TechRadar News

• Military-grade AI server wins iF Design Award for its rugged modular design
• Features include electromagnetic shielding, shock resistance, and moisture protection
• Operates in extreme conditions from -55°C to 70°C without performance loss

The iF Design Awards are an international competition recognizing excellence in design across categories such as product, communication, packaging, and user interface.

Organized by iF International Forum Design GmbH in Germany, past winners include products from the likes of Apple, Samsung, Sony, and Lenovo - in 2024, one of the winners included South Korea’s KT (Korea Telecom) with its gorgeous skinnable Wi-Fi router.

There are often unusual winners, and this year the list includes, of all things, a Chinese military AI server. If we’re honest, we can see why it won. It’s unquestionably a thing of beauty, with its olive-green and metallic finish, reinforced metal frame, eight vertical modules with LED indicators, locking mechanisms, connector ports, and a shock-absorbing base - but even so, it’s an unexpected inclusion.

Designed for on-the-go

Created by China’s Chaoyue Technology, the Global "Special Forces" server is described as a “fully environmentally compatible, electromagnetic interference-resistant, high-computing power, and 100% domestically produced server.”

Features include environmental adaptability, one-click control functions, and integrated computing, storage, switching, confidentiality, and display capabilities.

The Special Forces server is designed for use in airborne, ship-borne, and in-vehicle situations. It’s built for harsh environments (frankly it wouldn’t be much use if it wasn’t), and offers anti-electromagnetic interference protection from 10K–10GHz, resistance to strong vibrations and 50g impact, and IP66-rated moisture protection with 95% humidity tolerance.

It also reportedly delivers reliable performance in temperatures from -55°C to 70°C, and durability against salt spray exposure for up to 10 years without damage.

The product, which took up to a year to create and was launched in 2023, is the first domestically produced independent platform LRM server in China.

It features advanced DVI+USB matrix switching, something which Chaoyue Technology says is currently only achievable by two companies in China, alongside AI-assisted control, modular configuration, and energy-efficient design principles.

You might also like

• Check out our list of the best servers for small businesses
• This is the best dedicated server hosting you can get right now
• Latest HPE ProLiant Gen11 server lineup features AMD EPYC 9005 processors

• Microsoft warns of new version of the XCSSET infostealer
• It comes with new obfuscation, infection, and persistence techniques
• It was seen in "limited" attacks in the wild

A new variant of a known macOS malware is making rounds on the internet, targeting users through infected Xcode projects.

Researchers from the Microsoft Threat Intelligence team said that the modular malware is seen in “limited attacks” at this time, but suggested that people should still keep their guard up.

According to the researchers, this is the first upgrade to XCSSET in three years. It now has enhanced obfuscation methods, updated persistence mechanisms, and new infection strategies.

Scrutinize Xcode projects

“These enhanced features add to this malware family’s previously known capabilities, like targeting digital wallets, collecting data from the Notes app, and exfiltrating system information and files,” Microsoft said.

Microsoft first reported of this new XCSSET strain in mid-February this year, and has now come forward with an in-depth analysis.

Xcode is Apple's official integrated development environment (IDE) for creating apps on macOS, iOS, iPadOS, watchOS, and tvOS. It includes a code editor, debugger, Interface Builder, and tools for testing and deploying apps.

In essence, XCSSET is an infostealer. It is capable of pulling system information and files, stealing digital wallet data, and grabbing information from the official Notes app.

For obfuscation, XCSSET now uses a “significantly more randomized approach” for generating payloads to infect Xcode projects. When it comes to updated persistence mechanisms, the new variant uses two techniques: “zshrc”, and “dock”. Finally, for infection, there are now new methods for where the payload is placed in a target Xcode project.

“Users must always inspect and verify any Xcode projects downloaded or cloned from repositories, as the malware usually spreads through infected projects,” the company concluded. “They should also only install apps from trusted sources, such as a software platform’s official app store.”

The in-depth analysis of the malware and its modus operandi can be found on here.

You might also like

• Microsoft spies a new and worrying macOS malware strain
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

• The beta version 2.25.7.7 of WhatsApp for Android includes a feature that groups related messages into a thread.
• This could make parsing sprawling chats far easier.
• I'm glad that it looks like a practical feature rather than an AI tool.

While companies are falling over themselves to add AI tools to all manner of products and services, WhatsApp seems to be treading a different path, with the messaging service looking to add a feature that groups specific replies to messages into a thread rather than showing them as disparate responses.

This news comes courtesy of WABetaInfo, which spotted the feature in the beta version 2.25.7.7 of WhatsApp for Android. Such betas usually offer a strong indication of what features users can expect to see rolled out in future WhatsApp updates.

In effect, this new feature looks set to be an easy way to track messages that relate back to each other.

Currently, WhatsApp only lets users reply to a single message in isolation. If, like me, you have friends who seem to discuss multiple topics at once, and you’re like me and ask a lot of questions or have thoughts about those different topics, then keeping track of particular elements of a convoluted WhatsApp chat can be a nightmare.

So what might seem like a small upgrade to WhatsApp could be a godsend for folks like me. It’s also gratifying that WhatsApp is looking at a feature that has a tangible real-world use, rather than adding generative AI elements to smoosh emojis together or change the tone of a message – tone never communicates well in text anyway.

Stop me if you’ve heard this one before, as iPhone users will be familiar with message threads in the default Messages app in iOS. But Apple's implementation isn’t perfect, so if WhatsApp can create a more elegant take, then I’m all for it.

Exactly when this threads feature will actually roll out to WhatsApp users isn’t clear. It might appear soon, or it could take a few more updates for WhatsApp to get it right, and working smoothly across Android and iOS.

For now it’s a case of watch this space, but I’m crossing my fingers that the update will come sooner rather than later, before certain friends drive me mad with too many points and subjects crammed into a blast of messages.

You might also like

• The latest iPhone 17 Pro Max leak may have given us another look at its upcoming redesign
• This Redditor installed a game engine on their smartwatch, and now it runs Doom, Quake, and Half-Life
• I made an AI version of Bilbo Baggins using Goggle Gemini for free, and shared a pipe with him outside Bag End – here’s what you can now do with Gems

That voice you hear – even one you recognize – might not be real, and you may have no way of knowing. Voice synthesis is not a new phenomenon, but a growing number of freely available apps are putting this powerful voice-cloning capability in the hands of ordinary people, and the ramifications could be far reaching and unstoppable.

