Unlike the quad-chainsaw-wielding Dreame Cyber X, CLOiD has primitive wheels, so it won’t be able to navigate complex home environments with any level changes. But he will fold laundry (slowly) and make breakfast, if by make breakfast you mean placing a croissant onto a plate. It’s unlikely that CLOiD will ever make it to market, but LG is showing it off as a prototype of future projects, hoping it will eventually become an “ambient care agent that supports everyday life”. I joke, but robots that can assist with care of the elderly are very much in demand in an aging society where the cost of care is skyrocketing.
Also making an appearance was the SwitchBot Onero H1, whose makers claimed it should hit retail later this year at less than $10,000. For a robot that can pick up dirty socks and place them in a basket, I think we can all agree that’s a bargain. But unlike the LG CLOiD, the SwitchBot Onero is unable to display heart emojis on its face.
We’ve all been there — you’re watching the game and need to go grab a beer, but that means being taken away from the excitement of watching people kick a ball for 30 seconds. But what if — hear me out — your TV could follow you to the kitchen? You’d never miss a thing!
Now that I’ve woven that compelling tale of a use case for you, let me explain the Hisense S6 FollowMe TV isn’t actually any of that, because despite the name, it’s not motorized and won’t FollowYou. Instead, you need to manually push it around. So it’s just a TV on a stand then? Well, yes, but that makes it sound like it’s not innovative at all. It also has a webcam, and battery, and Wi-Fi 6 — but doesn’t ship with a chain to attach it to your neck, disappointingly. The real strangeness here is that Hisense felt a smart TV on wheels was an innovation worth showing off at the largest consumer electronics show. Don’t fret about upgrading: here’s a TV stand with wheels for less than $50.
CrowdStrike has been granted a motion to dismiss a consumer class action lawsuit brought by shareholders who were affected by the now-infamous 19 July 2024 outage – prompted by a faulty sensor update – which crashed Windows PCs around the world, causing widespread disruption and billions of pounds worth of losses.
The suit, filed on behalf of CrowdStrike investors in August 2024, accused the defendants – who included the company’s founder and CEO, George Kurtz – of making false and misleading statements over the efficacy of the Falcon platform at the centre of the outage.
It also alleged failings over software testing and quality assurance, and claimed that CrowdStrike was seeking to maximise its profit by rushing untested updates.
Handing down his decision at the federal district court for the Western District of Texas in the city of Austin, US district judge Robert Pitman said the plaintiff’s claims were dismissed in their entirety in part because the shareholders had failed to establish any plausible motive of intent to commit securities fraud on CrowdStrike’s part.
It his judgment, Pitman said the court agreed with CrowdStrike that the statements it made were neither false or misleading when considered in the context from which the plaintiffs removed them. He wrote that the court concluded that if anybody was being misleading, it was the plaintiffs.
Rejecting other arguments, Pitman also said that corporate mismanagement did not, standing alone, give rise to a multibillion-dollar claim. “We appreciate the Court’s thoughtful consideration and decision to dismiss this case,” said CrowdStrike chief legal officer Cathleen Anderson in a brief statement.
Reuters earlier reported that the office of New York State comptroller Thomas DiNapoli, who led the lawsuit, is reviewing options following the decision.
Multiple actions
Last June, Pitman dismissed another lawsuit brought by airline passengers who experienced delays and cancellations after the sudden blackout.
Pitman’s decision in this case was made on the basis that the US Airline Deregulation Act preempted the claims the plaintiffs were making against CrowdStrike. His ruling has effectively shielded CrowdStrike from consumer suits related to disruptions experienced by its customers.
However, a third lawsuit brought by US airline Delta is still working its way through the US legal system.
Delta was particularly badly affected by failures arising from the outage, and was highly critical of CrowdStrike in the wake of the incident after it was forced to cancel thousands of flights and spend millions compensating stranded passengers and putting them up in airport hotels.
CrowdStrike and Microsoft have both claimed that Delta in fact rejected their offers of help during and after the disruption.
Single point of failure
The 2024 CrowdStrike outage stands as a stark reminder of the potential for a global catastrophe arising from unscheduled technical outages given the interconnected nature of cloud platforms, which enable single points of failure to cascade rapidly.
The scale of the problem was aptly demonstrated in November 2025, when web traffic management firm Cloudflare was hit by its worst outage in six years.
In an incident that bears some similarities to the CrowdStrike outage, Cloudflare had made a minor and seemingly innocuous change to a feature configuration file used by its Bot Management security system, which caused the file to grow larger than expected and propagate across the Cloudflare network, causing crashes and disrupting daily life on the world wide web for millions.
A month earlier, Amazon Web Services customers experienced a 15-hour outage after the cloud giant suffered a series of cascading technical issues at a northern Virginia datacentre powering its US-East-1 region, and in June, Google Cloud went on the blink for many after an incorrect change to an application programming interface management system triggered a crash loop.
Speaking to Computer Weekly’s sister title, Search Cloud Computing, Forrester principal analyst Lee Sustar said that such incidents are a “preview of what’s to come” and predicted more similar stories in 2026.
Sustar said a combination of hyperscalers pivoting from legacy environments to graphics processing unit-centric datacentres to manage AI workloads, and ageing infrastructure, would likely lead to “major multi-day outages” in the near future.
Apple has been tweaking the iPhone design almost every generation. With the iPhone 14 Pro, the company ditched the notch in favor of a Dynamic Island. On the iPhone 15 Pro, Apple added a titanium frame to its phones. On the iPhone 17 Pro models, a new camera plateau and the return to aluminum marked these smartphones.
Now, the well-known Chinese leaker Digital Chat Station claims the iPhone 18 Pro and iPhone 18 Pro Max models will have a redesign, as Apple could be working on a new under-display cutout area. This leak is corroborated by other reports that say Apple will make a more discreet Dynamic Island while placing Face ID sensors under the screen. Other rumors suggest the iPhone will receive another major change in 2027, as Apple plans to celebrate the 20th anniversary of the first iPhone release.
iPhone 18 — 6.27″ LTPO 120Hz display with Dynamic Island
iPhone Air 2 — 6.55″ LTPO 120Hz display with Dynamic Island
iPhone 18 Pro — 6.27″ LTPO 120Hz display with a newly redesigned under-display cutout area
iPhone 18 Pro Max — 6.86″ LTPO 120Hz display with a newly redesigned… pic.twitter.com/0Ix3CuLyuP
— Ice Universe (@UniverseIce) January 14, 2026
A design change won’t come to other iPhone models
José Adorno/BGR
Following Apple’s trend of updating the design of the iPhone Pro first, the Digital Chat Station leaker says that both iPhone 18 and iPhone Air 2 will remain with the same look as their predecessors, which means the company is not expected to add under-display components, like Face ID, or move the Dynamic Island somewhere else on the screen on these models.
While it’s still unclear whether Apple will move all the Face ID and camera components under the display on the iPhone 20, these rumors claim the company will do that with the TrueDepth system, which powers facial recognition, on the iPhone 18 Pro models.
Interestingly enough, Apple has been full of new ideas with its iPhones. For example, the iPhone Air was just introduced, but the company is expected to release the iPhone Fold this year. Then, in 2027, Apple might introduce an iPhone with an all-new design. Considering that, back in 2017, Apple decided to make a premium iPhone X model while offering the simpler iPhone 8, the rumored smartphone releases could make Apple’s release schedule as busy as it was back then. With two Pro models, one regular version with a new release schedule, the iPhone Air, and a future iPhone Fold, it becomes unclear where Apple will fit the iPhone 20 release and what the other models will look like.
