News
For years, users of the best iPads have been contending that their tablets aren’t as capable as they could be, and a big part of that argument centers on window management. After all, the discussion goes, how can you manage multiple apps and workflows if you struggle to keep them all under control?
Well, at its Worldwide Developers Conference (WWDC) earlier this month, Apple attempted to answer those critics emphatically, bringing a huge update to iPadOS 26 that included improved productivity features and multiple overlapping windows, plus several Mac-like additions, including a menu bar and the multi-colored “traffic light” buttons for minimizing and maximizing app windows. It was a massive, long-awaited change.
This got me thinking: while I don’t believe that Apple has any intention of merging iPadOS and macOS for the time being (Craig Federighi even said as much in a recent interview), something is looming on the horizon that’s going to test that resolve to the limit. It might not be long before Apple is forced to reckon with the possibility of cannibalizing its own operating systems in order to make something better.
The foldable future(Image credit: Future)Despite unveiling a bunch of new features at WWDC that brought its iPad and Mac software releases closer together, Apple’s actual hardware remains solidly distinct. You’re unlikely to confuse an iPad tablet for one of the best MacBooks (unless you’ve attached it to a Magic Keyboard mount, of course).
Yet if the rumors are to be believed, this could all change in the next couple of years. That’s because according to Bloomberg reporter Mark Gurman’s latest Power On newsletter, Apple is set to release a 19-inch foldable device by 2028 that sits somewhere between Apple’s tablet and its laptop.
The idea is you can fold it out flat and use it like a giant tablet, or crease it along its hinge and use it as a laptop. Apple has never made something like this, but companies like Huawei have already launched their own attempts. We even reviewed the original Asus Zenbook Fold a couple of years back, when this tech was still in relative infancy. Don’t be surprised to see more in the coming years.
A product like this raises some serious questions regarding software. Right now, Apple says that merging iPadOS and macOS wouldn’t work because macOS isn’t designed for touch and iPadOS should retain its simplicity. In other words, a sturdy boundary should divide the two platforms. But what do you do to your software when your hardware has already obliterated those boundaries?
That might become the reality by late this decade. Depending on what Apple does to its software at that time, Federighi might have to eat his words.
Testing Apple’s resolve(Image credit: Apple)Over the years, Apple has given a few different reasons why it’s not planning to merge iPadOS and macOS, but they’ve all come back to the idea that the two systems are better off apart. By forcing tablets and computers to use the same platform, Apple says, you end up making too many compromises to get the operating system working well for everyone. The result is a bland mixture that loses everything that makes iPadOS and macOS unique and powerful.
But there might be another thought that’s holding Apple back: the fear of cannibalizing its own products. This makes a certain degree of sense – while neither the Mac nor the iPad bring in as much raw cash as the iPhone, they’re still massive moneymakers for Apple. Why jeopardize that?
Interestingly, this might not be a way of thinking that Apple founder Steve Jobs would have embraced. Under his watch, the company occasionally launched devices that overshadowed its other products (or made them outright obsolete). Take the iPhone, for example: for all of Apple’s rivals’ talk of releasing the next “iPod killer,” it was the iPhone itself that leapfrogged the iPod and led to its premature demise.
(Image credit: Future / Lance Ulanoff)That wasn’t some accident – it was a very deliberate decision on Apple’s part. I’m currently reading the Steve Jobs biography by Walter Isaacson, and coincidentally came across the following Jobs quote right around the time Apple was bringing Mac-like features to iPadOS 26: “If you don’t cannibalize yourself, someone else will.” Destroying your own bestsellers every now and then was a way to keep your lineup fresh, embrace exciting advances, and avoid getting stuck in a technological rut.
When the foldable iPad-Mac hybrid arrives a few years down the road, will Apple be able to call upon that mindset when it comes to the software that runs on this device? Will it cannibalize iPadOS and macOS with something newer, stronger, and better suited to its foldable?
Or will it pick an existing operating system and maintain its insistence that iPadOS and macOS must be kept forever discrete, even on the clearest example yet of the company overlapping its hardware?
None of us can know for sure what will happen, but I’ve no doubt that Apple is frantically trying to work out the best approach inside its top-secret labs. This foldable will be the sternest test yet of Apple’s ability to craft a software experience that works for the hardware at hand – and could help us finally learn whether it will ever merge iPadOS and macOS.
You might also like- Chatbots are filling classrooms with fake students
- These chatbots are costing colleges millions in fraudulent loans
- AI is fueling a huge increase in identity theft
The rise of Artificial Intelligence has had a transformative impact on classrooms and college campuses across the world, but a new report from the Associated Press claims chatbots are now posing as students in order to collect financial aid, and are even “attending” online classes - and taking spaces from real students.
Analysis of fraud reports has uncovered a whopping 1.2 million fraudulent applications in California alone, leading to 223,000 suspected fake enrolments. California in particular is a target because of the large number of community colleges.
