Technology

• The UK's AI Safety Institute is now the AI Security Institute
• The renaming represents a shift towards cybersecurity
• Anthropic has also partnered with the Government

The new Labour Government has gone all in on AI since taking power in 2024, and was supported by the UK’s AI Safety Institute (AISI) - but not for much longer. The institution will remain, but the government has announced that this will now be renamed to UK AI Security Institute - signalling a definite shift towards cybersecurity.

But what’s the difference? Well security is focused on defending against things like cyberattacks, and mitigating security risks including how the technology can be leveraged against national security or can enable crimes such as fraud or developing chemical weapons.

Safety has a slightly wider scope - also aimed at mitigating the risks of cyberattacks, but going further to protect against misinformation spread by chatbots, and aims to assess the societal impacts of AI models, rather than just the immediate cybersecurity threats.

New partnerships

The UK’s widely publicised Plan for Change, released in January 2025, leant heavily on AI, introducing ‘Growth Zones’, handing public data over to train models, and aiming to see AI ‘mainlined into the veins’ of public services - but not once did the document mention the words ‘harm’, ‘safety’, or ‘threat’, TechCrunch noted.

Despite these omissions, the work of the AI Security Institute is still the same, says Secretary of State for Science, Innovation, and Technology, Peter Kyle;

“The work of the AI Security Institute won’t change, but this renewed focus will ensure our citizens – and those of our allies - are protected from those who would look to use AI against our institutions, democratic values, and way of life.”

As part of the new plan, the government has agreed to a new partnership with AI firm Anthropic, working to ‘realise the technology’s opportunities, with a continued focus on the responsible development and deployment of AI systems.’ This will include ‘insights’ on how AI can ‘transform public services and improve the lives of citizens’, as well as drive scientific development.

This is part of the UK’s ambition to attract tech investments from around the world - trying to foster an environment perfect for AI innovation, seemingly free of safety regulations.

You might also like

• Take a look at our pick of the best AI tools around
• Check out our recommendations for the best antivirus software
• Make AI work smarter, not harder for employees

EA's Hazelight Studios follows up its lauded hit It Takes Two with a new two-player collaboration game in its signature style.

• The PS5 just had its best financial quarter results ever
• The console shipped a staggering 9.5 million units in Q3 FY24
• Monthly active users also saw a healthy uptick in the same period

It seems that folks are still plenty eager to add a PlayStation 5 console to their gaming setups, as the system has just enjoyed its best financial quarter on record.

This is per financial results recently published by Sony for its Q3 FY24. Page 10 of the report shows that the company shifted 9.5 million PS5 consoles during this financial quarter. While that's an astronomical jump from Q1 (2.8 million) and Q2 (3.8 million) these Q3 numbers were most likely bolstered by the release of the PS5 Pro.

This brings the total number of PS5 consoles shipped to around 15.7 million for this financial year so far. Last financial year, Sony shipped a total of 20.8 million PS5 units, and with a quarter left to go for this financial year, the console could well be on track to beating that, too.

This brings the PS5's lifetime sales so far up to 75 million which, per GameSpot, is just shy of the 76.5 million PlayStation 4 consoles shipped over the same amount of time. In that case, it should be stressed that the PS4 Pro launched earlier in that generation than the PS5 Pro did in this one. PS4 family units were also cheaper than their PS5 counterparts.

Another highlight from the financial report show that there has also been a decent uptick in monthly active users. Q3 FY24 peaked at 129 million monthly active users compared to the previous year's 123 million from the same quarter. You could likely look at the success of live service titles like Call of Duty: Black Ops 6 and Helldivers 2 for getting a few million extra folks to log on each month.

Sony has two high-profile exclusive games launching this year in Ghost of Yōtei and Death Stranding 2: On the Beach, though as both currently lack a release date beyond a vague '2025' window and it's unclear whether or not they'll land in Q4 FY24.

You might also like...

• The latest Astro Bot update adds five new levels and a 'constant best resolution' at 60FPS for PS5 Pro
• Housemarque's next game is Saros, launching on PS5 and PS5 Pro in 2026
• Lies of P's Overture DLC is finally coming this year, closing the book on my favorite soulslike

If you start focusing on quality sleep now, your future self might thank you.