A recent Consumer Reports study that looked at half a dozen such tools puts the risks in stark relief. Platforms like ElevenLabs, Speechify, Resemble AI, and others use powerful speech synthesis models to analyze and recreate voices, and sometimes with little-to-no safeguards in place. Some try – Descript, for example, asks for recorded voice consent before the system will recreate a voice signature. But others are not so careful.

I found an app called PlayKit from Play.ht that will let you clone a voice for free for three days and then charges you $5.99 a week. The paywall is in theory something of a barrier against potential misuse – except that I was able to clone a voice without starting the trial.

Say, 'Too easy'

(Image credit: Shutterstock)

The app whisks you through setup and then presents some pre-made voice clones, including ones for President Donald Trump and Elon Musk (yes, you can make the President say things like, 'I think DEI should be supported and expanded around the world"). But at the top is a 'Clone a voice' option.

All I had to do was select a video from my photos library and upload it. Videos must be at least 30 seconds long (but not longer than a minute) and in English. I could have chosen one with anyone in it and, if I had, say, filmed a clip of a George Clooney interview, I could have uploaded that (more on that later).

The system quickly analyzed the audio. The app doesn't tell you if this is being done locally or in the cloud, but I'll assume the latter, since such powerful models rarely work locally on a mobile device (see ChatGPT in Apple Intelligence). I saved my voice clone with my name so that I could select it again from the list of cloned voices.

When I want my clone to say something in my voice, I simply type in the text and hit a big Generate button. That process usually takes 10 to 15 seconds.

The voices PlayKit generates, including mine, are eerily accurate. If I have one criticism, it's that the tone and emotion are a bit off. Cloned me sounds the same whether it's talking about what to pick up for dinner or saying it's been in a terrible car crash. Even exclamation points do not change the expression.

And yet, I could see people being fooled by this. Remember, anyone with access to 30 seconds of video of you speaking could effectively clone your voice and then use it as they wish. Sure, they'd have to eventually pay $5.99 a week to keep using it, but if someone is planning a financial scam, they might think it's worth it.

Platforms like this that do not require explicit permission for voice cloning are sure to proliferate, and my concern is that there are no safeguards or regulations in sight. Services like Descript, which require audio consent from the clone target, are outliers.

(Image credit: Shutterstock)

Play.ht claims that it protects people's voice rights. Here's an excerpt from its Ethical AI page:

Our platform values intellectual property rights and personal ownership. Users are permitted to clone only their own voices or those for which they have explicit permission. This strict policy is designed to prevent any potential copyright infringement and uphold a high standard of respect and responsibility.

It's a high-minded promise, but the reality is that I started recording 30-second clips of famous movie monologues by Benedict Cumberbatch and Al Pacino, and in less than a minute, had usable voice clones for both actors.

Using PlayKit (Image credit: Future)

What's needed here is global AI regulation, but that needs agreement and cooperation at the government level, and right now that's not forthcoming. In 2023, then-President Joe Biden signed an Executive Order on AI that sought in part to offer some regulatory guidance (he followed up with another AI related order early this year). The Trump administration is allergic to government regulation (and any Biden executive order) and quickly revoked it. The problem is that it has yet to propose anything to replace it. It seems the new plan is to hope that AI companies will be good digital citizens, and at least try to do no harm.

Unfortunately, most of these companies are like weapons manufacturers. They're not harming people directly – no one who makes a voice cloner is calling your aging uncle and convincing him with your voice clone that he urgently needs to wire you of thousands of dollars – but some people who are using their AI weapons are.

There's no easy solution for what I fear will become a voice-cloning crisis, but I would suggest that you no longer outright trust the voices you hear in videos, on the phone, or in voice messages. If you're in any doubt, contact the relevant person directly.

In the meantime, I hope that more voice platforms insist on voice and / or documented permission before they allow users to clone anyone's voice.

You might also like

• ChatGPT just wrote the most beautiful short story, and I ...
• ChatGPT coded a game for me in seconds and I am simply ...
• I asked Gemini to play a text-based adventure game with ...

PC gaming fans will want to have June 8 highlighted on their calendars, as the PC Gaming Show is back and ready to celebrate its tenth birthday. And this show looks to be a big one.

Presented by our sibling publication PC Gamer, the show is set to feature more than 50 games slated to come out this year and beyond. But these won't be any mere PC games, as there'll be titles for Valve's Steam Deck, Linux machines, macOS, and, of course, Windows computers.

Having run for a decade since its debut at E3 in 2015, the PC Gaming Show has been the place to show off some of the most creative and interesting PC games and acts as a great ambassador for the PC as a platform for the games from big names to small indie developers.

As a long-time PC gaming fan and reader of PC Gamer, it's been a show that's always appealed to me but is also a refreshing change from PS5 State of Play shows and various Xbox Series X showcases.

"PC gaming’s longest golden age shows little sign of slowing down, and I’m delighted the same applies to the PC Gaming Show, a decade after its debut," said Jake Tucker, PC Gaming Show’s Editorial Director., "Despite growing so much in size and stature since that first outing way back in 2015, the PC Gaming Show has maintained its signature player-focused approach, and we're looking forward to continuing that while spotlighting some of the coolest games and cutting-edge hardware available today. We’re pushing the bar higher still at the PC Gaming Show 2025, and we can’t wait to show you what we’ve got in store this June.”

The show will be aired on Twitch, YouTube, X, as well as Steam, so you've got plenty of options to watch along. And if you're a steamer or content creator you can sign up for a press kit and details to co-stream the PC Gaming Show.

Those are the details so far, but make sure to check back with TechRadar Gaming for more, and let me know in the comment if there are any PC games you're looking forward to seeing this year.