If your credentials have been compromised in a data breach, take advantage of the best identity theft protection software available to make sure you're protected against this type of attack.
Save up to 68% on identity theft protection for TechRadar readers!
TechRadar editors praise Aura's upfront pricing and simplicity. Aura also includes a password manager, VPN, and antivirus to make its security solution an even more compelling deal.
Preferred partner (What does this mean?)View Deal
Unrecoverable lossesThese “ghost students” don’t just apply for a loan and hope for a payout, though. Chatbots have been deployed to online classrooms, where they stay “just long enough” to collect the financial aid check from the college.
Some lecturers even report having online classrooms with barely any real students as bots have filled the class to the enrollment limit.
According to researchers, criminals stole at least $11 million in local, state, and federal financial aid that ‘cannot be recovered,’ just from California colleges in the last year, so the problem is widespread.
This isn’t the first time we’ve heard of AI scams fueling a rise in identity theft and fraud, with hackers seen ‘leveling up’ their tactics, with more sophisticated attacks leveraged at a higher frequency than ever before.
Many of these scammers will use information obtained from data breaches, such as names, addresses, ages, and social security numbers. With this data they will apply for grants and loans using real credentials.
In March 2025, over 300 people were fired from the Federal Student Aid Office by the Trump administration, and the department’s fraud investigation unit has lost over 20% of its staff through attrition and retirements since October 2025.
You might also like- Take a look at our picks for the best malware removal software around
- Check out our choice for AI tools
- Identity fraud attacks using AI are fooling biometric security systems
Wireless chargers make a super convenient accessory for any Apple gadget fans who want to keep their desktop looking tidy, especially when charging multiple devices at the same time. However, you may find that the same wireless charger that’s keeping your surfaces clutter-free quickly becomes an eyesore all of its own. Let me explain.
Two key considerations when shopping for a wireless charger tend to be charging speeds, naturally, and aesthetics, because it’s something we’ll likely be looking at every day. However, not long into my stint testing almost thirty of them, I learned that it doesn’t matter how much you may like the design of a wireless charger if the finish means it’ll look filthy fast and, worse still, if the materials are difficult to clean.
If you’re looking for a wireless charging station, for example, then it’s pretty likely it’ll be spending its life on your nightstand or your desk, so it’ll inevitably pick up dust day-to-day. That in itself isn’t a revelation, of course, but it turns out that some silicone areas can hold onto dust longer than Taylor Swift holds onto a grudge.
Now, I’m not entirely against silicone; it certainly has its uses, particularly when used over Qi and Qi2 charging pads to help keep devices in place as they charge. However, the way it’s used and the finish it has make all the difference when it comes to keeping a wireless charger dust-free.
(Image credit: Future)For example, non-slip rings, like those pictured above, prove a nightmare for dusting, as lint loves to cling to the rougher finish and raised edges. Whereas the super-soft surfaces of the Anker MagGo Wireless Charging Station 3-in-1 Foldable Pad didn’t pose a problem – though the pesky ridges on the silicone straps were another matter.
Even some of the best wireless chargers, like the Anker MagGo Wireless Charging Station 3-in-1 Stand, can make things a little trickier than they need to be by sticking a silicone area in the middle of an otherwise wipeable surface. Instead, I much prefer when a design includes a recess to help keep wireless earbuds in place as they charge. The ESR Qi2 3-in-1 Watch Wireless Charging Set (HaloLock) is a prime example of this.
Another example of a satisfyingly silicon-free wireless charger is the UGreen 2-in-1 Qi2 Magnetic Foldable Fast Charging Station. Unfortunately, though, the surface of the Qi2 charging pad falls victim to my second-biggest bugbear – fingerprints.
I wish I could give you a straightforward answer on what material to avoid if smudges and fingerprints are going to get on your nerves, but I’m yet to find a definitive answer. You may assume that any matte black plastic is bound to show up fingerprints; some types certainly do, to the point that they never fully wipe clean again, but some eternally appear free of ever having been handled.
The easiest solution, if these imperfections are the sort of things that could prove bothersome to you, is to opt for a charging station that can be easily packed away when not in use. The design of the aforementioned 2-in-1 Qi2 Magnetic Foldable Fast Charging Station is ideal here, as it can be folded down into a neat little cube very easily.
Alternatively, if you’ve got an Apple Watch to charge too, the ESR Qi2 3-in-1 Travel Wireless Charging Set folds down flat, and proved really useful as both a desktop and a travel charger. Otherwise, keep an eye on my guide to the best wireless chargers, as I always take durability and ease of cleaning into account when evaluating potential additions.
The Sony Bravia 8 II is the company’s top OLED TV for 2025, and at $3,500 / £2,999 / AU$3,999, it’s priced at the level you’d expect for a flagship Sony TV.
