Feed aggregator

This week’s Nvidia RTX 5000 reveal at CES 2025 in Las Vegas was a pretty mixed bag for me. On the one hand, the generational price drop on the new RTX 5070 was an extremely welcome sight, as was the backward compatibility for DLSS 4 across all RTX cards. On the other hand, the RTX 5090 costs 2,000 bucks, and Nvidia’s new Multi Frame Generation tech – while impressive – will be exclusive to next-gen GPUs.

Still, there was one footnote to Nvidia’s slate of CES reveals that really piqued my interest: a renewed commitment to its SFF-Ready scheme. This program has actually been floating around for a while - it was a small part of the Nvidia press presentation I attended at Computex last year - but I hadn’t really paid it much heed up till this point.

Why? Well, as much as I love compact PCs, Nvidia’s SFF (Small Form Factor) program felt a bit half-assed at the time. The scheme determined key specifications for other industry bodies – primarily third-party GPU makers and case manufacturers – to follow, creating a sort of certification that assures users that their chosen graphics card will fit inside whatever compact case they buy.

Not all RTX GPUs are created equal – sometimes, smaller is better. (Image credit: Nvidia)

It was a good idea in theory, but there wasn’t much an experienced PC-builder couldn’t reasonably extrapolate – most of the current-gen cards supported were twin-fan RTX 4070 and 4070 Ti models, with a small handful of third-party 4080 cards. The only supported first-party FE (Founders Edition) cards from Nvidia were the 4070 and 4070 Super – if you ever saw an RTX 4090 FE in person, you’d understand perfectly that there was zero chance that ultra-chunky GPU would fit inside an SFF case.

But with the RTX 5000 series, that’s all about to change.

Small is the new big

Yes, the RTX 5090 FE is much smaller than its predecessor – despite being an absolute monster of a GPU in performance (and price) terms, the Founders Edition of Team Green’s new flagship graphics card returns to a two-slot configuration and measures 137 x 304 x 40mm - just about small enough to fit within Nvidia’s SFF-Ready size limits.

The list of viable third-party cards is also longer and more inclusive than it was for the RTX 4000 generation (you can view the list on Nvidia’s website), with almost every third-party manufacturer offering at least one RTX 5080 variant that can fit inside a compact PC case. It appears that the only 5090 that qualifies is the FE version, but that’s still a significant improvement considering that even the RTX 4080 FE was too large.

Seriously, this thing is flippin' gigantic. (Image credit: Future)

As console performance improves with the likes of the PS5 Pro and the inexorable march of PC gaming handhelds continues, there’s a rising demand for ‘living room PCs’ – the sort of compact system that can fit comfortably in your TV stand and be used for gaming with a traditional gamepad. The renewed inclusivity of the SFF-Ready scheme is a step in the right direction here.

While I might still balk a little at the price tag (sorry Nvidia, $2,000 is practically into professional-grade hardware territory), there’s no denying that the RTX 5090 would make for an absolute beast of a living room gaming machine. Hell, even the RTX 5070 should outperform any currently available console with the help of DLSS 4. OK, article over, I need to get on the phone with Nvidia right now.

You might also like...

• Nvidia CES 2025 Keynote live blog: all the latest on the RTX 5000 reveal and more
• AMD announces new Radeon RX 9070 XT and RX 9070 graphics cards at CES 2025
• Razer Blade 16 announced at CES 2025 with an Nvidia RTX 5090 GPU – but Intel has been dumped

• The HDMI 2.2 specification has been announced at CES 2025
• HDMI 2.2 can support up to 96Gbps bandwidth and 16K resolution
• It follows on from HDMI 2.1, which was first announced in 2017

The HDMI 2.2 specification has been officially announced at CES 2025 by HDMI Forum Inc, and it's said to deliver up to 96Gbps bandwidth and 16K resolution, as reported by FlatpanelsHD.

After initial reports in December suggesting HDMI 2.2's reveal at CES 2025, HDMI Forum Inc officially announced the new HDMI specification at the tech event. It's the successor to HDMI 2.1, which was first announced in January 2017 but first came to prominence in 2020 and 2021, when it began to appear more commonly on gaming consoles such as the PS5 and Xbox Series X, and on TVs such as the LG OLED CX.

HDMI – which stands for High Definition Media Interface – is a commonly used connection port that you'll find on displays such as the best TVs and the best monitors. It enables connection between said displays and other audio-visual devices, such as the best 4K Blu-ray players and best soundbars, and gaming consoles.

HDMI 2.2 is said to support a multitude of resolutions including 4K, 8K, 10K, 12K and even 16K, the latter two of which are new for HDMI 2.2. It will also have improved bandwidth, reaching 96gbps compared with HDMI 2.1's 48gbps and DisplayPort's 80gbps. HDMI 2.2 will also support 4K at 480Hz.

Alongside the HDMI 2.2 specification reveal, HDMI also confirmed that the Ultra96 HDMI cable, which supports 96gbps and HDMI 2.2 features, will be available later in 2025.

Also introduced is HDMI Latency Indication Protocol (LIP), which is said to improve synchronization between audio and video devices such as AV receivers or soundbars and TVs.

HDMI 2.2 is set to be available to supporting manufacturers, such as TV and display brands, in the first half of 2025.

What does HDMI 2.2 mean for TVs and gaming devices?

TVs such as the LG C4 have four HDMI 2.1 ports, but it'll be a while before devices implement HDMI 2.2 (Image credit: Future)

So what does HDMI 2.2 mean for TVs and gaming devices, and when are we likely to see it? The answers: currently, not much; and not very soon. Although HDMI 2.2 will be available in 2025, it'll take manufacturers some time to actually implement the tech into their devices.

It took HDMI 2.1 between three and five years to go from announcement to implementation, and it's only now commonly found in more budget displays after initially appearing in more premium devices such as the best OLED TVs. It wouldn't be surprising if HDMI 2.2 runs to a similar timeframe.