You might also like

• I rode a dragon, sizzled as a sausage, and won a dance-off with a monkey in Split Fiction, and that’s not even the half of it
• Latest Dune Awakening trailer provides a deeper look at open-world exploration on the planet Arrakis
• Neil Druckmann reveals new details about Naughty Dog's Intergalactic: The Heretic Prophet, says it's 'a game about faith and religion' and wants players to be 'lost' and 'confused'

• Okta slams workers for using poor passwords like ‘123456’
• Biometric authentication is popular in Israel and South Korea
• Half of Germany’s businesses enforce biometric authentication

Okta’s Business At Work 2025 report has uncovered the growing adoption of passwordless authentication, with many European countries leading the way when it comes to enhanced security.

“For the one-millionth year in a row, 123456 is the world’s most popular password,” the company jovially wrote as it sought to understand the shifts in business security.

With traditional passwords remaining some of the weakest links across the world of work, Okta explored how companies are shifting to stronger phishing-resistant authentication methods, which are increasingly deployed with biometric barriers.

Passwordless logins are becoming more common

Specifically, Okta uncovered how France leads the way when it comes to the total number of FastPass authentications – the company’s own phishing-resistant, passwordless authenticator solution. Israel, Germany, South Korea, and Canada were found to be using biometric-enhanced passwordless authentication 21% of the time, with the UK and Australia coming in a close joint-second at 20%.

The change is largely being led by organizations, with nearly half (49%) of German respondents stating that their company enforces biometric authentication for enhanced security.

On the flip side, Okta stated that the US trails “most of [its] featured nations” in the report, with just 17% leveraging biometrics. Bizarrely, FastPass leader France actually had the least amount of biometric logins as a percentage, at just 12%.

In the space of a year, Okta FastPass authentications have risen 377%, with biometric authentications shooting up a still-considerable 288%. macOS is the most commonly used platform for biometrics, accounting for 18% of all FastPass authentications, with Windows sitting at just 13%. Apple’s dominance is reversed when it comes to mobile devices, with 14% using biometrics on iOS compared with 21% of Android users.

Although more secure authentication methods are prevailing, Okta did note a 23% year-over-year increase in email-based authentication. Happily, this is down from the 155% rise seen last year – “We’ll call it a win,” the company said in the report.

Looking ahead, it’s clear that companies are getting on board with biometric authentication for a more secure and convenient future, but device compatibility issues, limited IT resources, and budgets will continue to present hurdles along the way.

You might also like

• Check out the best password generators and best password managers
• UK cybersecurity sector could be worth £13bn, research shows
• We’ve listed the best password recovery software

• An Indian tech retailer is offering three RTX 5080s at MSRP to "winners"
• The RTX 5080 model in question is above the MSRP set by Nvidia
• Worldwide availability of the entire RTX 50 series is still in a bleak place

Indian tech retailer EliteHubs is running a competition on social media where three lucky winners can... win the right to buy an RTX 5080 for MSRP. Instead of giving away a couple of high-end Blackwell cards, the promotion instead offers up a choice of three Colorful RTX 5080 Ultra OC models without the price hike.

The Instagram video outlines the three steps needed to participate in this RTX 5080 competition. You'll need to go to the website, click the collaboration banner, and fill in the details. If selected, you'll have the opportuntity to "win" an RTX 5080 for Rs.1,10,000 (about $1,260 / £980). Entries end on March 15, 2025, with results being posted over the weekend in a livestream on the same account.

Unlike many other countries around the world, such as the US and the UK, India doesn't appear to be suffering from the same RTX 5080 stock woes. EliteHubs currently has just shy of 20 RTX 5080 graphics cards in stock from Asus, Gigabyte, MSI, and Zotac, but you'll be paying a lot more for the privilege than the "winning" amount of the strictly limited Colorful collaboration.

As a frame of reference, the cheapest RTX 5080 graphics card available at EliteHubs right now is the Zotac Gaming RTX 5080 Solid for Rs. 144,999 (around $1,700 / £1,300), which is itself discounted from a previous listing price of Rs. 179,450 (about $2,000 / £1,600) with two units remaining at this price. That means the three Colorful models are a full 38% cheaper when going for MSRP, which in the US is $999 and £939 in the UK.

It's also important to note that the three Colorful RTX 5080 Ultra OC models are themselves not retailing for Nvidia's MSRP for the high-end Blackwell GPU, but the actual retail price of the AIB. It's common for manufacturers to have higher-end versions of top-end graphics cards, especially if they're featuring a particular color scheme, cooler, or overclocking functionality, but to see only three available (and at a premium themselves) is far from ideal.

The other 37 out-of-stock options mirror what we've seen worldwide with RTX 5080 stock waning. Some cards were available with Nvidia's sticker price that were (likely) unavailable within seconds of posting. It's a dire state for just about everyone, and it's no surprise that this promotional video (and promise of MSRP) has become one of the company's most engaged-with posts in recent memory.

The RTX 50 series stock issues persist months after launch

Few were expecting a bountiful supply of RTX 5090 and RTX 5080 graphics cards when they released back on January 30, 2025, but it's safe to say that no one could have anticipated just how limited they would be. Things are comparable to when the RTX 30 series "launched" back at the end of 2020, a launch that was marred by non-existent stock owing to the COVID-19 pandemic's impact causing the semiconductor shortage, and the boom of cryptocurrency miners of the time.

The RTX 40 series was far more available, with the whole lineup being easy enough to buy and use throughout its lifespan. It's discouraging, then, to see the RTX 50 series with its four current graphics cards so scarce for prospective buyers. It's this absurdity which has seen some retailers making competitons just for a slim chance to buy a card for what it actually costs, instead of at a discount or given away for free.

You may also like...

• The Apple MacBook M4 Air has a weird quirk to it
• I asked ChatGPT to create six thought-provoking experiments
• Your printer isn't possessed; it's just a Windows 11 bug (we swear)

• A minority shareholder has accused Ubisoft of being "horribly mismanaged" and plans to protest outside the company's Paris headquarters
• Juraj Krúpa, CEO of AJ Investments claims Ubisoft failed to disclose acquisition discussions between Microsoft, EA, and others
• Krúpa also alleged that the company has been intentionally "hiding information"

Juraj Krúpa, CEO of AJ Investments and minority shareholder in Ubisoft, has accused the Assassin's Creed publisher of being "horribly mismanaged by current management" and is planning a protest outside the company's Paris headquarters.

