TechRadar News

A critical development in artificial intelligence is agentic AI, and these tools are becoming a part of everyday life faster than we realize. Unlike traditional AI models that follow predefined rules, agentic AI systems can make decisions, perform actions, and adapt to situations based on their programming and the data they process, often without human input.

In comparison, traditional AI models are unable to adapt to unexpected changes without a form of retraining, testing and validation, which requires human intervention. A traditional AI model will typically be built to fulfil a specific role, such as a classification task where it will identify if a person is likely to default on a loan payment or not.

From self-instructed house robots to focused data analysis, agentic AI promises convenience and efficiency. However, with that shift comes much deeper access to our personal data, which raises new concerns over transparency, trust, and control.

Agentic AI: The shift from support to self-direction

Agentic AI behavior is goal-driven – it must determine how to achieve its primary goal and sub-goals, requiring it to prioritize tasks and solve problems independently of humans. For example, a house robot might be instructed to “keep the house clean”. The system will then act independently to assess different areas of the home and undertake tasks where appropriate, without requiring constant human intervention. As part of this process, the robot will identify individual sub goals, such as tidying the living room or vacuuming dirty floors, while making its own decisions to achieve the objectives.

Alternatively, businesses might employ agentic AI to analyze datasets over a specified period and identify trends and preferences, which can then be applied to other company strategies. For instance, creating targeted email campaigns. Automating the process brings instant access to valuable insights, increasing human capacity for time spent on more strategy-focused tasks.

These examples illustrate the power agentic systems have to identify opportunities, execute strategies, and adapt to different goals. But as the use of agentic AI increases, businesses and developers should carefully evaluate our trust in them and how they influence our human-machine relationships.

The evolving human-machine link

Agentic AI is fundamentally transforming the human role in tasks through automation, creating a more balanced human-to-machine relationship. As agentic systems utilize deep learning and complex image and object recognition, they can operate in increasingly dynamic environments and solve complex problems autonomously, without any human involvement.

Reduced human intervention not only offers greater efficiency in work, both at home and in businesses, but also frees up time to focus on more strategic initiatives. However, as trust builds towards these automation tools, reduced human oversight comes with a potential risk of over-reliance. While we benefit from agentic AI’s efficiency, we must also ensure we have the opportunity to upskill and educate ourselves.

Whilst agentic systems can operate independently, they still require value and goal alignment to maintain control over the outputs and ensure it aligns with the desired outcome, not just what was instructed of them. Otherwise, there is concern that these systems may take dangerous shortcuts or bypass other infrastructure to achieve efficiency.

The future of AI ethics and privacy demands

In turn, this raises numerous ethical concerns surrounding agentic AI. One key debate is data privacy and the security of sensitive data – this can vary in severity depending on the industry in which the organization operates.

For example, agentic AI has already been deployed by cybersecurity companies to detect and correlate threats by analyzing network activity in real-time and then autonomously responding to potential breaches.

However, organizations who implement this must provide data to the system, raising questions over the security and privacy of their information. Without human oversight, organizations must consider if they are comfortable with agentic AI making business-altering judgement and being on the frontier of their most valuable assets.

Bias in agentic AI can occur because of human input and data. When such a system is entrusted with making moral decisions that have real-world consequences, it faces significant ethical considerations. While existing or emerging legislation provides guidance, there is still much to be done to unpack and implement these concepts within operationalized AI systems fully.

Furthermore, systems with access to highly sensitive data have raised security concerns as they can have vulnerabilities exploited. This has been highlighted by recent cyber attacks, and information within digital environments is at risk. In such complex systems, who is held accountable if things go wrong?

Key considerations for ethical AI use

While we are guided by ethical principles and emerging legislation, it is essential that we have safeguards in place for both agentic and traditional AI systems. When automating manual tasks and analyzing datasets, it is crucial to identify and mitigate bias in both data and algorithms with consistent human oversight. Organizations that utilize agentic AI should strive for ethical practice which can be supported through training and continuous auditing. This helps ensure fairness and prevents harm while creating transparency around how automated decisions are made.

We've featured the best private browser.