Although HDMI 2.2 will support higher resolutions, such as 8K (at 240Hz compared with HDMI 2.1's 8K, 60Hz) and apparently even 16K, the problem is the lack of devices available that will support it. 8K TVs are seeing their popularity wane, with many manufacturers abandoning the premium display tech. Only Samsung really flies the 8K flag each year. This is due to premium price tags, and the fact that there are very few 8K movies or TV shows to take advantage of the 8K display.

8K gaming is more readily available than movies or TV shows, with the PS5 Pro and some of the best graphics cards, such as the AMD Radeon RX 7900 GRE, and even the newly announced Nvidia GeForce RTX 5090, supporting 8K resolution, but the number of 8K games (especially at higher frame rates) is scare. Plus, you'd need some powerful and pricey gear to run them.

But do expect to see HDMI 2.2 in future spec lists. While manufacturers won't get the full potential out of HDMI 2.2 for a while, it serves as an excellent marketing tool. HDMI 2.1 is common on the best gaming TVs now, both premium and budget – so listing HDMI 2.2 as a new feature sounds more enticing.

HDMI 2.2 will be the future of TVs and gaming, even if is still likely a few years away, especially before the full benefits can be widely enjoyed. It's not looking like we're getting 10K out of HDMI 2.1 any time soon, so expect 12K and 16K from HDMI 2.2 to be a ways off.

You might also like

• The LG M5 wireless OLED TV might be weirdly the perfect pick for PC gamers, especially with the Nvidia RTX 5080 on the horizon
• Samsung announces 2025 OLED TVs, with next-gen Glare Free tech and 165Hz gaming support – the Samsung S95F, successor the S95D has arrived
• How high do you want your frame rates? Nvidia boasts that RTX 5090 GPU can drive Valorant at over 800fps with jaw-dropping low latency

We’re covering all of the latest CES news from the show as it happens. Stick with us for the big stories on everything from 8K TVs and foldable displays to new phones, laptops, smart home gadgets, and the latest in AI.

And don’t forget to follow us on TikTok for the latest from the CES show floor!

The veteran TV doctor is back to lead a different sort of ER drama.

• DAMAC founder Hussain Sajwani to invest $20 billion in US data center infrastructure
• Trump and Sajwani have indicated this sum could more than double
• Investment will “keep America on the cutting edge of technology and artificial intelligence"

US president-elect Donald Trump has confirmed plans to invest $20 billion into US data center infrastructure.

Rather than coming from the government, the cash will be provided by Emirati billionaire and founder of property development company DAMAC Hussain Sajwani, Trump revealed in a press conference at Mar-a-Lago.

The first phase of the investment is set to target Arizona, Illinois, Indiana, Louisiana, Michigan, Ohio, Oklahoma and Texas, however Trump and Sajwani didn’t rule out future investments.

US data center funding

Addressing the press on stage, Trump said the $20 billion investment would become available “over a very short period of time.” He also said DAMAC has indicated a willingness to “double or even somewhat more than double the amount of money” as a mark of the company’s confidence in the US.

No detailed timeline or further specifics have been shared yet, but we know that the data centers will primarily be designed to support AI and cloud technologies.

Trump added the investment would “keep America on the cutting edge of technology and artificial intelligence,” a notable statement given the country’s ongoing battle with China when it comes to artificial intelligence and the powerful chips that the technology requires.

Speaking on stage, Sajwani confirmed Trump’s comments that DAMAC would be prepared to invest more than $20 billion if market opportunities are satisfactory.

However, the news doesn’t come without its downsides – data centers are widely criticized for their high energy consumption and usage of other natural resources like water for cooling. In its last environmental report, Google noted a 48% rise in data center emissions, blaming the rapid expansion of AI for the negative impacts.

In his speech, Trump declared companies investing over $1 billion into the US would benefit from expedited environmental reviews, helping them to establish more quickly.

More broadly, hyperscalers have also committed to spending big in the industry. In the past month, AWS has committed to spending $21 billion to support its infrastructure in two US states, with the promise of generating more jobs.

You might also like

• Check out our roundup of the best AI writers
• Access the best chips with the best cloud computing services
• AWS reveals $10bn cloud and AI investment in Ohio, bringing hundreds of jobs

Meta’s AI characters users might seem useless, but fake social media users can sometimes offer valuable insights into real human behavior.

• Riot Games has released a new in-game cinematic for its next League of Legends season
• The five-minute video also appears to act as a prologue to Arcane's Noxus-set sequel show
• Mel Medarda is seen alongside other iconic Noxus characters, including Darius and LeBlanc

A new in-game League of Legends (LoL) cinematic appears to have teased our first look at one of Arcane's sequel projects.

The five-minute-long video, titled 'Welcome to Noxus', which was released yesterday (January 7) is, first and foremost, a fully CGI flick that markets the impending arrival of LoL's first season of the year. However, while the promo video has been created to increase the hype ahead of season 1's debut on January 9, it simultaneously acts as an epilogue to Arcane season 2 and/or a prelude to the highly-rated Netflix show's first follow-up.

Set to Teya's 'Bite Marks' track, the cinematic, which was developed in partnership with Fortiche – the French animation studio that made Arcane alongside Netflix and Riot – is filled with references to the trio's first animated LoL adaptation. The most obvious of those call-backs is Mel Medarda, the daughter of Noxian warlord Ambessa. As I mentioned in my Arcane season 2 act 3 ending explained piece, the series' final episode revealed Mel was traveling back to her homeland of Noxus to seek answers to, well, everything that happened to her in season 2. That includes how she acquired mage powers, why her now-deceased mother left Noxus with the aim of returning to conquer it with Piltover's Hextech, and what role The Black Rose has in the Medarda family's past, present and future.

Speaking of The Black Rose, there are numerous references to the clandestine organization throughout LoL's latest trailer. We get glimpses of LeBlanc, Vladimir, and Elise – three of The Black Rose's four most infamous members – and shots of Mel fighting off some of the cabal's dark magic-wielding magicians. Oh, and don't overlook LeBlanc's comment about "Piltover being a lesson" for The Black Rose to learn from. Clearly, this is a nod to events that transpired in Arcane seasons 1 and 2, so there's no denying that the Netflix series' storyline is reshaping, and maybe even retconning aspects of, the wider LoL universe.