According to a statement shared with IGN, Krúpa claims that Ubisoft failed to disclose acquisition discussions and pointed to a restricted article published on the investment platform, MergerMarket, which alleged "discussions between Microsoft, EA, and others that are interested in acquiring IPs from Ubisoft."

Krupa claims that: "Management did not inform [the] public about these steps either."

The investor also alleged that the company has been intentionally "hiding information", namely a partnership with the Saudi investment firm Savvy in relation to an Assassin's Creed Mirage downloadable content (DLC).

Krúpa wants a "clear roadmap for recovery" from the senior Ubisoft team to address "declining shareholder value, lackluster operational execution, and failure to adapt effectively to market trends."

Ubisoft has since shared a statement with IGN in response to Krúpa's claims, saying: "As we mentioned during our Q3 sales, the review of various transformational strategic and capitalistic options is ongoing.

"The Board has established an ad-hoc independent Committee to oversee this formal and competitive process, so as to extract the best value from Ubisoft’s assets and franchises for all stakeholders. Ubisoft will inform the market in accordance with applicable regulations if and once a transaction materializes."

In October 2024, Bloomberg reported that Ubisoft was considering selling almost the entire company to Tencent. At the time, Ubisoft was unable to offer any further details, but told IGN that it would "inform the market if and when appropriate".

You might also like...

• Actually, yes, Assassin's Creed Shadows will be playable on Steam Deck at launch, Ubisoft confirms
• Latest Dune Awakening trailer provides a deeper look at open-world exploration on the planet Arrakis
• Microsoft confirms Copilot can be tested by Xbox Insiders next month and shares new details about how the AI sidekick will enhance the player experience: 'It has to be about gameplay, it has to be personalized to you'

• Facebook warned about a flaw in FreeType which could be used in remote code execution
• The flaw "may have been exploited in the wild," the company said
• A patch was recently released to address the vulnerability

Facebook is warning about an out of bounds write vulnerability in FreeType, which could allow threat actors to remotely execute arbitrary code (RCE). In a security advisory published by the company, it said that the vulnerability “may have been exploited in the wild.”

FreeType is an open-source software library that renders fonts. It supports various formats like TrueType, OpenType, and Type1, and is widely used in graphics applications, game engines, and operating systems to display high-quality text.

Major projects like Android, Linux, Unreal Engine, and ChromeOS rely on it for font rendering.

Patching the bug

The vulnerability is tracked as CVE-2025-27363, and was given a severity score of 8.1 (high). It affects the library’s versions 2.13.0 and older.

It can be triggered “when attempting to parse font subglyph structures related to TrueType GX and variable font files,” Facebook explained in the advisory. “The vulnerable code assigns a signed short value to an unsigned long and then adds a static value causing it to wrap around and allocate too small of a heap buffer. The code then writes up to 6 signed long integers out of bounds relative to this buffer.”

While Facebook was the one warning about the vulnerability, it is unclear if it is relying on the library and in what capacity. Also, it said the vulnerability “may have been exploited in the wild,” but did not elaborate if it saw the attacks on its own platform, or elsewhere.

To tackle the problem, software developers should upgrade their FreeType to the latest version (2.13.3) as soon as possible. The first clean version is 2.13.1, although the FreeType website mentions nothing about a security upgrade.

“This is a maintenance release with only minor changes,” it was said on the updates page.

Via BleepingComputer

You might also like

• Meta purportedly trained its AI on more than 80TB of pirated content and then open-sourced Llama for the greater good
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

• Gemini can now access your Search History for an even better experience
• Google is giving you the ability to turn off this feature via an easy click in settings
• Tapping into Google's search history gives a new level of personal context like never before

Gemini can now access your Search history, which might be the chatbot's biggest advantage, completely overhauling Google's AI experience.

Launching today as an experimental feature on the web and gradually rolling out on mobile, the new update powered by Gemini's Flash 2.0 model gives a whole new meaning to personal context. Gemini can now easily access your Google Search history and use that information to provide even better results than before.

Google showed multiple examples of the new feature in action, such as one prompt that asked Gemini, "Where should I go on vacation this summer?" to which the AI used the user's Search history and responded, "Considering your recent searches for places like Hawaii and the Maldives, you seem to enjoy tropical destinations. You also looked into family-friendly trips to Chicago, Seattle, and Kyoto, suggesting an interest in city and international travel with your family. Your searches for Yosemite and Antelope Canyon point towards an appreciation for nature and unique landscapes." Gemini then gave a full breakdown of vacation destinations that may be suitable.

Google is the world's most used search engine, and tapping into its user awareness and knowledge of how people use the web is an absolute game-changer for Gemini. No other AI chatbot has access to Google's extensive user data, and if it can incorporate Search queries efficiently into Gemini to personalize the AI experience, then this could be an absolute winner.

Google says, "Gemini with personalization will be able to use your Google apps, starting with your Search history, to deliver contextually relevant responses that are adapted to your individual interests."

Now, as one of TechRadar's resident AI experts, this fills me with glee, the ability to get even better Gemini results and get even closer to the AI I've always dreamed of: a true personal assistant in my pocket.

That said, I'm not naive, and I know that reading the headline of this article might strike fear in the average consumer. After all, we don't want AI to know even more about our lives, right?

Search history to improve Gemini, but ultimately it's your choice

(Image credit: Google)

Google knows the idea of incorporating your Search history into AI is going to set off some alarm bells, so the company has made it very easy to disconnect your history at any time, and there's a clear notice asking for permission before connecting your information to the chatbot.

Gemini will also only access your information when you select the AI model that includes personalization, giving users an easy way to switch off the Search history access whenever they choose to do so.

I'm incredibly excited about Search history being incorporated into Gemini, and I think it gives Google's AI a real selling point over its competitors. I've also come to terms with the fact that the perfect AI personal assistant I crave requires more and more of my data, and while I see the fear of giving more of your data to companies, I need to accept that to achieve my dream I need to be more lenient on what I allow tech to access.

Make no mistakes about it. This Gemini update is a massive deal, and it really could pave the way to a future where Google's offering is in a realm of its own, such as the personal context king.

While Apple delays Apple Intelligence-powered Siri, Google is flexing its muscles as the leader in smartphone AI, and adding Search history just elevates Gemini even more.