• Gen research claims malicious apps on major app stores are on the rise
• New banking trojans, predatory loan apps, and similar, have been spotted
• Official app stores are still a safer bet than third-party ones

App stores are becoming an increasingly major security concern, as more and more malicious apps are discovered in stores owned by major brands, experts have warned.

A new report from cybersecurity researchers Gen claims it observed new mobile bankers (phone applications designed to steal banking information) launched in the fourth quarter of 2024, often disguising themselves as Visa apps, dating apps, or Chrome browsers. Gen added that the well-known BankBot banker compromised 236% more devices compared to the previous quarter.

Besides banker apps, malicious loan apps surged as well, the researchers said. These promise quick money, but demand excessive permissions (SMS, photos, sensitive data) which they later use in extortion, threatening to publish victim data unless they pay up.

Generally safe

Gen also noted cybercriminals are adopting novel distribution tactics in app stores. Recently, they discovered a new malware strain disguised as a Body Mass Index (BMI) calculator. The app was spread through the Amazon App Store.

Major app stores like the Apple App Store, Google Play Store, and Samsung Galaxy Store are generally safe because these companies invest heavily in security measures to protect users. They use automated malware detection, manual app reviews, and strict developer policies to prevent harmful apps from reaching their customers.

Features like Google Play Protect scan apps in real-time, while Apple’s notarization process ensures apps are vetted before approval. Regular security updates, user reporting systems, and app removal policies further enhance safety.

Obviously, no platform is completely risk-free. We’ve seen malicious apps make their way into these stores in the past, and we’ll probably see them in the future, as well. However, these app stores provide a significantly safer environment than third-party or unofficial sources.

You might also like

• In defense of the Apple app store
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

• Final Fantasy Crystal Chronicles has suddenly shut down on iOS
• Square Enix says the problem lies with an unsolvable in-app purchase bug
• The specific problem has been speculated on by users

There’s been some surprising news in the world of Final Fantasy, as the game’s developer Square Enix recently shut down Final Fantasy Crystal Chronicles on iOS without a warning. The move only struck the game on iOS, with other platforms being unaffected, leaving iOS users wondering what on earth had happened.

According to Square Enix, the issue centered on in-app purchases within the game. In particular, a blog post by the company revealed that “we have been receiving reports of customers unable to access the additional content they purchased for the iOS version of Final Fantasy Crystal Chronicles.” This problem affects the full, paid version of the game too.

Unfortunately, Square Enix has taken the drastic step of shutting down the game on iOS entirely. The developer said that “Further investigation revealed that we are unable to completely fix the bug and implement the new changes, making it unlikely to resume service for the game.”

Square Enix believes the problem lies with “changes made to the in-app purchases model.” While this seems a little vague, some have speculated that the decision could be due to a recent change to Apple's in-app purchase verification. If true, this could be what has caused the disruption to Final Fantasy Crystal Chronicles – although it’s not clear why Square Enix couldn’t update the game to be compliant.

How to get a refund

(Image credit: Square Enix)

If you’ve bought in-app purchases in Final Fantasy Crystal Chronicles on iOS and have been affected by this bug, Square Enix has laid out the steps you can take to get a refund.

To be eligible, you need to have made an in-app purchase in the game in January 2024 or later. The list of affected items includes the full version unlock, the Magic Wallet, the Memory Crystal Pack, and more. The full list can be found in Square Enix’s blog post.

To get your money back, Square Enix advises you to contact Apple Support and request a refund from them. Square Enix’s support team will be available to help until August 31, 2025.

It’s unfortunate to see the game be completely pulled from the App Store. While we would hope that Square Enix is able to find a fix, judging by the company’s own words, it doesn’t sound like that’s going to happen.

You might also like

• Square Enix is ditching PlayStation exclusivity to 'aggressively pursue' a multiplatform strategy
• Final Fantasy 7 Rebirth review - fully reincarnated
• Final Fantasy 7 Remake just got a patch that changes the ending and some minor details

It's easy to fall for these top rates.

• Google Gemini can now remember previous conversations
• The memory functionality is rolling out now to Google One AI Premium subscribers
• ChatGPT has had similar functionality for a year, and it's one of the best features of any AI chatbot

Google has just added an upgraded memory feature to Gemini that allows you to ask the AI chatbot questions based on past conversations.