Are LoL's revamped takes on Vladimir and LeBlanc going to be the ones we see in Arcane's follow-up project? (Image credit: Riot Games/Fortiche)

There are plenty of other Noxian characters shown throughout the teaser, too. The assassin known as Katarina is the individual who spars with the arachnid-like Elis, while the hulking Freljord troll called Trundle is seen squaring off with, and then being defeated by Darius, who's one of the most iconic Noxian champions in LoL. The latter duel certainly suggests that Noxus is continuing its global expansion by trying to conquer the Freljord, which could be a secondary location that's explored in future spin-offs of Arcane, aka one of the best Netflix shows ever made.

Curiously, there's no mention or look at Noxus' Grand General Swain, who was teased in Arcane's last-ever episode by way of the six-eyed raven. Over on LoL's official Reddit page, some fans have been left baffled by his absence from the teaser, especially with how integral he is to this Runeterran region. Nevertheless, you can be sure that he'll be part of season 1 of LoL's 2025 calendar.

Darius could have an important role to play in any Arcane sequel show (Image credit: Riot Games/Fortiche)

Ever since Arcane season 2 ended last November, fans have been clamoring for more information on the projects that Riot, Netflix, and Fortiche are cooking up. Right now, there's very little we know about these productions, which are still in the early stages of development, but some details have been revealed by Arcane's co-creator Christian Linke since the show's run on the world's best streaming service came to an end.

Speaking exclusively to TechRadar before season 2's debut on November 9, 2024, Linke and fellow co-creator Alex Yee claimed it'll be a long time before the next League of Legends adaptation lands on Netflix. In a separate interview with TechRadar that took place after its final chapter, Linke also denied that three Arcane spin-offs were in the works, before adding that he couldn't say when they'd be released or which LoL champions would appear in future projects. Thanks to this new LoL cinematic, it seems we may have some answers concerning the latter.

I covered Arcane season 2 extensively before, during, and post-release, so be sure to read more exclusives from yours truly about the series and its sequels in the section below.

You might also like

• Arcane co-creator vows 'we will learn from it' after fan frustrations of the Netflix show's 'rushed' final season
• Arcane co-creator says 'I can neither confirm or deny' if a big Jinx season 2 finale fan theory is true, but I think his reaction says otherwise
• Arcane season 2 rewrote the 'truth' of some vital League of Legends lore, but the Netflix show's creators deny they're breaking the game series' canon

About 12 out of 15 of the best fitness apps actively share your personal data with third parties, de-facto selling out your privacy. Among these, Strava and Fitbit are the most data-hungry, collecting 84% of all potential data points.

These are some of the worrying findings from new research released by Surfshark, one of the best VPN services on the market, after looking at data collection and sharing practices of the most popular fitness mobile applications.

"Our research shows that free apps share significantly more data with third parties compared to paid apps, highlighting the importance of evaluating privacy implications," said Tomas Stamulis, Chief Security Officer at Surfshark.

The hidden price of at-home training

To determine the real price of (often free) at-home training, the Surfshark team analyzed the 15 top fitness mobile applications around. These include exercise trackers, workout apps, and personal training platforms.

Experts sourced the data collection information for each app from its Apple App Store page on December 30, 2024. The App Store provides a list of 35 unique data points categorized into 16 unique data point categories. The team looked at the data set according to the number, type, and handling of the data points collected by each app.

Surfshark unveiled a pretty worrying scenario for mobile fitness fanatics. As mentioned earlier, 80% of the analyzed apps share users' tracked data with third parties. These details include device locations, emails, user IDs, device IDs, or profiles. Nike Training Club leads the category, with four types of tracking data shared with third parties. These are coarse location (approximate, generally within a city block), some sensitive info, device ID, and product interaction.

In Apple's own words, "Tracking refers to the act of linking user or device data collected from your app with user or device data collected from other companies’ apps, websites, or offline properties for targeted advertising or advertising measurement purposes. Tracking also refers to sharing user or device data with data brokers."

Most apps (13 out of 15) also collect health and fitness information directly linked to users – Centr and Peloton are the only two applications that don't.

Overall, as the graph above shows, the analyzed apps collect an average of 12 different types of information out of the 35 potential data points available – with the least privacy-friendly storing nearly twice as much.

Let's look at the data. Strava and Fitbit came out as the most data-hungry apps, for example, both gathering 21 unique types of data. In comparison, the most private workout application, Centr, collects just three types of data (User ID, Product Interaction, and Crash Data) with only one of these contributing to user tracking.

Worse still, three apps collect some very sensitive information such as racial or ethnic background, sexual orientation, pregnancy or childbirth details, disability status, religious or philosophical beliefs, trade union membership, political opinions, genetic information, or biometric data. These include the Nike Training Club app.

Location data is another piece of information many fitness apps collect. Four apps, including popular running applications like Runna and Strava, collect precise location data linked to the user. Five apps collect only coarse location data, with two of these (Nike Training Club and Peloton) sharing this information with third parties.

As mentioned earlier, free applications collect and share the most data. After all, the only way they can make a profit is to sell your data to data brokers or run invasive ads on the app. This is why Stamulis from Surfshark suggests upgrading to a paid subscription whenever possible.

He also recommends considering whether the app can function without granting permissions that may not be truly necessary. "If such options aren’t provided, it raises important questions about the intent behind the data collection," he added.

• Honda reconfirms its commitment to line-up of bold new EVs
• The Honda 0 SUV will launch in North America in early 2026
• Asimo operating system will bring advanced levels of automated driving

Fear not faithful reader, you are not suffering from chronic CES déjà vu. Because, yes, Honda has already revealed its plans for the upcoming 0 Series electric vehicle range, showcasing what it called the Saloon and the Space-Hub at last year’s show... complete with eerily similar concept cars.

But the Japanese marque says that things are moving on at pace and it has reaffirmed its commitment to the futuristic and gloriously wedge-like shapes with the unveiling of its 0 SUV and 0 Saloon prototype models at this years Las Vegas-based tech-fest.