The Sony Bravia 9, the company’s flagship mini-LED model, was one of the best TVs I reviewed in 2024, so I was very curious to get my hands on Sony’s new flagship OLED. A main reason was that Sony had claimed the new model would be 150% brighter than its Sony Bravia 8 predecessor, an advancement made possible by the company’s switch from a standard W-OLED panel, the type used in last year's Bravia 8, to a QD-OLED panel for the Bravia 8 II.
While our Sony Bravia 8 OLED review was positive overall, the TV’s peak brightness measured significantly below what we’d seen from the best OLED TVs of 2024, such as the Samsung S95D and LG G4. New flagship OLED TVs in 2025 are now even brighter than last year’s models, with the LG G5 measuring 2,268 nits peak brightness, a level that surpasses many of the best mini-LED TVs, when I tested it.
Brightness mattersThe LG G5 (shown above) has a picture that looks bright even with bright room lighting conditions (Image credit: Future)The LG G5 features a new “four-stack” OLED display panel, which LG calls the Primary RGB Tandem Structure. Unlike previous panels, which use two blue OLED layers plus a third with red, green, and yellow elements, LG’s new design produces light via individual red, green, and blue layers. This design results in improved color detail and, notably, increased peak and fullscreen brightness. (See chart below for a benchmark comparison between the Sony Bravia 8 II, LG G5, and competing OLED TVs.)
The G5’s exceptional brightness had a real impact on its picture quality when I reviewed it. Movies with HDR had a near-3D quality due to the picture’s powerful contrast, which made bright highlights in pictures gleam with a high level of intensity. Colors also looked bright, which gave them a vivid quality without looking unnatural or boosted.
Another important factor with the G5 when I tested it was its ability to retain strong contrast when viewing in bright room lighting conditions, something helped by its anti-reflective screen. This made it a great TV for viewing daytime sports, and movies and darker TV shows also held up very well in bright lighting.
I’ve just started testing the 65-inch model of the Sony Bravia 8 II, so I’m only able to make preliminary judgments about its performance at this point. But as you can see in the brightness benchmark chart above, it falls short of Sony’s claimed 150% peak brightness boost over last year’s Bravia 8, which maxed out 817 nits peak and 182 nits fullscreen brightness.
That’s not to say the Bravia 8 II isn’t bright for an OLED TV. Its peak brightness (measured in Cinema mode, the most accurate available picture preset) is about the same as the 65-inch Samsung S90F, that company’s mid-range OLED TV for 2025, and another TV that uses a QD-OLED display panel. Fullscreen brightness is notably lower on the Bravia 8 II compared to the S90F, however.
As I said above, I’m just starting my subjective testing of the Sony Bravia 8 II, so I’ve yet to get a full sense of its capabilities. The Bravia 8 II’s lower brightness compared to the LG G5 (and also several flagship mini-LED TVs I’ve recently tested) means its picture has less of a vibrant pop when viewing in daylight conditions, and its color, while undoubtedly accurate (see picture accuracy chart), also appears a bit less vibrant.
Is the price right?Sony's Bravia 8 II has a compelling picture, but its measured brightness falls below its premium OLED TV competition (Image credit: Future)At $3,500 / £2,999 / AU$3,999 for the 65-inch model, the Sony Bravia 8 II is priced around the same as new flagship OLED TVs such as the LG G5 and Samsung S95D. It has a premium design, along with a great set of audio features such as Acoustic Surface Audio+, which turns the TV’s OLED panel into a speaker, and Acoustic Center Sync, which lets it be used as a center channel when paired with compatible Sony speaker systems and soundbars.
It’s hard to ignore, though, that the new Samsung S90F, an OLED TV with comparable brightness plus a superior set of gaming features, costs $1,000 less at $2,499 / £2,699 / AU$4,299 for the 65-inch model. That’s quite a bit of cash that could otherwise be spent on 4K Blu-rays and other home theater goodies.
Does the Sony Bravia 8 II justify its premium price? We’ll soon have our review wrapped up, and at that point will provide complete thoughts on its performance and value.
In the meantime, the premium OLED TV competition is looking pretty tough for 2025, and Sony’s flagship model has plenty to prove.
You might also likeAs we expected, WWDC 2025 – mainly the opening keynote – came and went without a formal update on Siri. Apple is still working on the AI-infused update, which is essentially a much more personable and actionable virtual assistant. TechRadar’s Editor at Large, Lance Ulanoff, broke down the specifics of what’s causing the delay after a conversation with Craig Federighi, here.
Now, even without the AI-infused Siri, Apple did deliver a pretty significant upgrade to the Apple Intelligence, but it’s not necessarily in the spot you’d think. It’s giving Visual Intelligence – a feature exclusive to the iPhone 16 family, iPhone 15 Pro, or iPhone 15 Pro Max – an upgrade as it gains on-screen awareness and a new way to search, all housed within the power of a screenshot.