The new "recall" feature is rolling out to all users who subscribe to Google One AI Premium, a paid monthly subscription that grants access to Gemini's best features. With recall, you'll be able to ask Gemini about previous conversations and pick up from where you left off, allowing the AI to feel more alive and aware of your history. Previously, Gemini had no recollection of previous chats, so you'd have to remind it of important details.

This huge upgrade to Google Gemini brings the AI chatbot up-to-speed with competitor ChatGPT, which has had a well-functioning memory feature for over a year now. The difference is, ChatGPT's offering is available for free and doesn't require a monthly subscription to access its functionality. That said, Google could be testing the recall feature before rolling it out to free Gemini users, although currently, we've had no information of that happening.

This new update comes off the back of Gemini's November update that added the ability for the chatbot to remember certain things about you based on your interests and personal preferences. Unlike this new update, you'd have to go to Gemini's "Saved Info" tab and pre-fill information for the AI to reference in conversations.

When ChatGPT introduced memory last year it completely changed the way I interacted with AI, allowing me to speak naturally with the chatbot and spot nuances where it was able to reference the past in very useful ways. Until now, I've been put off from using Gemini because of its lack of memory, but that's all changed. Gemini's recall feature is rolling out in English to Gemini One AI Premium users now (although I don't have access yet), and Google says the update will be available for other languages in the coming weeks.

The context we needed

Gemini's ability to remember previous conversations gives Google's AI chatbot a whole new level of usefulness. In the past, I've been frustrated by Gemini's lack of context to my prompts when I've asked similar questions in old chats. This was never an issue if you used one single chat with Gemini, but considering the range of models from Gemini 2.0 Flash to Gemini 2.0 Flash Thinking Experimental, I can quickly rack up multiple discussions at once.

Now, Gemini will be able to take information from all of my chats and have the personal context to reference them in any way I need. Things like "Remember that time I talked to you about train travel? What was the route you told me to take?" can now be used in Gemini, and that's a huge step in making AI more conversational and more accessible.

Talking about accessibility, hopefully, Google plans to roll out this memory feature to free users, as I truly believe a memory function is one of the most important features for any AI chatbot. Until then, I'll still recommend ChatGPT to my friends and family, after all, OpenAI's model has the memory ofan elephant, and without paying for it.

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

Staggering sales drops, swastika-daubed EVs, companies culling fleet models, and fan-forum owners selling their cars—Elon Musk's alt-right antics are seriously impacting his electric car business.

• Sophos has laid off 6% of its workers to align the two businesses
• It’s also blaming an evolving cybersecurity landscape
• Both are privately owned, marking a major shift

Just over a week after it officially completed its acquisition of Secureworks in a deal worth $859 million, Sophos has laid off around 6% of its workers.

The post-merger layoffs are a result of Secureworks no longer being a public company, with some further jobs also cut to prevent certain overlaps between the two entities.

Given that Sophos is estimated to have had around 4,500-5,000 workers (per The Register), the 6% headcount reduction could equate to around 300 positions.

Sophos lays of 6% of workers

Besides the effects of the merger, a company spokesperson also revealed that an evolving cybersecurity landscape was to blame for some job losses: “In addition to aligning our business goals, changes in the cyberattack landscape are driving an urgent shift in security needs.

“With persistent increases in both targeted and opportunistic cyberattacks, organizations of any type and size are now battling both everyday cybercrime, such as identity theft, data theft and ransomware, and state sponsored attacks, which used to be more focused on specific enterprise or public sector targets.”

It’s not the first time the company has laid off workers – widespread pandemic-induced tech layoffs in 2023 saw Sophos say goodbye to 10%, or 450, or its workers in the first month (via layoffs.fyi).

The Register reports that, in this year’s layoffs, workers lost their jobs from both Sophos and Secureworks, and that they were handled “as well as is possible.” No details have been shared regarding redundancy packages.

Secureworks’ most recent full-quarter revenue stood at $288.8 million, up 4% year-over-year, with CEO Wendy Thomas revealing a 30% rise in ransomware groups compared with the year before.

Now that it’s private, we’ll have no concrete idea as to the company’s performance, much like that of Sophos, which was snapped up by private equity firm Thoma Bravo in 2020 for $3.9 billion. Cybersecurity giant McAfee was also acquired by Thoma Bravo in 2021 for $14 billion – the private equity firm’s cybersecurity portfolio is now worth $54 billion.