While the Saloon looks a lot like last year’s concept, albeit with a flashy new paint job and tweaked headlight design, the 0 SUV arguably gives us our best look at what this upcoming line of forward-thinking EVs will look like.

Gone is the boxy, almost MPV body shape of last year’s concept, as we are introduced to a more traditional-looking SUV front end, complete with a sweeping, almost estate or shooting brake-esque rear.

The Japanese marque claims the 0 SUV will arrive in North American markets in early 2026, with the low and long Saloon model following shortly after. Global markets, including Japan and Europe, are also touted to receive both vehicles.

(Image credit: Honda)

Honda previously hinted at its “Thin, Light and Wise” philosophy, which it hopes will reduce the size and weight of battery packs to ensure a line-up of lighter, more elegant and more efficient electric vehicles.

Although it didn't go so far as to reveal battery specs, Honda instead honed in on the ‘Wise’ element of the aforementioned mantra at this year’s show, revealing that it has developed an in-house operating system (Asimo OS) that it says will allow for Level 3 ‘eyes-off’ automated driving.

Those with long enough memories will recall Asimo, the cute little robot that was often found kicking a ball or falling down stairs. Well, Honda says it has borrowed bits of Asimo’s brain in order to develop the latest OS.

The company claims that advances in its robotics technology, including developments in systems that recognize external environments and human behavior, will equip its future vehicles with genuinely useful AI technology that can predict what a driver wants and create more reliable driver assistance systems.

Analysis: Honda goes back to being bold

(Image credit: Honda)

Although Honda has yet to confirm any technical detail of its upcoming 0 Series, which will encompass seven individual models by 2030, its commitment to the latest tech and bold styling is proof that the marque is serious about EVs this time.

Its Honda e was arguably ahead of its time, but the brand has stagnated in recent years, churning out reliable but relatively vanilla SUVs and small city cars with combustion engines or basic hybrid systems.

Only the Civic Type-R has remained as a shining example of what the renowned engineering firm is capable of when it comes to pure driving thrills, but even that faces an uncertain future.

It feels like the time is right for Honda to return to offering reliable but exciting vehicles that stand out from the crowd. We can't be the only ones thinking that an all-electric Series 0 Saloon Type-R would be a serious statement of intent from the brand.

TechRadar will be extensively covering this year's CES, and will bring you all of the big announcements as they happen. Head over to our CES 2025 news page for the latest stories and our hands-on verdicts on everything from 8K TVs and foldable displays to new phones, laptops, smart home gadgets, and the latest in AI.

And don’t forget to follow us on TikTok and WhatsApp for the latest from the CES show floor!

You might also like

• The Sony Afeela EV is finally here, but it could struggle to catch the Lucid Air – here's why
• I review EVs for a living – here are 5 ways I'm slashing my charging bills in 2025
• Goodbye, buttons – BMW's new Panoramic iDrive system gives all of its cars a massive head-up display with AI assistant

Prosecutors dropped the two criminal cases against Trump after he won the 2024 election, and the final report by Smith may be the last chance for prosecutors to explain their decisions.

(Image credit: Drew Angerer)

• CISA addS three new bugs to KEV - two in Mitel’s MiCollab, and one in Oracle WebLogic Server
• The bugs allowed crooks to read sensitive files and take over vulnerable endpoints
• Federal agencies have until late January 2025 to deploy the patch

The US Cybersecurity and Infrastructure Security Agency (CISA) HAS added three new flaws to its Exploited Vulnerabilities Catalog (KEV), signalling in-the-wild abuse, and giving federal agencies a deadline to patch things up.

Two of the three flaws are found in Mitel’s MiCollab unified communications platform. One is a critical path traversal vulnerability, tracked as CVE-2024-41713.

By abusing this bug, threat actors can run admin actions and access user and network information.

A deadline to patch

"A successful exploit of this vulnerability could allow an attacker to gain unauthorized access, with potential impacts to the confidentiality, integrity, and availability of the system. This vulnerability is exploitable without authentication," MiCollab said.

"If the vulnerability is successfully exploited, an attacker could gain unauthenticated access to provisioning information including non-sensitive user and network information and perform unauthorized administrative actions on the MiCollab Server."

The second bug is tracked as CVE-2024-55550, another path traversal vulnerability granting admin privileges. The impact of this bug is limited, however, since it doesn’t allow threat actors to escalate privileges, or access files with sensitive information. Therefore, the severity of this bug was assigned to “medium” - 4.4/10.

The third bug is found in Oracle WebLogic Server, and is tracked as CVE-2020-2883. It was patched in April 2020, and grants threat actors the ability to remotely access vulnerable endpoints.

Now, with all three vulnerabilities being added to KEV, federal agencies have until January 28 to apply the fixes, or stop using the products altogether. 8. "These types of vulnerabilities are frequent attack vectors for malicious cyber actors and pose significant risks to the federal enterprise," CISA said.

Mitel’s MiCollab is a popular unified communications platform, and as such - a major target for cybercriminals. In early December this year, the company patched a three-month-old zero-day vulnerability that allowed crooks to read sensitive files.

Via BleepingComputer

You might also like

• Mitel collaboration software zero-day strings along a previously patched vulnerability
• Here's a list of the best antivirus tools on offer
• These are the best endpoint protection tools right now

• An accurate recreation of the Nintendo Switch 2 logo may have surfaced
• It indicates that the logo will be the same as that as the Nintendo Switch, albeit with a large 2 added
• It follows an abundance of other recent Nintendo Switch 2 leaks

It’s clear that Nintendo has lost control of the narrative surrounding the Nintendo Switch 2, with other companies showing off detailed mockups allegedly based on the unreleased console at CES 2025 and an apparent photo of its controller hitting social media just a few days ago. Now we might even know what the console’s logo looks like, all before it's been officially unveiled.

According to Felipe Lima, the co-creator of Brazilian Nintendo fan site Universo Nintendo, the Nintendo Switch 2 logo will look an awful lot like the emblem for the original Nintendo Switch. It will feature a prominent '2', however, clearly differentiating it from its predecessor.