It’s a companion feature to the original set of Visual Intelligence know how – a long press on the Camera Control button (or customizing the Action Button on the 15 Pro) pulls up a live view of your iPhone’s camera and the ability to take a shot, as well as “Ask” or “Search” for what your iPhone sees.
It’s kind of a more basic version of Google Lens, in that you can identify plants, pets, and search visually. Much of that won’t change with iOS 26, but you’ll be able to use Visual Intelligence for screenshots. Following a brief demo at WWDC 2025, I’m eager to use it again.
(Image credit: Jacob Krol/Future)Visual Intelligence makes screenshots a lot more actionable, and could potentially save you space on your iPhone … especially if your Photos app is anything like mine and filled with screenshots. The big effort on Apple’s part here is that this gives us a taste of on-screen awareness.
Screenshotting a Messages chat with a poster for an upcoming Movie night in the demo I saw revealed a glimpse of the new interface. It’s the iPhone’s classic screenshot interface, but on the bottom left is the familiar “Ask,” and “Search” is on the right, while in the middle is a suggestion from Apple Intelligence that can vary based on whatever you screenshot.
In this case, it was “Add to Calendar,” allowing me to easily create an invite with the name of the movie night on the right date and time, as well as the location. Essentially, it's identifying the elements in the screenshot and extracting the relevant information.
Pretty neat! Rather than just taking a screenshot of the image, you can have an actionable event added to your calendar in mere seconds. It also bakes in functionality that I think a lot of iPhone owners will appreciate – even if Android phones like the best Pixels or the Galaxy S25 Ultra could have done this for a while.
Apple Intelligence will provide these suggestions when it deems them right – that could be for creating an invite or a reminder, as well as translating other languages to your preferred one, summarizing text, or even reading aloud.
All very handy, but let’s say you’re scrolling TikTok or Instagram Reels and see a product – maybe a lovely button down or a poster that catches your eye – Visual Intelligence has a solution for this, and it’s kind of Apple’s answer to ‘Circle to Search’ on Android.
You’ll screenshot, and then after it’s taken, simply scrub over the part of the image you want to search. It’s a similar on-screen effect to when you select an object to remove in Photos ‘Clean Up’, but after that, it will let you search it via Google or ChatGPT. Other apps can also opt in for this API that Apple is making available.
(Image credit: Jacob Krol/Future)And that’s where this gets pretty exciting – you’ll be able to scroll through all the available places to search, such as Etsy or Amazon. I think this will be a fan-favorite when it ships, though not entirely a reason to go out and buy an iPhone that supports Visual Intelligence ... yet, at least.
Additionally, if you’d rather search for just the whole screenshot, that’s where the ‘Ask’ and ‘Search’ buttons come in. With those, you can use either Google or ChatGPT. Beyond the ability to analyze and suggest via screenshots, or search with a selection, Apple’s also expanding the types of things that Visual Intelligence can recognize beyond pets and plants to books, landmarks, and pieces of art.
It wasn’t all available immediately at launch, but Apple is clearly working to expand the capabilities of Visual Intelligence and enhance the feature set of Apple Intelligence. Considering this gives us a glimpse into on-screen awareness, I’m pretty excited.
You might also like- I spoke to Apple’s software engineering VP for the inside story on how iPadOS 26 finally became a real Mac alternative
- The best value Mac just got even cheaper – the M4 Mini is an absolute bargain at Amazon today
- I just experienced super-smooth Cyberpunk 2077 at Ultra settings on a Mac, but the developers say there’s more to ‘squeeze out’ of Apple Silicon
This year’s Wimbledon tennis championships is set to be the most interactive and exhaustive for sports fans thanks to a major AI upgrade.
The iconic tournament, long a mainstay of the British summer, is introducing a range of AI-powered upgrades and services for fans through its Wimbledon.com website and app.
This includes an all-new feature allowing fans to explore information about certain matches in almost real-time, and an upgraded tool looking to predict the possible winners of every match.
Match Chat and moreIBM has now been the technology partner for the All-England Lawn Tennis Club (AELTC), tournament organizers for Wimbledon, for 36 years, with 2025 marking another significant milestone for the pair.
Following in the footsteps of recent AI-powered innovations such as 2024’s Catch Me Up, which utilized Watsonx's generative AI to create player-based updates in the form of "cards" available on the Wimbledon.com website or mobile app, and AI commentary introduced in 2023, the two organizations hope these new advances will offer fans old and new greater insights into the tournament.
"The way sports are being consumed is ever-evolving,” Kevin Farrar, Head of Sport Partnerships at IBM UK, told a pre-tournament briefing attended by TechRadar Pro, “our challenge is to see how we tap into that”.
(Image credit: IBM / AELTC)New for 2025 is Match Chat - a new conversational interface which lets fans explore info about the match in nearly real-time.