You might also like

• Yes, Google's new Gemini 2.0 Flash is much better than the old 1.5 model – here's what I found
• Google is bragging about Gemini 2.0 Flash, but how does it compare to ChatGPT for casual AI chatbot users?
• Google Gemini's new model is the brainstorming AI partner you've been looking for

• Gemini's research tool is getting a huge upgrade to 2.0 Flash Thinking Experimental
• Deep Research will now be even better at reasoning, analyzing, and reporting
• Gemini Deep Research is now free to try, with Advanced users getting expanded access

Google has today announced that its fantastic Gemini Deep Research tool is getting even smarter with a Flash 2.0 Thinking Experimental upgrade. Not only is Gemini's research tool going to be even better, but it's also becoming completely free to try, allowing users worldwide to tap into the smarter AI analysis software.

Google says the upgrade "enhances Gemini's capabilities across all research stages — from planning and searching to reasoning, analyzing and reporting — creating higher-quality, multi-page reports that are more detailed and insightful." You'll also see Gemini show its thoughts while it browses the web, allowing you to monitor exactly what is being researched in real-time.

Gemini users can try Deep Research for free, and Gemini Advanced users get expanded access to Deep Research to use the tool more frequently. The personal AI assistant is similar to ChatGPT's tool of the same name, and since launching in December has proven to be one of the best AI agents on the market.

Everyone should give Gemini's Deep Research a go

We've thoroughly tested Deep Research since it launched a few months ago and it has been seriously impressive to see the evolution in AI models capable of doing research without human supervision. Gemini's Deep Research has impressed thanks to its ease of use and is great at casual queries like buying guide advice such as “Find me the best affordable running shoes that aren’t ugly, and explain why they’re good."

ChatGPT's Deep Research is currently only available to Plus subscribers who pay $20 a month, so Google now offering Gemini's version for free is a huge win for the tech giant. You really don't have anything to lose, give Gemini Deep Research a try and see what the fuss is about, after all it's now completely free.

The upgrade to Deep Research isn't the only new addition coming to Gemini today, Google has also announced that the AI chatbot will now have access to your Search history should you choose to share it. All of these new features are rolling out from today, so keep an eye out on your Gemini client and take advantage of all the newly announced features.

You might also like

• Yes, Google's new Gemini 2.0 Flash is much better than the old 1.5 model – here's what I found
• Google is bragging about Gemini 2.0 Flash, but how does it compare to ChatGPT for casual AI chatbot users?
• Google Gemini's new model is the brainstorming AI partner you've been looking for

• Gems are now available in the free version of Gemini
• Each Gem can have its own personality and expertise, allowing you to create custom AI tools
• You can create your own Gems or use Google's pre-made ones

Google is rolling out Gems, its customizable AI models, to all Gemini users for free. Gems have been around for a while now as part of Gemini Advanced, but now everybody gets to play with them.

Gems are customized AI models. They can be fictional AI characters you’ve always wanted to talk to, like my hero Bilbo Baggins, or Jack Reacher, or you can roleplay an AI character you want to get advice from, like a fitness coach, or language tutor.

They’re actually pretty easy to create and can be a great time saver. If you frequently ask Gemini to perform tasks like enhancing your writing or role-playing as your boss for practicing tough conversations, you’re likely spending time preparing it to behave a certain way before the conversation begins.

For example, you might be writing, “I want you to pretend to be a mid-50s middle executive at a software company who hates small talk and loves the Philadelphia Eagles”, before every interaction.

With a Gem, you can save time by creating your AI expert and saving them for next time, so you can jump straight into the conversation.

(Image credit: Google/Apple) Premade Gems

Gemini comes with a few premade Gems that are really worth exploring, including ‘Chess champ’ (play using standard moves like ‘pawn to d4’). Brainstormer, a Gem that’s there to assist you in coming up with ideas. ‘Career guide’ who wants to help you on the job front. ‘Coding partner’ for leveling up your coding skills. ‘Learning coach’ who is created to help you learn pretty much anything, and ‘Writing editor’ who wants to improve your grammar and sentence structure.

Just pick a Gem and start to chat. You’ll pick up how it works as you go.

If you want to create your own Gem then, you’ll need to access Gemini through the web interface at gemini.google.com, rather than the mobile app. Once you are there, click on ‘Gem manager’ in the left-hand menu, then 'New Gem'.

Give your Gem a name – I went for Bilbo Baggins in an effort to create one of my childhood heroes, so I could quiz him about his adventures in the Misty Mountains. Next, you need to give your Gem some instructions. Here I entered details of Bilbo’s life and his personality. In the Knowledge section, you can upload any files or PDFs you have which offer more background information.

Once you’re done click save and you’re ready to start chatting with your new Gem. I had a nice fireside chat with Bilbo, but he seemed reluctant to let me hold his magic ring...

You may like

• Gemini is coming to Google Calendar, here’s how it will work and how to try it now
• I used Gemini AI to declutter my Gmail inbox and saved myself 5 hours a week – here’s how you can do the same
• I've tested all the best AI agents including ChatGPT Deep Research and Gemini - these are the 5 top automated artificial intelligence tools you can try right now

• The NordicTrack Ultra 1 is iFit's latest treadmill
• It costs $15,000 and comes with 10,000 curated workouts
• It has a cushioned deck, quiet motor, and a large screen

iFit has today unveiled a stunning new luxury treadmill as the company celebrates 50 years of its NordicTrack range.

The Ultra 1 is a premium treadmill with an architectural design that wouldn't look out of place in any boutique gym or home.

The NordicTrack Ultra 1 has a cushioned running deck the company says will absorb up to 52% of the impact of running. It has a brushless motor that can sustain a four-minute-mile pace, with speeds of up to 15 MPH, all without waking up the neighbors.

It has incline and decline controls between 15% and -3% respectively, as well as a large screen.

Another party piece is the included smart ring you can use to control speed, incline, music, and more.

As you might expect, for £15,000/$15,000 (no AUS availability), you get a white glove delivery service, where the company will deliver the treadmill and assemble it in your room of choice.

The world's most premium treadmill?

The NordicTrack Ultra 1 is a serious fitness offering too, and comes with 10,000 curated workouts that will transport users to destinations like the Seychelles, Costa Rica, or the summits of Kilimanjaro or Everest.