In a recent X / Twitter post which we have machine translated, Lima claims that the information comes from “a reliable source” that was able to send them an image of the console itself. They even offer their own recreation of the logo, presumably in an effort to avoid a copyright takedown notice, which they state is “exactly like” the real thing.

Previous leaks have also hinted that the new console will feature very similar branding to the Nintendo Switch, which lends Lima’s claims some credibility. With alleged photos of the controllers out in the wild, it also doesn't seem implausible that they might have been able get their hands on an image showing the whole thing.

I do have to say, however, that if this leak is accurate it’s hardly the most exciting news. I have always loved the imaginative branding of Nintendo consoles, whether that’s the brilliant Nintendo DS or older favorites like the Nintendo GameCube. Labeling a product the 'Nintendo Switch 2' seems pretty tame and much more in line with the naming conventions of competitors like Sony and its PlayStation lineup.

Even so, it could provide an interesting insight into how Nintendo views the upcoming system. This is definitely shaping up to be much more of an iterative release than what we’ve seen in the past, but I’m still optimistic that some cool new feature (potentially the ability to use the Joy-Con controllers like a mouse) will help set it apart.

You might also like...

• The Nintendo Switch 2 could be revealed "earlier than planned" thanks to recent leaks, per industry analyst
• Will the Nintendo Switch 2 feature AI upscaling? This patent suggests it’s on the cards
• New Nintendo Switch 2 leak gives us more information about the console's potential design, dock, and official name

2024 has been a year of rapid AI adoption, with many businesses scrambling to capitalize on the latest advancements for fear of being left behind. However, despite significant investment, organizations often struggle to realize tangible benefits from their AI initiatives. In fact, reports suggest that while 68% of large companies have integrated AI, a quarter of IT professionals regret rapid AI adoption, and two-thirds wish they had chosen technologies more carefully.

Arguably, the root of this issue lies in a lack of control. Organizations are struggling to implement AI tools in a manner that not only brings benefits, but also does not compromise their data privacy. In 2025, businesses need to ensure they choose the right AI tool for the job while retaining the control and privacy their data needs.

Identify why you want an AI tool

Before embarking on an AI initiative, it's crucial to define clear objectives. What specific problem are you trying to solve? What value do you expect to derive from AI? Is it threat intelligence, enhanced decision-making, or improved customer experience? It is only once these goals are identified that a business can know what type of AI they need.

Crucial to this is finding the right tool for the job. The first step is to understand that while Large Language Models (LLMs) have been dominating the headlines and fueling the hype, they are not the only form of AI model. Instead, there are a number of different tools available that are focused on specialist tasks and solutions that may not only be more suitable, but also more able.

This is because specialist AI is designed to tackle a specific task as opposed to being trained to offer a solution for everyone – both professionally and personally. What’s more, unlike LLMs, which are trained on vast, often uncurated datasets, specialist AI models focus only on relevant data, resulting in higher accuracy and efficiency. Finally, specialist AI models are more efficient in terms of computational resources and energy consumption, making them more cost-effective, less environmentally impactful, and faster to implement.

It is crucial to consider all options when seeking the right tool for the job to make sure you retain control of your data and are focused on the job, not the hype. After all, if you choose the wrong tool then you will lose control of your data the second you sign in.

The right data and the right privacy

A heavily touted advantage of LLMs is that they are trained on vast amounts of data and are therefore able to provide insights and generate content for organizations from all industries and regions. However, while this is indeed an advantage for people looking for a tool able to provide such scope, in most business cases this is in fact a negative.

This is because being trained on such huge pools of data can cause a reduction in the quality, accuracy, and integrity of that data. What’s more, it is often difficult to discover what data the LLM was specifically trained on in order to validate it. This is a particular challenge for businesses who need a high degree of transparency and accuracy with their outputs as LLMs have been shown to be prone to hallucinations and biases as a result of learning from such vast and varied data.

Specialist AI tools, meanwhile, can offer users the option to choose the data the model is trained on with the customer able to see and curate those sources with transparency. For example a Small Language Model (SLM) AI tool can be fed a number of sources in the form of thesauruses so it can accurately understand the specific needs of a user – this includes not just languages in a formal sense, but also the ability to understand the technical jargon and expertise of a company’s industry as well as that company’s own annotations and coded shorthands. This can offer a highly efficient approach for an organization when it comes to AI adoption as it is the tool that is adapted to the user, rather than staff having to be trained for the tool.

A further aspect to consider is the privacy of that data. It is crucial that any data an organization gives an AI tool to tailor its training and make it work for them is kept private and confidential and not shared externally. This is important not just to protect a business from breaches and to keep their sensitive and confidential information secret, but also for regulatory and legal reasons with many industries having strict control over many aspects of financial, health, and PII data. This also goes for data used as part of prompts and AI analysis once the tool is being used too, with any data that passes through or is subjected to an AI tool needing to be secure and private.

For example, LLMs often require vast amounts of data to be shared with third-party providers. This can pose significant risks to sensitive information, particularly for businesses operating in highly regulated industries. In contrast, private AI models, such as specialist AI, can be deployed within a secure, zero-trust environment, ensuring that data remains confidential and protected from unauthorized access.

By opting for a private AI solution, organizations can safeguard their intellectual property and maintain control over their data, mitigating the potential for data breaches and reputational damage. They are therefore able to use the AI with even their most confidential and regulated data as opposed to having to limit it to publicly available material, thus maximizing the potential gains of the tool.

Integration, control and security

It is imperative that an organization has complete control over how the AI is implemented into their workflow and system with all data access tightly controlled and transparent. This is particularly important in industries working with sensitive and regulated data as they need to be able to report on how that data has been used and who has had access to it.

The importance of this has been highlighted in 2024 by a number of surveys and reports uncovering the prevalence of data exposure due to AI tools. For example, research by Syrenis found that 71% of AI users regret sharing their data with AI tools after realizing the extent of what was shared, while a RiverSafe survey of CISOs revealed that one in five UK companies exposed sensitive corporate data as a result of employees using AI tools.