The tool is built with technologies on watsonx Orchestrate, taking AI agents and large language models (LLMs), such as IBM Granite, which have been trained on the Wimbledon editorial style and language of tennis - so “Gentlemen’s” and “Ladies’” singles, rather than Men’s/Women’s.
Fans will be able to use a number of pre-written prompts, or ask their own questions (such as, ‘who has served the most aces in the match?’, or ‘who is performing better in the match?’), with replies delivered almost immediately.
IBM says the Match Chat training also ensures the tool stays focused on the tennis - users will only be able to ask it questions about matches at the tournament, so there’s hopefully no chance it will get distracted if you ask it where the tastiest strawberries are.
“Whenever we're designing something new, it always starts with the fan first...we think this is going to be a really engaging experience that addresses a number of different kinds of fans,” noted Chris Clements, Digital Products Lead at the AELTC.
"At its heart, sport is a human thing, it's an emotional thing - we're using AI to be able to tell these stories to be told more effectively."
First introduced in 2024, the “Likelihood to Win” tool is also getting an AI boost, and will now alter its projected win percentage even throughout a single game, generating projections from AI-powered analysis of player statistics, expert opinion and match momentum.
The 2025 Wimbledon Championships run from June 30 - July 13, 2025, with the app available to download on Android and iOS now, as well as across the Wimbledon.com website.
More from TechRadar Pro- OM System's OM-5 II is a modest upgrade of the OM-5
- Available in three colorways, including a limited edition Sand Beige
- Body only price is £1,099 / AU$1,699.95 (US pricing TBC)
I'm a fan of OM System's Micro Four Thirds cameras. They're compact, travel-friendly, compatible with a huge range of superb lenses, deliver incredible image stabilization for easy handheld shooting, plus their computational photography modes are addictively fun.
OM System cameras hit the mark on many fronts. But what they have also hit, it seems, is a ceiling. Case in point – the new OM System OM-5 II. It comes two and a half to three years after the OM-5, but you wouldn't know it – there's so little to differentiate between the two cameras.
That's no bad thing per se, we still rate the OM-5 as a top travel camera. But where Panasonic is adding meaningful improvements to its Micro Four Thirds cameras, especially for video capture in the Lumix GH7 and Lumix G9 II, in the OM-5 II we get USB-C charging, some video color profiles, and a rejigged menu. That's just about it.
I can't say I'm surprised. Ever since OM Digital Solutions acquired Olympus, the most notable updates we've seen in new cameras is OM System rebranding. I was still hoping for something bigger in the OM-5 II, though. If OM System was properly investing in the Micro Four Thirds system, there has been enough time since the acquisition for it to have started introducing new tech.
The OM-5 II is a highly rugged camera, ideal for travel and the outdoors (Image credit: OM System)Instead, what we get is the same 20MP MFT sensor with 5-axis image stabilization, a modest 1.04m-dot touchscreen and run of the mill 2.36m-dot EVF, albeit packaged in a retro and rugged body. The OM-5 II still looks the part, and I'm a fan of the limited edition Sand Beige – it looks fab.
At least the legendary Olympus brand hasn't been killed off altogether, and continues to live on under a new name, because I still believe there's a place for such cameras.
Micro Four Thirds cameras, especially the inherited Olympus design ethos, hit a certain quality / portability sweet spot. I just wish OM System was giving fans a little more to be excited about going forward.
Can we ever expect meaningful upgrades again?The glass-half-empty types of have preaching a doom and gloom for Mirco Four Thirds for some time now.
'Micro Four Thirds isn't dead', comes the response from fans who love what the camera system represents; superb build quality, a wide range of optics for specialist interests such as wildlife, birding and more, all in a lightweight system which weighs a fraction of full-frame.
But the fact remains, perhaps more specifically for OM System rather than Panasonic – we haven't seen any decent updates to its new cameras for years.
It's part of the Micro Four Thirds system, with many compatible lightweight lenses, such as the 12-45mm PRO, above, with which it is available as a kit (Image credit: OM System)If OM System was indeed investing in future MFT cameras, I think we would have started to see it this year. Earlier in the year it launched the OM-3 – the first in a series with a slightly different retro styling. It was a delight to use, but not because of big technological improvements, but because Micro Four Thirds remains a really fun and versatile system to shoot with.
I'm glad MFT is here to stay for another few years until the next update cycle lands. However, at that point I'm slightly concerned that we'll discover the system has sung its final song, to live on only through its fans.
Do you love Micro Four Thirds photography? What do you think of its future? Let me know in the comments below.
You might also likeArtificial Intelligence (AI) touches virtually every industry, but it’s become a foundational element in today’s customer experience (CX) strategies. Contact centers, customer support platforms, and digital engagement tools rely on AI to enable faster response times, more personalized interactions, and to uncover valuable insights from massive amounts of customer data. Conversational AI, real-time voice analytics, and intelligent routing are just a few of the innovations transforming how organizations connect with their customers.