You might also like

• We’ve listed the best endpoint protection software and the best antivirus software
• $19 billion Vodafone and Three merger approved by UK competition watchdog
• Downloaded something dodgy? Consider the best malware removal

The recently launched UK government’s AI Opportunities Action Plan highlights the immense scope of AI to impact business, society and individuals in the coming years, and the need to move quickly to take advantage of the technology. Looking at UK organizations' cloud strategies for the next 18 months, preparing for AI adoption was cited as a priority for 88% of IT manager respondents.

For organizations that want to keep up with any technological innovations, flexibility and agility are essential. This is no different with AI. Although this technology is sometimes described in almost mystical terms, it is better to treat it ‘simply’ as a new application. And like any modern and complex application, AI requires an open, vendor-agnostic, and hybrid-agnostic IT infrastructure.

Hybrid cloud for flexible AI use

AI tools are, of course, anything but ‘simple’. They are very complex and involve large datasets and advanced algorithms for which a traditional IT infrastructure is not suitable. An on-premises or private cloud environment can quickly become costly due to the necessary hardware investments, while a single public cloud comes with flexibility challenges. An additional consideration is whether organizations want to store all their data in the public cloud.

These challenges and risks can be better managed in a hybrid environment. ‘Hybrid’ does not only mean using a private and public cloud side by side. A hybrid cloud environment means that the different clouds are aligned and integrated within one flexible infrastructure. This orchestration is the key to uniting the benefits of on-premises, private cloud, and public cloud.

A similar hybrid cloud approach offers several advantages for AI use. First, it provides much-needed flexibility: companies can choose the best environment for each AI workload, optimizing performance, cost, and security based on specific requirements. Second, it supports scalability. An open hybrid cloud platform can be more easily scaled in response to growing AI demands, allowing companies to continue innovating without infrastructure constraints. Third, it optimizes costs. By leveraging both private and public cloud resources, companies can better optimize their cloud computing costs.

An open infrastructure for trust in AI

Trust plays a special role in the use of AI. Whether due to a lack of knowledge about how AI models work or the many headlines about hallucinations and data leaks – more than any other application, you must lay a foundation of trust to move AI applications out of the experimental phase and into production.

The solution to trust issues around AI starts with a secured and reliable AI production chain, a transparent path from idea to production, with automation where possible. Automation is important to streamline the development process and prevent human errors. Additionally, transparency and explainability in AI algorithms are important, so organizations can better understand how AI-driven decisions come about. Organizations can then build upon this with certain open source tools like InstructLab, which allow developers not only to gain insights but also to directly contribute to models with skills and knowledge without requiring specific data science expertise.

Small language models (SLMs) also play a role here. They are trained on smaller data sets, typically tailored to specific industry domains, and can enable developers to get started with AI on their laptops, removing the need to risk using unsafe or uncurated data.

Keep your options open

As AI development accelerates, companies will need to pursue flexibility and openness. This means organizations must keep their options open regarding infrastructure. A consistent platform layer that supports the use of open standards and interoperable systems from multiple vendors makes it easier for companies to switch to new tools and technologies without compromising on innovation speed.

We've compiled a list of the best cloud storage.

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

Chargers funded through the program were due to be just a small share of those opening this year. The longer-term effects aren’t yet clear.

Captain America: Brave New World has landed in theaters worldwide and I bet you're already wondering when it might debut on Disney Plus. Indeed, Marvel's latest movie has just taken flight in cinemas as of today (February 14), but, if you're not planning to watch it on the biggest screen possible, you'll want to know when it could come to Disney's primary streaming service.

Right now, I can't tell you when it'll be released on one of the world's best streaming services. However, I can use some of its predecessors' Disney Plus launch dates to predict its arrival. Before you continue scrolling, though, read my Captain America: Brave New World review to see if it's worth watching, plus my Captain America: Brave New World hub and Captain America 4 cast and character guide for details on its cast, story, trailers, and more.

When do we think Captain America 4 will debut on Disney Plus?

You won't like us when we're angry, Marvel – especially if a streaming launch date isn't revealed soon (Image credit: Marvel Studios)

As expected, Captain America 4 doesn't have a confirmed Disney Plus release date yet. It's just launched in theaters globally, so it'll be a while before it's available to watch at home. That won't stop me from guessing when it'll join its Marvel Cinematic Universe (MCU) siblings on the platform, though.