To put it bluntly, if an AI tool, or indeed any tool, harvests a business's data or shares their information externally, then that business is at risk of a breach and could be at risk of failing compliance requirements.

When implementing new AI tools pay close attention to how they are integrated within a business's existing architecture and ensure that it does not require data to be stored outside of your control. For example, if a business chooses to use a cloud-based AI tool, it is crucial to ensure they have the ability to either host that cloud structure on their own system, or to prevent third-party access to the data and protect it from cyberattacks such as ransomware. This can be achieved through combining the cloud provider’s infrastructure with your own decentralized storage, for example blockchain, and implementing strict access control and encryption.

These same encryption and access measures can also ensure you have control over what data is accessed and by who, ensuring that your information is protected by least privileged access with nobody able to access data they do not need. Homomorphic encryption can also ensure that data can remain encrypted at rest, in transit and in use with search and computations possible on the fully encrypted data. However, while the security and privacy of the data is crucial, it is also important to check the scalability and speed of the system to ensure the AI is able to provide the real-time insights and services that are needed in today’s market.

Final thoughts

The successful implementation of AI hinges on a balanced approach that prioritizes control, data privacy, and security. By carefully selecting AI tools tailored to specific needs, prioritizing data quality and transparency, and implementing robust security measures, organizations can harness the power of AI while mitigating potential risks.

As the AI landscape continues to evolve, it is imperative to stay informed about emerging technologies and best practices to ensure that AI is used responsibly and ethically. By adopting a proactive and strategic approach, organizations can unlock the full potential of AI and drive innovation while safeguarding their interests by retaining control.

We've featured the best AI website builder.

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

• January is historically a busy month for fitness and gyms
• The second Tuesday of January is often referred to as Quitter's Day
• Strava says it has been the most common activity break for its users every year for more than five years.

If you've returned from the holidays to find your gym mobbed by the teeming masses of January fitness enthusiasts, then you don't have long to wait before people start to fall off the wagon and things quieten down again.

Every year, people flock to the gym hoping to kickstart their New Year's resolutions and finally tackle that health push they've been talking about. As we all know though, plenty of those who've started won't finish. It's a tale as old as time.

Not only is the January fall-off an age-old tale in fitness circles, it's also a scientifically proven phenomenon. Strava has the data to prove it, such that it can predict the exact date every year, and Quitter's Day 2025 is fast approaching – here's when your gym will start to quieten down.

When is Quitter's Day 2025?

(Image credit: Future)

Strava says that Quitter's Day falls on the second Tuesday of January each year. For more than five years, it's been the most common activity break for Strava users, and this year it falls on January 14.

So if you head to your gym on that day and think to yourself that things are a little quieter than they've been for the last couple of weeks, you probably aren't imagining it.

Not that we're rooting for anyone to fail, of course. In fact, Strava has released a slew of helpful tips backed by new data to help people keep up their fitness commitments in 2025.

Strava, one of the best fitness apps, especially for outdoor adventurers, says the most popular goal-setting trend is a week-by-week approach. In any given month only one-third of Strava runners who met or were close to meeting their goals actually tracked perfectly against them. There are also seasonal cycles – runners are most likely to be on track in the summer months (in the US and UK) of July through September, while January and February are the months athletes are least likely to be hitting their goals.

Tips from Strava to stay on top of your goals include joining a running club or working out in a group, introducing variety to keep things fresh, and making rest a priority.

So if you are starting to feel that January struggle, try those helpful tips, and know that if you make it past January 14 then that's a big deal. And if you're a more committed fitness enthusiast you can start counting down the days until you're able to find a bit more space in the locker room, and – finally – some free treadmills.

You may also like

• Tech Resolutions 2025 – 15 ways to survive January and upgrade your life with tech this year
• Look away now if you're a personal trainer: tech, data, and AI are forecast to be the biggest fitness trends of 2025
• I'm a fitness editor, and I swapped weights for resistance bands for six whole months – and I don't regret it

It’s official: Samsung’s next Galaxy Unpacked showcase is set for January 22, meaning we haven’t got long to wait before the curtain is lifted on what could be some of the most exciting (and bar-setting) consumer tech of 2025.

But what, specifically, are we expecting to see unveiled in San Jose this year? Naturally, Samsung itself hasn’t confirmed anything just yet, but countless leaks and rumors have got us excited for the potential reveal of new Galaxy phones, software features, and maybe even an all-new premium mixed reality headset.

Below, we’ve rounded up five of the biggest announcements to expect at Samsung Galaxy Unpacked 2025.

Samsung Galaxy S25 Ultra

The Samsung Galaxy S24 Ultra (Image credit: Future | Alex Walker-Todd)

At this point, the Samsung Galaxy S25 Ultra has been the subject of more leaks than the Titanic, and we’re fairly certain that this presumed successor to the Galaxy S24 Ultra will spearhead Samsung’s next flagship smartphone lineup.

As for what upgrades the Galaxy S25 Ultra could bring to the table, we’ve heard talk of a more rounded design (akin to the Google Pixel 9 Pro XL and iPhone 16 Pro Max), a 6.9-inch display, a new 50MP ultra-wide camera, 25W wireless charging, and a new 16GB RAM capacity (in the 512GB model, at least).

These upgrades will almost certainly be underpinned by the all-new (and all-powerful) Snapdragon 8 Elite chipset, which has already made a name for itself in the OnePlus 13 and Asus ROG Phone 9 Pro. Having put Qualcomm’s latest mobile silicon through a series of benchmark tests, TechRadar’s US Mobiles Editor, Philip Berne, said the Snapdragon 8 Elite “will change everything about smartphones,” so the Galaxy S25 Ultra looks set to be a very powerful handset indeed (expect it to dominate best Android phone conversations in 2025).

Samsung Galaxy S25

The Galaxy S25 could look a lot like the Galaxy S24 (pictured) (Image credit: Future | Roland Moore-Colyer)

The Galaxy S25 Ultra won’t be the only new smartphone kid on the block this month. The new flagship model will almost certainly be joined by two base-level devices, which we’re expecting to be the Samsung Galaxy S25 and Galaxy S25 Plus.