While there are plenty of benefits to AI, one thing remains true: AI will never be entirely free from bias. This is because AI is only as accurate as the data it was trained on - which is ultimately created, trained, and maintained by humans - humans, who unconsciously bring their own assumptions and blind spots into the AI systems they build.
This doesn’t mean AI can’t be trustworthy, responsible or fair. It simply means organizations need to implement strong guardrails and standards for monitoring and refining AI models to ensure fairness, inclusion, and neutrality. Mitigating bias is essential across industries, but is especially important in CX - not just for stronger performance and efficiency, but to build and maintain long-term customer trust and regulatory compliance.
Reducing AI bias improves agent performance and efficiencyWhen using AI to automate customer service tasks or assist human agents, even the smallest of biases in data can lead to low-quality experiences. For example, speech recognition tools might struggle to understand different accents and dialects, leading to frustrating customer experiences. Sentiment analysis might misread emotional cues, resulting in inaccurate responses or escalation to the wrong agent. Intelligent routing workflows can unintentionally prioritize certain customer profiles over others if historical training data skews unfairly.
These inconsistencies don’t just impact customers, but agents as well. Human agents may have to step in more often to correct AI mishaps or hallucinations, increasing their cognitive workload and decreasing employee morale, reducing the overall efficiency that AI-powered tools promise to deliver. Additionally, it decreases trust in the technology for agents, potentially leading to negative perceptions of how AI is used and how it is impacting their work.
To address these challenges, organizations need to start by using diverse datasets to train AI models and ensure they can adapt to evolving inputs. From there, constantly auditing and refining data allows organizations to weed out biases before they creep into outputs, ensuring more fair, accurate results. Additionally, monitoring real-time customer feedback across multiple channels gives organizations a strong idea of where customer frustrations are occurring and allows them to take another look at the data feeding those interactions.
Ethical AI builds customer loyalty and supports complianceToday’s consumers are more tech-savvy and privacy-conscious than ever. While recent data shows that more than half of consumers say AI alone doesn’t negatively impact their trust, how customer data is used with it can.
Organizations can address these concerns by adopting privacy-first principles to maintain trust and show commitment to responsible AI practices. Taking steps like encrypting sensitive data, restricting access through strong identity controls, and anonymizing customer data used in AI training models are great examples of a privacy-first approach. Transcripts, voice recordings, and behavior patterns must be handled with care - not just to build trust, but to comply with privacy laws like the GDPR, CCPA and the EU AI Act.
Transparency with consumers is equally as important, especially as it relates to how and what data is collected. Giving customers control over their data, ensuring transparent AI governance, clearly disclosing the use of AI chatbots or tools, and providing seamless escalation to human agents when needed, fosters a sense of trust among customers. Organizations that share how AI is used and decisions are made are likely to earn long-term customer loyalty.
What is easily forgotten is that there is an entire industry segment called Workforce Engagement Management and part of that is coaching agents and getting customer feedback. The ethics of best practice are already in place. Whether it is a virtual agent or a real agent, the principle of improving and compliance still applies. What AI can bring is that the time between the potential error and the review of that mistake can be almost instantaneous. We can also use AI to check AI and compare the ethical answer with the actual answer. Just make your AI agents trainable as you would with your human agents.
Responsible AI enables responsible innovationAI-driven innovation seems to move at the speed of light, but innovation doesn’t have to come at the expense of responsibility. Unsurprisingly, the most forward-thinking organizations are those that embed ethical principles into the innovation process from day one. Achieving this means fostering open collaboration between developers, data scientists, business stakeholders, and IT teams to ensure that both innovation and security are balanced.
Establishing a clear AI governance framework or roadmap helps align stakeholders around a clear vision for ethical AI. When standards and processes are both clearly defined and consistently applied, organizations can scale innovation more responsibly and confidently.
Bias in AI is a complex issue that nearly every organization has or will face - but it’s not an unsolvable one. Feeding diverse datasets into AI training models and then consistently auditing the data helps to mitigate bias. While truly bias-free AI may be difficult to achieve, understanding the challenges and continuously working to limit bias leads to stronger customer loyalty, enhanced compliance, and more opportunities to innovate at scale.
This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
- The long-awaited Philips Hue AI assistant is now available in some countries
- It will be launched in the UK in July, and globally by the end of August
- The Philips Hue Play Wall Washer light is on sale today in the US and UK
If you want to have more fun with your smart lights, there's good news – the long-awaited Philips Hue AI assistant is finally here, letting you pick lighting scenes or create brand new ones with simple voice commands. Signify, the company behind Philips Hue, first teased the assistant back in January, and last week a few iPhone users found themselves with early access, but the launch is now official.