By and large, Marvel movies tend to join Disney Plus' film library three to four months after their initial theatrical launch. For example, Deadpool and Wolverine was released on the service on November 12, 2024. Given it landed in cinemas on July 25 of that year, that means 110 days passed between those two dates. Other recent Marvel films, including The Marvels and Guardians of the Galaxy Vol. 3, assembled on Disney Plus within a similar timeframe.

Marvel Unlimited monthly subscription: was $9.99 per month now $4.99 at Marvel
The ultimate digital comics subscription for Marvel fans is offering new and returning customers 50% off their first month, which is much cheaper than a streaming subscription to watch all the best superhero movies. To gain access to a library of more than 30,000 comics, use the code 'SPIDEY50' at checkout. You're able to use the Marvel Unlimited app on all iOS and Android devices, including the web, too. Be quick, though, because this offer expires on Monday, March 3.View Deal

I'm confident, then, that Cap's next adventure will be added to the streamer's back catalog around 100 days after its February 14 theatrical debut. If I'm right, the superhero movie should land on the platform in late May or early June.

That said, not every MCU film has followed the same release schedule format. Doctor Strange 2 joined Disney Plus just six weeks after it arrived in cinemas. Meanwhile, Black Widow earned a day-and-date release – essentially, it launched simultaneously in theaters and on the streaming giant – in mid-2021 due to the pandemic. It's possible that Brave New World could arrive sooner or later than I've predicted. I suspect, though, that its Disney Plus release will depend on how popular it proves to be among cinemagoers (and how much money it'll make) in its first few weeks on the silver screen.

What is Captain America: Brave New World's runtime?

Me and who running to our local theater to watch Captain America 4 (Image credit: Marvel Studios)

Captain America: Brave New World is one of the shortest MCU movies ever. It clocks in at a zippy 1 hour 58 minutes. That includes the fourth Captain America film's end credits sequence and its mid- and post-credits scenes. Speaking of which...

How does Captain America: Brave New World set up future Marvel movies and TV shows?

Does Sam Wilson's first big-screen outing as Captain America set up any plot threads for Avengers 5? (Image credit: Marvel Studios)

To answer that, allow me to direct you to my Captain America: Brave New World ending explained piece. In it, you'll find out how the Marvel Phase 5 movie ends, and whether its mid- and/or post-credits stingers lay the foundations for what's to come in future MCU projects, including Avengers: Doomsday, which will arrive in theaters in May 2026.

Do I need a Disney Plus account to watch Captain America: Brave New World?

Brave New World is the 35th movie that Marvel Studios has made (Image credit: Marvel Studios)

Yes. You'll need to sign up to Disney Plus to watch Captain America 4 when it's eventually released on the service. Our Disney Plus price guide is full of information on how to sign up and which tier is best for you.

When you sign up, you'll gain access to the best Disney Plus shows and best Disney Plus movies around. Some of the best Marvel movies of all time appear on the latter list, including two of Chris Evans' three Captain America flicks. For more details on those films and the MCU as a whole, be sure to read my guide on how to watch the Marvel movies in order.

For more Marvel coverage, read my guides on forthcoming films and series including Daredevil: Born Again, The Fantastic Four: First Steps, and Spider-Man 4.

The launch of the AI Opportunities Action Plan shows the UK’s commitment to making AI work for everyone. Powering the tools we use to communicate, the platforms that help us collaborate, and the systems driving vital business decisions, AI’s ability to transform productivity and simplify operations is both significant and proven. Yet, this impact isn’t always felt across the board. For many organizations and individuals, the way AI is being implemented means it often falls short and sometimes outright misses the impact it promises to deliver.

In the push to adopt AI, businesses have often focused on adding tools rather than integrating them. The result is disconnected systems that create unnecessary friction. While 96% of C-Suite executives believe AI drives productivity, 77% of employees find it complicates workflows, only making work harder and falling short of expectations. Too often, AI is treated as an isolated feature or add-on, operating independently of the workflows it’s meant to improve.

The real opportunity lies in rethinking how AI is implemented. In 2025, “productized AI” will take on a defining role in how businesses make the most of this technology – embedding intelligence directly into workflows, systems, and platforms so it operates in alignment with how people already work. It’s not about pre-built or one-size-fits-all solutions – it’s about delivering AI that feels natural, efficient, and fully integrated into everyday processes.