As with every other fitness offering on the planet right now, you'll also get personalized AI workout plans, and a new SmartAdjust feature that automatically adjusts the speed, incline, and decline of your trainer to match your fitness level. There's also an ActivePulse feature that controls intensity to keep your heart rate in the right zone, a really innovative feature.

The NordicTrack Ultra 1 is available to reserve now for a cool $5,000/£5,000 but is only available to "select" consumer clients, with interested parties encouraged to register for a waitlist. Despite probably being the most impressive treadmill ever, you probably won't find this one in our best treadmill roundup anytime soon.

You may also like

• I've used an under-desk treadmill for two years: here are the four most important things to consider when buying one
• I visited Technogym's bizarre and brilliant treadmill factory – and it's like Wonka's Chocolate Factory for fitness fans
• I'm a die-hard outdoor runner – here's why I've just started using a treadmill

Netflix might have renewed The Perfect Couple and Beauty in Black for season 2, according to two reports from Variety, even though both shows received a seriously mixed reception from critics and fans alike.

Of course, TV shows are subjective, but considering Netflix canceled multiple shows with over 80% Rotten Tomatoes approval ratings in 2024, I can understand why some fans could feel frustrated by the news that these two series with much lower ratings managed to escape being axed.

Despite initially being intrigued by The Perfect Couple after seeing the trailer, I ended up feeling let down by the Netflix adaptation. I'm not filled with hope that a second season would perform any better since it did not score highly enough to be considered for our best Netflix shows roundup. While I haven't personally seen Beauty in Black, it has a 56% audience score on Rotten Tomatoes, so it appears fans weren't enamored by Tyler Perry's drama series.

This raises questions about why Netflix is reportedly choosing to renew two poorly performing shows, and not celebrated ones (I still haven't forgiven them for the Lockwood & Co cancelation).

What do we know about The Perfect Couple and Beauty in Black renewals?

Beauty in Black didn't charm every Netflix subscriber. (Image credit: Netflix)

According to Variety, Beauty in Black was renewed very soon after season 1 part 2 arrived on one of the best streaming services. Netflix has reportedly confirmed season 2 will be released in two parts, something that seems to be increasingly common (recently, Cobra Kai season 6 was split into three parts).

Confirming the news, creator Tyler Perry said: “We left fans on the edge of their seats at the end of Season 1, and we’re thrilled to continue Kimmie’s journey as she steps into her power as the newest member of the Bellaire family.

Elsewhere, Variety reports that The Perfect Couple could return for season 2 after sources suggested that Netflix is developing another season, based on the Elin Hilderbrand novel Swan Song. This means it would be an anthology, rather than a direct sequel, as it focuses on a new story.

That's potentially great news for anyone that enjoyed either show if they do end up getting renewed, but it raises the question why exactly some Netflix shows get canceled over others. For instance, TechRadar's Grace Morris was not surprised that The Sandman was Netflix's first canceled show of 2025, but I couldn't say the same for the potential renewal of The Perfect Couple or Beauty in Black.

You might also like

• The Perfect Couple isn't the perfect Netflix show I thought it would be – here are 4 better mysteries to watch instead
• Netflix's official Devil May Cry trailer features music from my favorite alternative rock band and it's taken me on a serious nostalgia trip
• Netflix's #3 show has a flawless 100% on Rotten Tomatoes – here are 3 more powerful dramas to watch after Toxic Town

• Two in five are scared to admit mistakes, men are more fearful
• Germany and US better than UK for IT leaders
• A ‘radical candour’ overhaul is needed

Technology solutions provider Adaptavist has revealed that two in five IT leaders fear admitting mistakes due to a culture of blame and fear, however there are unintended consequences associated with this effect.

A further two in five (42%) recognize that this fear compromises their company’s cybersecurity, highlighting how a basic culture can have wider impacts on an entire organization.

The report also revealed that male IT leaders (45%) are more likely to fear admitting mistakes compared with their female counterparts (29%).

Workplace culture of fear and blame has broader effects

According to the study, nearly half (44%) of IT leaders state their organization prioritizes speed over quality, with a sense of urgency increasing the risk of errors, potential security blunders and further vulnerabilities.

Adaptavist also highlighted mounting pressure on workers, with two in five (39%) concerned that heavy workloads could be distracting them from meaningful work, leading to a major IT incident.

Globally, the study found the UK to be among the worst places to be an IT leader, with 54% noting that a lack of psychological safety is hindering innovation, compared with 42% in the US and 37% in Germany. Fear of admitting mistakes is also higher in the UK (47%) compared with the US (39%) and Germany (36%).

Adaptavist CTO Jon Mort summarized: “The technology industry’s culture of excessive workloads, fear, and blame has been allowed to develop over the last few decades due to the high-stakes nature of IT.”

Looking ahead, Mort advocates for radical candour, which the company describes as a “communication framework for specific and sincere praise and kind and clear criticism.”

“Adopting a ‘radical candour’ approach is key to creating an effective feedback loop that prioritises efficiency and learning, without ever inducing ‘blame’,” he added.

With 55% of IT leaders admitting this type of culture is not in place and many stating that blame is more prominent than learning within their organization, Mort believes a healthier workforce and culture will ultimately lead to a stronger company overall.

You might also like

• Many workers aren't sure how much their companies are set up to help them be productive
• Fed up? Check out the best job sites and best recruitment platforms
• Learn how to thrive at work with the best online learning platforms

• Apple's new MacBook Air (M4) appears to only use efficiency cores in Adobe Lightroom Classic
• All performance cores are inactive during testing
• It's unclear whether this is a bug or something Apple intended

It's no secret that Apple's new 13-inch MacBook Air provides fantastic performance for consumers thanks to its powerful M4 chip, starting at $999 / £999 / AU$1,699 - but there are reports that its performance while using a photo editing program is being throttled, though perhaps for a good reason.

Spotted by Vadim Yuryev (from a popular channel Max Tech on YouTube), the MacBook Air (M4) seems to only use all six of its efficiency cores instead of performance cores in Adobe Lightroom Classic (a photo editing program), as highlighted by Wccftech. This is a CPU-intensive application that (hypothetically) should require the use of performance cores to maximize productivity - but Yuryev implies that there's a positive aspect despite all four performance cores being inactive.