The assistant is available now for users in the Netherlands, Belgium, and Luxembourg whose app is set to English, and will be rolling out to Hue users in the UK in July. According to Signify, the global rollout is planned for the end of August.
With the AI assistant, you can either say or type what type of lighting effect you'd like to see, or what mood you want to set, and the app will either suggest something from its existing gallery of presets, or create something brand new if there's nothing that quite fits the bill.
Clean sweepThat's not all – Signify has also launched a new lamp that bathes your walls with a gradient of light. We got our first peek at the Philips Hue Play Wall Washer earlier this month when it was listed on Amazon ahead of its official release date.
Now it's available to buy, and despite measuring just 3.6 x 3 x 6.2 inches / 9.1 x 7.6 x 15.7cm (that's even smaller than the Philips Hue Play Light Bar), it promises to drench your whole wall with a smooth gradient of colored light that you can customize via a 3D drag-and-drop interface in the app, allowing you to set the direction and brightness of the light to suit your room.
The Wall Washer works with the Philips Hue Play HDMI Sync Box or Hue Sync desktop app for PC, allowing you to synchronize lighting effects with movies, TV shows, and games.
Alternatively, you can use it alone, or synced with other Philips Hue smart lights to help create a mood. It uses the same ColorCast light system as the Philips Hue Twilight lamp (one of the best smart lights we've ever tested), projecting a smooth gradient onto nearby surfaces. The video below gives you an idea of the overall effect.
The Wall Washer also looks appealing when switched off. Unlike the Hue Play Light Bar, which has a plastic case, the Wall Washer is finished in black or white matt aluminum.
The Wall Washer is available now in the US and UK, priced at $219.99 / £169.99 for a single light, or $384.99 / £299.99 for a pack of two. It's due to launch in Australia in September, but official pricing has yet to be announced.
You might also likeThe increase of remote and hybrid working, as well as digitization and networking of a wide range of devices and systems, has made IT landscapes much more complex. Employees use so many devices – desktop computers, laptops, tablets, phones – that it’s all too easy to unwittingly give out information.
While organizations being exposed to cyber criminals is nothing new, over half of businesses in the U.S and UK have been targets of a financial scam powered by ‘deepfake’ technology, highlighting ‘deepfake’ scams are a high concern.
In today’s digital landscape, CEOs and CFOs have large digital footprints. They have speeches, interviews and videos across many social media and business channels like YouTube and LinkedIn as well as corporate websites.
And while generative AI has transformed the way people can work and create the vast amount of online content now available is providing criminals with endless material to generate convincing deepfakes, which are being used by scammers worldwide.
In May 2024, British engineering group Arup was duped into transferring $25 million to cybercriminals. The employee attended a video call, where everyone looked and sounded like familiar coworkers and bosses. But everyone in the call was a deepfake, AI-generated imitations of real people used to manipulate the employee into making the transfer.
This wasn’t an isolated incident, either. Advertising group WPP were also targeted for a deepfake scam but thankfully, it was unsuccessful. The group's CEO detailed the attempted fraud in an email to leadership, warning them to look out for calls claiming to be top executives.
The number of deepfake attacks in the corporate world has surged in recent years. The use of rapidly advancing and now widely available technology is making it possible, and people in workplaces are susceptible to falling for it.
Why does this matter to youThis deepfake technology presents a growing threat to businesses, particularly through financial fraud and so when scams like these happen, the damage isn’t just monetary, it can also come back on you. If you were the one who let the scammer in, accidentally shared sensitive data, or approved a fraudulent request, you could be held accountable, even if you didn’t realize what was happening.
AI-generated deepfakes exploit the element of trust, so while cybercriminals might be targeting your employer, you may be the entry point. Corporate deepfake fraud undermines business confidence and public trust.
Defending Your Employer (and Your Job)Given how quickly these threats are evolving, organizations and their employees must develop adequate safeguards and policies to stay safe from exploitation.
Take Your Time and Confirm
Make sure you scrutinize and verify before responding to requests received digitally, especially if they include a request to disclose sensitive information or conduct financial transactions. If you’re encouraged to respond to any requests via phone or video call, call back using the channels you’re familiar with to confirm the task.
Watch for Signs of Unusual Behavior
If a co-worker’s voice sounds a bit off or their camera seems strangely blurry, it may be a sign of something unusual. Other signs that can indicate something is amiss include unnatural blinking or speech that is out of sync with their lips. AI and deepfakes can be deceiving, but they’re not perfect.
Create a culture of cyber awareness
Encourage conversations with your colleagues that allow you to take a step back, pause and raise concerns whenever you feel concerned about a request. And while AI can be useful for a myriad of tasks, workplaces need to have detailed guidelines on its use.
Verify Attendees Before Letting Them In
If you've been invited to a meeting, double-check the invite to ensure you know who the sender is. If you’re hosting a meeting, it’s worth enabling waiting rooms or lobbies so you can approve who joins.