“AI overload” is the new “tool sprawl”

New AI tools are emerging rapidly, offering exciting new ways to rethink how we work. From scheduling meetings and assisting in hiring decisions to analysing datasets and predicting customer behavior, they show just how deeply AI is embedding itself into the workplace. However, simply having the tools in place isn't always enough to drive real change.

We've seen this before with “tool sprawl” – when disconnected technologies pile up across organizations, often doing more to confuse than help. “AI overload” is the next phase of this, where businesses are stacking up AI tools without fully considering how they’ll work together. While more than 80% of businesses now rely on AI as a core part of their operations, only 35% have integrated tools across multiple departments.

The challenge of “AI overload” is both operational and cultural. Employees are left switching between platforms, reconciling data from multiple systems, and navigating interfaces that don’t communicate with each other. Take a marketing team juggling several AI-driven tools: one for customer segmentation, another for campaign automation, and a third for analytics. Each tool performs a specific function, but they often don’t connect, wasting time and increasing the risk of error. This fragmentation stops AI from learning and improving, as it misses the chance to connect the dots and build a complete picture of the customer journey.

As a result, productivity slips, costs creep up, and employees grow frustrated with tools that feel more like obstacles than enablers. Over time, this can break down trust – not just in the tools, but in the broader promise of AI.

What is “productized AI”?

As we know, people adopt products, not technology. To make technology accessible, it must be delivered in a format that feels intuitive and easy to use. When it comes to AI, the key is to embed this new technology into tools and systems that people are already comfortable with. This is what we mean by “productized AI” – it’s about enhancing productivity seamlessly, without forcing teams to switch platforms or learn complicated new processes.

For IT teams, this means prioritizing AI tools with greater interoperability – those designed to complement, rather than compete with, existing systems. This ensures AI’s potential is fully realized without creating additional silos or complexities.

Take project management as an example. Rather than operating as a separate tool, AI can work within current project management platforms to not only flag delays or track deadlines but actively resolve issues. By reallocating resources, automating updates, and suggesting actionable next steps, AI becomes an active participant in workflows, not just an observer.

This approach brings key advantages: contextual intelligence, where AI understands the environment and tailors its recommendations to fit the organization's processes; simplicity, by eliminating the need for yet another tool; and scalability, with AI evolving alongside workflows to stay relevant over time. This is particularly critical as businesses face increasing pressure to do more with less. With stretched teams, tight budgets, and rising expectations, “productized AI” offers a practical solution, providing the efficiency and agility organizations need to compete.

Measuring AI success in real terms

The effectiveness of AI integration and its real-world success ultimately come down to one key factor: measurable impact, not features or algorithms. Does it save time? Reduce manual effort? Make work easier and more intuitive? The conversation needs to shift to how well AI supports people in their work, not how impressive its technical capabilities seem on paper. With 85% of data, analytics, and IT leaders under pressure to quantify the ROI of generative AI, the focus is firmly on proving the measurable value of these investments.

Measuring these outcomes requires a clear, structured framework. Start by defining specific objectives: Which processes is AI optimizing? You can then identify metrics that reflect meaningful impact, such as improvements in efficiency or reductions in errors. Evaluate how AI tools interact with each other across workflows. Are they creating a joined-up experience, or are they introducing friction by operating in isolation? And finally, track these metrics over time to ensure the entire AI ecosystem remains effective and adapts to evolving needs.

The future of AI lies in how well it integrates, adapts, and enhances the way we already work. When AI feels less like a tool and more like a natural extension of the workplace, that’s when its true value will be realized.

We've compiled a list of the best free project management software.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

• Microsoft’s Copilot Voice has been upgraded with 40 new languages
• The AI has also improved its real-time responses
• Microsoft wants to encourage people to engage with Copilot in their everyday lives

Microsoft Copilot Voice has become a lot more cosmopolitan. The AI assistant has added support for 40 new languages and improved its real-time responses in a bid to make conversations feel more natural and comfortable for users.

Copilot Voice debuted in October, adding a vocal component to the AI, but with more power than the previous standard form of voice assistant. It can handle multi-turn conversations, recognize interruptions, and even adjust its tone based on emotional cues. It’s also free, which is a pretty big selling point in a world where AI subscriptions are becoming the norm. OpenAI has Advanced Voice Mode for ChatGPT, while Google’s Gemini Live offers its vocal interface.