Like their expected Ultra-branded sibling, both the Galaxy S25 and Galaxy S25 Plus are likely to use the Snapdragon 8 Elite chipset (there’s a slight chance that Samsung could deploy the Exynos 2500 in certain regions) and boast more rounded designs than their respective predecessors. Other rumored upgrades include 12GB of RAM for all storage configurations and support for Qi2 magnetic wireless charging.

Interestingly, we’ve heard talk of a Samsung Galaxy S25 Slim potentially debuting at Galaxy Unpacked this year, but the tipsters are divided over that model’s ETA. If indeed we do see a svelte version of Samsung’s new base model announced in January, it isn’t likely to begin shipping until later in the year, but it’s nonetheless exciting to think that a smartphone-related surprise could be on the cards in San Jose.

Galaxy AI

Circle to Search on the Galaxy S24 Ultra (Image credit: Samsung)

Samsung is going hard on AI-related marketing for its upcoming Galaxy Unpacked showcase, suggesting big Galaxy AI upgrades are on the agenda. Accompanying copy for the company’s event announcement reads: “Get ready for AI that is more natural and intuitive. The next evolution of Galaxy AI is coming and it’s going to change the way you interact with the world every day. The new Galaxy S series is about to set the bar once again for mobile AI experiences now and into the future.”

In other words, some big new software features look set to debut alongside those aforementioned phones. Among them will almost certainly be a major upgrade for Samsung’s Bixby voice assistant, and prominent leaker @UniverseIce claims that Galaxy Unpacked will see Samsung “tell Apple what leading AI is" and that "many new AI functions of [the] S25 have not been leaked so far.” Apple Intelligence, who?

Samsung Galaxy Ring 2

The original Galaxy Ring was announced at Galaxy Unpacked 2024 (Image credit: Samsung)

The Samsung Galaxy Ring was teased at last year’s January Unpacked event and only went on sale as recently as July, but reports suggest Samsung could pull the same trick with the Samsung Galaxy Ring 2 at this year’s January Unpacked event.

Of course, we don’t expect to hear (or see) much about this potential successor to the best smart ring on the market in San Jose. But Samsung will reportedly offer us an "introduction" to the Galaxy Ring 2 at Unpacked, with the new device rumored to boast more accurate sensors, improved AI, and better battery life.

Project Moohan

Samsung's Project Moohan mixed-reality headset (Image credit: Google)

Project Moohan, Samsung's mixed-reality headset and “the first platform built entirely for the Gemini era,” was technically announced in December last year, but there’s a chance that Samsung could give Unpacked attendees a chance to try out the Android XR-based product (or some version of it) for themselves on the ground in San Jose.

Project Moohan is still just a developer kit, but Google and Samsung have confirmed that "the first products based on this platform will launch in 2025,” so a close-to-finished wearable device must surely be just around the corner. Here’s hoping, anyway.

You might also like

• Samsung Galaxy S25 specs predictions: all the key rumored specs for every model
• Samsung Galaxy S25 price rumors: how much is the S25 line likely to cost?
• Samsung Galaxy S25 rumored features: the key tipped upgrades for the S25 line

In today’s fast-evolving world, the drive for digital transformation has become more complex and nuanced than ever. Organizations aiming to harness the benefits of this transformation now face a landscape shaped not only by technological advances but also by economic uncertainly, an unstable geo-political landscape and environmental challenges.

It’s why the role of the Chief Data Officer (CDO) has never been more important. Tasked with guiding organizations through this digital overhaul, CDOs are pivotal in aligning data management with strategic objectives.

However, the challenges that come with this task are so vast and intricate that they cannot meet these demands alone – technology in all its forms, partners including governments, and employees – all play a role in modernizing organizations to drive value for all stakeholders.

Harnessing new technology – the role of AI

One of the most powerful tools in the CDO’s arsenal today is artificial intelligence (AI). With intuitive, code-free interfaces, AI offers immense potential to streamline data management and bridge gaps in technical knowledge across teams.

Generative AI, in particular, can greatly enhance user experiences by quickly synthesizing and delivering valuable insights. However, the potential of these technologies is strictly contingent on the quality of the underlying data. For Generative AI to be effective, it requires rigorous data quality standards to avoid common pitfalls like data drift and hallucination, where the AI might generate irrelevant or inaccurate information.

Integrating Generative AI into the Shared Services sector has elevated key functions such as payroll, skillset analysis, deployment strategies, and even manufacturing capabilities. Such applications underscore the role of AI not just as an automation tool, but as an essential collaborator in digital transformation.

The need for greater collaboration and its role in digital transformation

Yet, technology alone cannot drive the kind of transformation that today’s organizations require. Meaningful change demands collaboration which is especially critical in sectors like defense, where data-driven strategies are central to mission success. By enhancing decision-making and situational awareness, advanced data analytics and AI tools have proven to be invaluable in defense operations, from logistics to supply chain management.

The ultimate objective of digital transformation in defense is to build a military force that is agile, informed, and ready to respond to a diverse set of threats.

However, achieving this goal is no small feat. Defense organizations, often reliant on complex legacy systems, face significant hurdles when integrating new technologies into existing infrastructures. Consider how different pieces of military equipment must work together in the field.

Cross-industry data standards are crucial to ensuring that equipment developed by different companies can interoperate seamlessly—whether in intelligence, surveillance, or tactical operations.

Collaboration tools have become essential in this context, allowing stakeholders to share critical documents, data, and insights in real-time. The industry has taken a step forward by working with government partners to establish shared data standards, a move that not only enhances decision-making but also lays the groundwork for improved interoperability across the defense sector.

While the defense sector’s needs are unique, its lessons around agile delivery, cross-organizational collaboration, and rigorous data management have the potential to drive innovation much more widely.

In a world where change is the only constant, digital transformation will remain challenging.

Only through a blend of advanced technology, robust data practices and strong collaboration, can CDOs and their organizations lay a solid foundation to overcome these challenges.

When harnessed effectively, these tools and strategies can unlock unprecedented potential, driving innovation and resilience across a great variety of industries and organizations.