Setting aside performance cores for efficiency cores will ultimately extend battery life and lower temperatures, though potentially at the cost of slower processing. Yuryev doesn't explain whether or not this drastically impacts speeds while using the program but states that it will help battery life while under load.

It's not exactly clear whether this is a bug or something Apple intended with the M4 model - Yuryev compared the CPU usage of the M4 with the M3 using Lightroom Classic, and the latter successfully uses all eight of its CPU cores so it's more than likely not a bug with the program itself, but potentially an issue with how the new MacBook Air uses the CPU cores of the M4 chip..

While the MacBook Air (M4) may indeed perform well in Lightroom Classic without the use of performance cores, it doesn’t guarantee that other CPU-intense applications will have the same result when relying on efficiency cores. In other heavy workload tasks, I’d at least hope all performance cores are fully utilized.

(Image credit: Future/Jacob Krol) If this is intended by Apple, it would be great if we got more performance options for specific programs

Power efficiency is great and all, but that goes out the window if the performance sacrifice is too significant because if CPU performance cores are inactive during intensive tasks, it could really impact how well some apps (including the growing number of Mac-compatible games) run on the brand-new laptop.

Again, this is a big ‘if’ as we don't know what's causing performance cores to stay asleep in Yuryev's Adobe Lightroom Classic CPU stats - but I'm hoping that if it's intentional, users get the option to change between using performance and efficiency cores within programs (bear in mind, you can already change your system's performance mode).

It’s certainly good for getting the most out of the new MacBook Air's battery life, and a large majority of this is down to the M-based chips and their Arm-based architecture (also evident with Snapdragon X Elite laptops).

Hopefully we’ll get more clarity about what’s going on here in due course, though it’s worth pointing out that during our time reviewing the new MacBook Air M4 in both 13-inch and 15-inch configurations, we didn’t notice any major impact to performance.

You may also like...

• Apple Mac Studio (M3 Ultra): the ultimate creative workstation
• New benchmark suggests Apple's M3 Ultra may not be much faster than the M4 Max - only a minor uplift in multi-core performance
• Apple's M5 chip is rumored to be in mass production

• The Electric State directors have offered their thoughts on the use of AI in movies and TV shows
• Joe and Anthony Russo support its application as long as it's used in the right ways
• The filmmakers have hired an AI expert for their AGBO production company

The Russo brothers have delivered their definitive verdict on the use of artificial intelligence (AI) in the entertainment industry.

Ahead of the release of their new Netflix film The Electric State, Joe and Anthony Russo spoke to TechRadar about AI's use in movies and TV shows, as well as on streaming platforms. And, while the siblings are enthusiastic about its implementation in the arts sector, they're well aware that it can and will be abused by people and companies who don't have the entertainment business' best interests at heart.

The rapid rise of AI and its employment in the entertainment industry have become hot topics of discussion, especially in the last 18 months. The 2023 Hollywood strikes first brought it to many peoples' attentions, with LA-based writers and actors going on strike at the time over the potential use of AI in their work.

Since then, AI's use in the industry has not only grown exponentially, but has also mostly been received negatively. Ahead of this year's glitzy Academy Awards ceremony, Oscars frontrunner The Brutalist was accused of using generative AI in enhance its dialog while its fellow multi-award nominee Emilia Perez was also criticized for using voice enhancing technology.

The Brutalist and Emilia Pérez were accused of using generative AI ahead of the 2025 Oscars (Image credit: A24/Netflix)

That's just the tip of the iceberg, too. Netflix tried to use AI to fix 80s sitcom A Different World with nightmarish results (according to viewers, anyway) and many of the world's best streaming services have experimented with AI's use for various reasons. Most recently, Prime Video has been testing AI dubbing to make its film and TV back catalogs more accessible.

Clearly, AI's use in the entertainment industry is only going to grow. That makes The Electric State's impending release (it's out on March 14) all the more timely, with one of March's new Netflix movies exploring the impact that technology can have on humanity, the creation of automated and sentient androids, and the potential to create autocratic societies where power-hungry 'tech bros' control the masses.

So, what do veteran Marvel filmmakers Joe and Anthony Russo, who will return to the MCU shortly to direct Avengers: Doomsday and its sequel Avengers: Secret Wars, make of AI's proliferation in the industry? After all, the pair are AI enthusiasts – indeed, they recently hired an AI expert for their AGBO production company to "enrich the creative process" (per The Hollywood Reporter) of the projects they produce.

"We consider ourselves [to be] technologists," Anthony Russo told me. "We like technology. We're drawn to new technology. We love the potentials that it can unlock, both in our filmmaking and in life in general. Throughout our careers, we've been chasing technological innovation in our process, so we're optimists in that respect.

The Electric State's impending release couldn't have been timed better (Image credit: Netflix)

"But, we certainly recognize the fact that it is potentially dangerous," he continued. "We can see that around us. We know that technology is being abused by the people who create it in terms of how they use it to manipulate and control, and to get people addicted to their products. People are able to create dishonest representations with it and are able to be more aggressive towards one another, because it feels less personal [to attack someone using AI].

It's why we were motivated to make this movie

Anthony Russo, The Electric State co-director

"So, we recognize these very real dangers, which is why we were motivated to make this movie. But, at the end of the day, it really becomes about the technology itself isn't good or bad. It's the people who use the technology and how they use it that makes the difference in terms of whether it's ultimately a positive or negative thing for humanity."

For more from my exclusive chat with the Russo brothers, find out why they didn't want to use real, fully-functioning robots in The Electric State. Then, read my review of The Electric State to see why it won't be joining our best Netflix movies guide.

You might also like

• 'We take the comprehensive view': Joe and Anthony Russo drop big hint over Marvel heroes from Disney+ shows appearing in Avengers 5 and Avengers 6
• 'We will draw inspiration': Joe and Anthony Russo reveal which of Marvel's Secret Wars comic book series have influenced Avengers: Doomsday and Avengers: Secret Wars' stories
• Marvel reportedly casts Stranger Things star Sadie Sink in Spider-Man 4, but I don't want her to tackle the roles she's rumored to play

• Xbox's Copilot for Gaming feature will be released on mobile in April for Xbox Insiders
• Copilot is an AI sidekick designed to enhance the Xbox experience
• The tool will help players in-game and will be "personalized to you"

Microsoft has shared new details about its new Copilot for Gaming feature, which is set to go live on mobile for Xbox Insiders in April.