Don’t Hesitate to Question Unusual IT Assistance
If someone appears in a meeting claiming to be from IT and begins asking you to install software or allow them access, be cautious. Instead, verify with your IT department through your usual work channels about what the procedure is to make changes to your device.
We list the best online cybersecurity course.
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
- New Magnetic Switch to turn the earbuds on and off without the case
- Up to 65 hours of battery life with ANC off
- £125 / €149 / $149
If you're looking for earbuds to wear on a mission to space, Audio-Technica have just the things: their new ATH-CKS50TW earbuds have an extraordinary 65-hour battery life in total, which is just about long enough to fly to the moon. That makes my AirPods Pro 2 look pretty feeble.
The buds' own batteries deliver 25 hours of continuous playback, and the charging case adds another 40. Those figures are with active noise cancelling (ANC) off, but with ANC enabled the numbers are still astounding: 15 hours from the buds and a further 25 from the case.
Again, for comparison, the AirPods Pro 2 give you six hours from the buds alone. The Sony WF-1000XM5 give you eight hours. Audio-Technica's new earbuds absolutely crush any of the best earbuds in this measure.
(Image credit: Audio-Technica)Audio-Technica ATH-CKS50TW2: key features and pricingThe icing on the long-lasting cake is that the new earbuds will cost only $149 / £125 (about AU$260).
One of the more unusual new features here is a magnetic switch, which powers the buds on or off by separating or joining their built-in magnets. The idea is to be able to turn off the buds without having to pop them in the case, although I'm not sure there are many people who've been cursing the tyranny of charging cases. It's nice to have the option, though.
The case is also compatible with Qi wireless chargers, which is somewhat rare among affordable earbuds – you don't get it from the Sony WF-C710N or the Nothing Ear (a) for example.
The earbuds feature hybrid ANC with hear-through and talk-through modes, Bluetooth LE Audio with the more advanced LC3 codec, custom-designed 9mm drivers with extended low-end response, and hybrid hard and soft silicone ear tips. They're waterproof and dustproof, rated IP55.
The specs and the battery specs in particular are impressive, but it's worth noting that Audio-Technica has had a few issues with earbud batteries in the past: its SQ1TW2 wireless earphones had a faulty batch that overheated and even produced smoke, and there has also been a recall of the charging case for the ATH-CK3TW earbuds – again due to overheating.
So while I'm glad to see game-changing battery life, if A-T is pushing the limits of battery tech here you might want to keep an eye out for any recalls, just in case. (No pun intended.)
You might also like- Salesforce research finds single-turn tasks see only 58% success, while multi-turn effectiveness drops to 35%
- Reasoning models like gemini-2.5-pro tend to outperform lighter models
- CRMArena-Pro has proven to be a challenging benchmark
Researchers from Salesforce AI Research have introduced a new benchmark – CRMArena-Pro – which uses synthetic enterprise data to access LLM agent performance in difference CRM scenarios.
It found LLM agents achieved around 58% success on tasks which can be completed in a single step, with tasks that require multiple interactions dropping in effectiveness to just 35% – barely more than one in three.
Although models like gemini-2.5-pro achieved over 83% success in workflow execution, the Salesforce researchers still highlighted some concerns with AI agents, suggesting they might not quite be up to scratch after all.
Are AI agents actually that good?The paper, entitled 'Holistic Assessment of LLM Agents Across Diverse Business Scenarios and Interactions', explained that LLM agents displayed near-zero inherent confidentiality awareness, noting that their performance in handling sensitive information is only improved with explicit prompting (which often came at the expense of task success).
They also criticized previous and existing benchmarks for failing to capture multi-turn interactions, addressing B2B scenarios or confidentiality, and reflecting realistic data environments. CRMArena-Pro is build on synthetic data validated by CRM experts, covering B2B and B2C settings.
In terms of analysis results, reasoning models like gemini-2.5-pro and o1 outperformed lighter models most of the time – Salesforce's researchers concluded that models that seek more clarifications generally perform better, especially in multi-turn tasks.
For example, while the average performance across the nine models tested (three each from OpenAI, Google and Meta) resulted in a score of 35.1%, gemini-2.5-pro scored 54.5%.
"These findings suggest a significant gap between current LLM capabilities and the multifaceted demands of real-world enterprise scenarios, positioning CRMArena-Pro as a challenging testbed for guiding future advancements in developing more sophisticated, reliable, and confidentiality-aware LLM agents for professional use," the researchers concluded.
Looking ahead, Salesforce CEO Marc Benioff views AI agents as a high-margin opportunities, with major corporate clients including governments betting on AI agents for boosted efficiency and further cost savings.
You might also like- These are the best AI tools and best AI writers
- Check out our roundup of the best productivity tools
- Many businesses are thinking twice on using AI bots