We've featured the best productivity tool.

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

• Amazon Web Services announces $11 billion investment plan in Georgia
• The funding will support AI data centers in Butts and Douglas counties
• At least 550 high-skilled roles will be created as a result, says Amazon

Amazon Web Services (AWS) has announced an estimated $11 billion in funding to expand its cloud computing and AI infrastructure in Georgia.

The company’s plans come in response to ever-increasing demand for compute power following a surge in artificial intelligence technologies, many of which are hosted on its servers.

Amazon says although its GPUs are capable of running AI and ML models, they can also be tweaked to efficiently run more traditional workloads, so their uses won’t just be limited to intense work.

AWS invests $11 billion in Georgia

As part of the investment, Amazon hopes to support at least 550 new high-skilled jobs in the region. The company as a whole, including its ecommerce business and other divisions, already employs 34,000 full and part-time workers in the state.

Details of AWS’s upcoming investments in Georgia have not been revealed, but the company did confirm Butts and Douglas counties will benefit from the cash injection, towards the center and north of the state.

Of the $1.2 trillion Amazon has invested across the US since 2010, Georgia has received $18.5 billion in funding, or 1.54%.

“We are excited to deepen our partnership with the state of Georgia and contribute to the continued advancement of its thriving technology landscape," noted AWS VP of Economic Development Roger Wehner.

Douglas County Commission Chairwoman Dr Romona Jackson Jones added: “We are pleased to welcome AWS to bring the most secure, resilient, and advanced cloud computing technology to the county, and with it hundreds of high-paying, high-tech jobs that will benefit our community and our citizens for the long term.”

You might also like

• These are the best AI tools and best AI writers
• Access the best cloud storage in our handy guide
• AWS reveals $10bn cloud and AI investment in Ohio, bringing hundreds of jobs

• Microsoft is focused on AI-powered Windows 11 and Copilot+ PCs for the foreseeable future
• October 14, 2025 is still Windows 10’s deprecation date, pushing users toward Windows 11 or newer PCs
• Microsoft's deep AI integration in Copilot+ PCs aims to redefine how we interact with our PCs but raises concerns

At CES 2025, we’ve seemingly just had confirmation that Microsoft is not releasing the next generation of Windows, which many of us assume will be called Windows 12, just yet. Instead, Microsoft is pushing ahead with its vision of AI-powered PCs (which it calls Copilot+ PCs) and intends to continue to integrate artificial intelligence into Windows 11.

Microsoft, eager to capitalize on AI buzz, is betting big on tools like ChatGPT and its own AI assistant, Copilot, which is accessible directly from the taskbar, becoming an everyday part of our digital work and personal lives.

According to Windows Latest, Microsoft claims that Copilot+ PCs are up to 50% faster than an M3 MacBook. From the Copilot+ PCs we’ve reviewed so far, such as the Surface Laptop 7, we’ve certainly been impressed with their performance, though the jury is still out on whether the AI features are all that useful.

(Image credit: Lenovo) Windows 10's departure and Microsoft's unwavering focus on its AI vision

There is also an emphasis on future versions of Windows 11 because Windows 10 is nearing its end-of-support date, which is on October 14, 2025. Yusuf Mehdi, Microsoft’s Executive Vice President and consumer Chief Marketing Officer, reinforced this at CES 2025, stating that Windows 10 is still on track to be deprecated on this date.

When October 14 rolls around, Microsoft will stop releasing regular updates (including security updates) for Windows 10 and after that, you’ll either have to buy a newer Windows 11 PC or pay for extra security updates for a limited time (and it’s clear that Microsoft would rather all Windows 10 users upgrade to Windows 11).

So, it looks like Windows 12 won’t be coming in the foreseeable future as Microsoft is focusing on getting its users to upgrade to Windows 11. It’s also focusing on winning over people with its Copilot+ PCs, offering a pretty bold vision of what these devices will be capable of and how much they’ll change how we interact with our PCs. Windows Latest points out that Windows 11 came out during the pandemic when people were buying new PCs and Microsoft saw an opportunity for a PC sales push, and that it’ll wait for another opportunity like this to launch Windows 12 (though hopefully not another global pandemic).

I can see why some people are perhaps disappointed that we’re not getting a bold new operating system, or reluctant to move on from Windows 10, which is still very much Microsoft’s most popular operating system. I also think that Microsoft has a long way to go to actually convincing people that its AI improvements are worth people’s time, as so far, features like Copilot and Recall, have been shown to be somewhat lackluster or just outright risky in terms of aspects like security. I don’t think it’s impossible for Microsoft to pull off its ambitions, but that ends up being so will depend to innovations that it has yet to show.

I do think Microsoft has a chance here to experiment with integrating AI-powered software and tools into Windows 11, and there's certainly huge potential for artificial intelligence to redefine how we interact with PCs.

However, experiments come with risks, and Microsoft would be remiss not to learn from this venture and it should pay attention closely to feedback from its users. If its involvement with AI doesn't succeed (or at least not immediately), I think it should consider designing Windows 12 with an option to operate without heavy AI integration, using Windows 12 as a clean break should its AI gamble not pay off.

In the meantime, I'm glad that Microsoft doesn't appear to be distracted with making Windows 12 whilst there's still a lot of work to be done with Windows 11 to make it more popular - especially as there will soon be a lot of Windows 10 users looking to upgrade.

YOU MIGHT ALSO LIKE...

• Microsoft really wants you to update to Windows 11 in 2025
• Microsoft nixes Windows 11’s “Suggested actions” feature: a smart idea that didn’t quite work
• Microsoft embarrasses itself with Windows 10 pop-up that hogs the desktop urging an upgrade to Windows 11 – then promptly crashes

Upgrade your kitchen with this elegant 10-piece stainless steel Calphalon cookware set for only $170.

This charger is perfect not just for your iPhone but also for tablets, MacBooks and more -- and it’s only $25.

President-elect Donald Trump is scheduled to be sentenced on Jan. 10 in New York. His latest legal move to block sentencing comes after a New York appeals court rejected a similar appeal Tuesday.

(Image credit: Scott Olson)