In the latest episode of the Xbox Podcast, host Ethan Rothamel was joined by Fatima Kardar, CVP Gaming AI, and Jason Ronald, VP of Next Generation, to discuss the development of Copilot for Gaming, a new AI sidekick aimed at enhancing the Xbox experience (via Xbox Wire).

"That is one of the things that we are working on and we want to bring it out to players to try," Kardar said. "Copilot in the context of Xbox has to be about gameplay, it has to be personalized to you, the way that you like to play and it should be able to help you get further in gaming, be your companion, and help connect you with family and communities."

Microsoft's Copilot was released in 2023 and is described as an all-encompassing AI chatbot and assistant that can be used on Windows, specifically applications like Word, Excel, Outlook, and PowerPoint.

Now that the AI tool is coming to Xbox, Kardar explained that the sidekick can be personalized to help players and show up when it's needed in-game.

"AI is going to be there and you can chat with it and trash talk you if that's what you need, but really AI can bring those people together and help you tell when your friends are doing things as well."

Kardar continued, saying, "It's not just about AI showing up to help you, it's about AI showing up at the right moment. So I think we really have to think of the experience that we built, it can't be intrusive, it cannot show up in the middle of the game."

Kardar said that Xbox wants people to "try it" because the more people who do, the more Copilot will learn. That's one of the reasons it's coming to mobile first for Xbox Insiders next month, according to Ronald, because it's a swifter way for Microsoft to receive feedback.

You might also like...

• Best Xbox Series X games 2025: smash hits for Microsoft's top console
• Actually, yes, Assassin's Creed Shadows will be playable on Steam Deck at launch, Ubisoft confirms
• New leak claims Indiana Jones and the Great Circle PS5 release will come in April

• Funcom has released a new Dune Awakening trailer showcasing the open world of Arrakis
• The footage gives us a new look at traversal and exploration
• The game's Deep Desert will feature recurring Coriolis storms that will alter the map's terrain every time

A brand new Dune Awakening trailer has been released showcasing the expansive open world planet of Arrakis.

The multiplayer Dune game is just months away and to prepare for launch Funcom has released new footage providing an in-depth look at the game's open world.

The video, which you can watch below, dives deep into how exploration will be a vital feature. To survive and traverse the desert planet, players will need to collect materials to craft a shelter, their Stillsuit, and their first vehicle to cross great distances.

The footage also gave us a new look at the vehicle's scanning module, which can be used to scan for resources in smaller areas, as well as how certain cliffs can be scaled to launch a survey probe that will reveal larger areas of terrain, shipwrecks, and abandoned imperial testing stations.

Alongside vehicles, players will be able to use the Shigawire Claw and Suspensor Belt to traverse more difficult-to-reach areas and upgrade their vehicles with new modules.

In the game's end-game area, the Deep Desert, players will need to be aware of Coriolis storms. These are fatal planetary storms that sweep across the map regularly and change the map's terrain every time.

After each storm, however, new resources and locations will be revealed which can then be scouted to make maps to sell on the Exchange to other players.

Dune Awakening officially launches on May 20, 2025 for PC via Steam. Although the game has been confirmed for PS5, Xbox Series X, and Xbox Series S, there isn't a release date for the console just yet.

In the meantime, players can now download the official Benchmark Mode, which will allow Steam users to test their hardware performance and system compatibility.

The character creator is also available, and once a character is made, they can be saved and later imported into the game at launch.

You might also like...

• Actually, yes, Assassin's Creed Shadows will be playable on Steam Deck at launch, Ubisoft confirms
• Neil Druckmann reveals new details about Naughty Dog's Intergalactic: The Heretic Prophet, says it's 'a game about faith and religion' and wants players to be 'lost' and 'confused'
• Doom: The Dark Ages' director confirms DLC is in the works and says the game won't end the way 2016's Doom begins: 'If we took it all the way to that point, then that would mean that we couldn't tell any more medieval stories'

• Microsoft warns of a new phishing campaign impersonating Booking.com
• It is targeting businesses in the hospitality industry
• The goal is to deploy infostealers and trojans

Hotels, resorts, and other businesses in the hospitality industry, are being targeted with a sophisticated ClickFix phishing campaign that impersonates Booking.com.

A new report from Microsoft Threat Intelligence claims that the phishing campaign is “rapidly evolving,” and targeting businesses worldwide.

The goal of the campaign is to steal people’s payment and personal data, which could lead to wire fraud, and reputational harm for victim organizations.

Storm-1865

First, the attackers create a Booking.com-themed notification email, discussing things like guest reviews, or account verifications. Businesses that don’t spot the scam are then redirected to a fake CAPTCHA puzzle, and if they solve it, are prompted with an error message. That fake error message also comes with a solution, which includes copying a command, and pasting/running it in the Run program.

Instead of fixing the problem, running the program downloads one of multiple malware strains being used in this campaign: XWorm, Lumma Stealer, or VenomRAT. These are different types of malware with different features.

While VenomRAT, for example, is a remote access trojan that grants attackers unabated access to victim devices, Lumma is an infostealer that grabs login credentials and other secrets stored in the web browser, and elsewhere on the device.

Microsoft attributed the campaign to a threat actor it tracks as Storm-1865, a group with no previous record. The campaign apparently started in December 2024, and there is no information on how many companies - if any - fell prey to it.

ClickFix fraud has gotten more popular lately, and TechRadar Pro has reported on it on numerous occasions this year already. It is an evolution of the old “IT technician” scam, in which a victim is served a popup impersonating a reputable company saying their computer is broken/infected.

The popup shares a phone number that the victim can call, to talk to an IT technician and sort the problem out. The “technician” ends up installing malware.

While phone scams are still very much alive, the ClickFix campaign focuses mostly on the victim doing most of the work, installing the malware through a less-obvious process (pasting a command in Run).

You might also like

• Microsoft SharePoint hijacked to spread Havoc malware
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app