The expanded language support is a big deal, especially for users outside of English-speaking markets. Whether you’re switching between languages or simply want an assistant who understands your native tongue better, this is a welcome change. This also points to Microsoft's strategy for making Copilot more of an international AI assistant through the Voice feature.

Speedy speech

You've got a lot *in* your hands, so let me help! Just get real-time updates with Copilot Voice pic.twitter.com/lF8B8UkQYJFebruary 13, 2025

Another key improvement is in real-time information retrieval. Voice assistants have always had a slight lag when pulling information from the web, often leaving users waiting while the AI “thinks.” With this update, Copilot Voice is now much faster and more responsive when answering questions, making interactions feel smoother and more natural. No more awkward pauses while you wait for an answer to a simple question.

The update also highlights Microsoft's efforts to enhance Copilot's place as a digital assistant, not just a glorified search engine. Copilot Voice might succeed after the failure of Cortana as Microsoft's AI voice assistant. The gap between what people expect from an AI assistant and what they actually get is closing, and voice AI tools will likely be a major facet.

You might also like...

• OpenAI’s Sora AI video generator set to launch as company announces '12 days of OpenAI
• Happy 2nd birthday, ChatGPT! Here are 5 ways you've already changed the world
• ChatGPT o1-preview can solve riddles faster than me and I kind of hate it for it

Here are the answers for The New York Times Mini Crossword for Feb. 14.

• AI features on the 2024 Kindle Scribe were only available to US customers
• Latest firmware with AI features is now rolling out to users in the UK, Canada and Australia
• These features are now also available on the 2022 Kindle Scribe

When the Amazon Kindle Scribe (2024) was released in December last year, I was rather miffed that the promised AI features weren't part of the package for users outside of the US. At the time, Amazon had no comment to make as to when the AI Summary and handwriting recognition features would be rolled out more widely, but that's now happened.

The latest firmware update (version 5.17.3) is now available to download on both the 2024 and 2022 Kindle Scribe models in the US, UK, Canada and Australia, meaning even the older model is now a far better note-taking device than before.

I've checked both the 2022 and 2024 models that I have but, at the time of publication, neither had the latest firmware available for download. However, Amazon has already listed it on its Help and Customer Service page in Australia where I am based, so it's only a matter of time before it gets pushed out to the devices itself. You can, of course, download the software package from the Amazon site in your region if you don't want to wait, but you will need to use the Send To Kindle service or email it to your Kindle to update your device.

Once installed, you should be able to see a star-like icon in your notebooks. Tapping this will give you two options: to 'Summarize' a note or to 'Refine writing'.

(Image credit: Sharmishta Sarkar / TechRadar) Read on, there's more

This firmware update also includes a new feature called Side Panel Margin. As the name suggests, this adds a side panel where you can write notes when you're reading an ebook. This feature is also part of the latest firmware rolling out to more users outside the US and is available for both the Scribe models.

The side panel feature is in addition to Active Canvas, which only adds a box for your handwritten note while also reformatting the page. The side panel is quite literally that: a full-length panel for your thoughts, with the ability to alter the width as much as you like.

(Image credit: Sharmishta Sarkar / TechRadar) A better note-taking tablet

This new update means the Kindle Scribe is now a far better e-paper tablet than it was, especially the 2022 original model. The promised rollout of the AI features brings the older Scribe on par with its successor and, given it's now cheaper to buy, it represents better value. I'm still partial to Kobo's Advanced Notebooks though, but I'm thrilled to see Amazon making an effort to make the Scribes more competitive.

The only bone I have to pick with Amazon now – only because I'm Aussie – is that none of the new Kindles now have Audible support Down Under. How about it then, Amazon?

You might also like...

• Find more ereader options in my dedicated best ereader list
• Not interested in the Scribe? Take a look what what I think is the best Kindle right now
• Find out more about the Kindle Scribe's Active Canvas feature
• Need more writing e-paper tablet options? Check out my Onyx Boox Go 10.3 review or my Kobo Elipsa 2E review

The move reportedly comes after US attorney general Pam Bondi reassured Apple and Google they would not be fined for hosting TikTok.

Like many others, I was an early adopter of ChatGPT and enjoyed exploring what it could do. When OpenAI debuted the ChatGPT Plus subscription, I was skeptical about whether I needed the greater access and additional features it promised. However, as the subscription option arrived, many of us using the free version started to encounter slower responses, unexpected usage caps, and other frustrations.

This was likely just because of the growing popularity of ChatGPT, though when particularly annoyed, I did speculate about a dark conspiracy. Either way, $20 a month started to seem reasonable, especially since testing out new ChatGPT features has been a professional need as much as something I personally want to do. I hesitated for a little while. If I had a dollar for every time I almost upgraded, I’d have had enough to pay for ChatGPT Plus for several years. Still, within a few weeks, I clicked on upgrade, and I can say I haven't regretted it at all.

Plus access

Since signing up, I've employed ChatGPT Plus, and its benefits have been extensive in my personal life. I can use it to come up with dinner ideas, entertain children, make New Year's Resolutions, and perform many other tasks. A lot of those would be more difficult or impossible without ChatGPT Plus.

When you sign up, the first and most obvious difference you'll note is that you have access to the ChatGPT models in beta. Currently, the default model is GPT-4o, but with a subscription, you can try the logic-heavy o1 model, the o3-mini model, which is even faster at advanced logic and reasoning, and the o3-mini-high model, which goes even further. I can always tell, even without looking, when a response uses the ChatGPT Plus-only models. The responses are sharper and more insightful, and they often feel like they have been run through an actual human editor.

If I stick to GPT-4o, I still benefit from subscribing to ChatGPT Plus. Those dreaded message limits disappear, which, if you’ve ever had ChatGPT cut you off just when a conversation was getting interesting, is very frustrating. If you hit the limit, the free tier sends you down to the GPT-4o-mini model. That model is fine, but it feels threadbare compared to the richer experience of GPT-4o. Similarly, the limits on uploading files and making and analyzing images also disappear when you subscribe. You might still hit a cap, but you'd have to really work at it.

Plus features

There are also options for GPT-4o exclusive to ChatGPT Plus. You can converse with ChatGPT in both standard and advanced voice modes. You can also play with customizing ChatGPT's abilities by making custom GPTs. The Tasks feature for ChatGPT that lets you schedule out missions for the AI in the future is also only available if you have a subscription to ChatGPT Plus, as it's currently in beta testing.

There is a laundry list of other ChatGPT features that were on the same list but have since migrated to the free tier. ChatGPT Plus subscribers were once the only ones able to use the editable Canvas page, the latest DALL-E image-making model, the web browsing function, and several other tools. Even the GPT-4o-mini model is a huge step up from earlier free-tier models, closing the quality gap significantly.

A fair argument against ChatGPT Plus is that those exclusive perks don't stay exclusive. So, if you’re the type of person who likes waiting for features to trickle down, maybe Plus isn’t necessary. But if you’re someone who wants to be that early adopter and play with the latest models and features immediately, then the subscription still holds a lot of value. It’s not just about what features exist; it's about when you get them and how smoothly they work.

Plus more

If you’re a casual user who just wants ChatGPT for quick questions, occasional brainstorming, or settling silly debates (like whether pineapple belongs on pizza), you’re probably fine sticking with the free tier. It’s gotten much better over time, and OpenAI keeps pushing updates that make it a pretty robust option.

But if you rely on ChatGPT every day for any number of projects at work or home, I'd say that ChatGPT Plus is absolutely worth it. The speed, model quality, unlimited access, and extra tools make a noticeable difference.

When I want a cute image to go with a hand-written greeting card, I don't want to worry about hitting a cap on DALL-E use. And if I want to know if the confusing array of parking signs beside my car say my car won't be towed in the next hour, I want the most powerful and logical AI model available.

If you like using ChatGPT, the real question is how much you use it and whether you’re willing to wait for features to roll out to everyone. If you’re someone who wants the best AI experience right now, ChatGPT Plus is a no-brainer. If you don’t mind waiting for features to trickle down, the free version will probably get there.

Just to be sure, I asked ChatGPT's most advanced model if the subscription was worth it, and ChatGPT declared that if it were human, it would happily pay.

You might also like

• Our predictions for AI in 2025 – what next for ChatGPT, Apple Intelligence and more
• If Sam Altman thinks OpenAI needs to raise ChatGPT prices, I might have to cancel my Plus membership
• Top 3 things you have to try with the new ChatGPT search

Watch for these signs that your gut health might need some attention.