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

• Alfie Allen shares 3 Body Problem creators have "pretty epic" plans for season 2 episodes
• He's one of the main new additions to season 2's cast
• Production is still in its early stages

3 Body Problem season 2 is currently in production, with an expected release date yet to be confirmed (though common sense tells us we can expect it in mid 2026 at the earliest). I've already been honest about not being able to remember the Netflix show's complex lore now, let alone by the time new episodes do eventually appear. But someone who is going to need to get their head around it all very quickly is Alfie Allen.

The Game of Thrones actor has been cast in a currently unknown role, alongside fellow newcomers David Yip, Jordan Sunshine, Claudia Doumit and Ellie de Lange. I'm speaking to Allen about his new series, Atomic, starring alongside Shazad Latif as a drug deal who runs into trouble when the cartel wants him to smuggle uranium. Hopefully, there's not too many similarities between the two shows, but 3 Body Problem season 2 is certainly going to pile on some unexpected pressure.

But what that pressure is exactly is unknown. We left season 1 with civilization threatened to be wiped out by the San-Ti, with the launch of Will's brain towards their fleet failing to stay on course (that will make more sense when you watch it). The conflict is far from over, and I can't help asking Allen what we can expect from upcoming episodes.

3 Body Problem season 2 episodes set to be 'pretty epic' according to new star Alfie Allen

“I'd second that without a doubt,” Allen tells me as I ask the pair about the two major projects they have on the horizon. “I'm just starting on 3 Body Problem season 2, I've only done two days [of filming]. But it's great. The stuff they've got planned out is pretty epic. It’s all so good.”

Given that we currently have no idea who Allen is playing or exactly how he's involved with the show, this literally could mean anything. It's less likely, but not impossible, that he's going to be on the side of the San-Ti, with the alien fleet needing new foes to fact as their ominous threat looms ever closer.

“The tone really changes once you get to the second season,” co-creators, writers, and executive producers David Benioff, D.B. Weiss, and Alexander Woo said in a statement to the streamer. “It’s much bigger. Things get wilder, the conflicts get more cosmic. Once people see it on the screen, they’re going to understand why we’ve been so excited for this season, in particular.”

It's not even much help to read the original book series given how much changed between season 1 and the first installment. It also presents another question – if production is only just beginning and season 2 is going to be even more involved than the first... how long are we actually going to be waiting for it?

I'm keeping my fingers crossed that season 2 won't be a repeat of Stranger Things season 5, but the fact that they have a similar level of sci-fi visual output, I'm nervous. Still, the cogs are moving, and I'm pretty sure we're all only just beginning to understand what we're letting ourselves in for.

You might also like

• Wednesday season 3 is confirmed by Netflix and I’m thrilled this Tim Burton series isn’t dead in the ground
• Emily in Paris season 5 will be a ‘tale of two cities’ as Netflix confirms release date, and that’s a disaster for Virgin River season 7
• Netflix renews Dept. Q for season 2, and I think it should time its release date to steal a rival HBO Max show's thunder

Looking for a different day?

A new NYT Strands puzzle appears at midnight each day for your time zone – which means that some people are always playing 'today's game' while others are playing 'yesterday's'. If you're looking for Monday's puzzle instead then click here: NYT Strands hints and answers for Monday, August 25 (game #540).

Strands is the NYT's latest word game after the likes of Wordle, Spelling Bee and Connections – and it's great fun. It can be difficult, though, so read on for my Strands hints.

Want more word-based fun? Then check out my NYT Connections today and Quordle today pages for hints and answers for those games, and Marc's Wordle today page for the original viral word game.

SPOILER WARNING: Information about NYT Strands today is below, so don't read on if you don't want to know the answers.

NYT Strands today (game #541) - hint #1 - today's themeWhat is the theme of today's NYT Strands?

• Today's NYT Strands theme is… Bee friendly

NYT Strands today (game #541) - hint #2 - clue words

Play any of these words to unlock the in-game hints system.

• BULB
• FAME
• RAIL
• CLAY
• GAMES
• MALE

NYT Strands today (game #541) - hint #3 - spangram lettersHow many letters are in today's spangram?

• Spangram has 10 letters

NYT Strands today (game #541) - hint #4 - spangram positionWhat are two sides of the board that today's spangram touches?

First side: bottom, 3rd column

Last side: top, 4th column

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

NYT Strands today (game #541) - the answers

(Image credit: New York Times)

The answers to today's Strands, game #541, are…

• DAISY
• BLUEBELL
• PRIMROSE
• MARIGOLD
• CHAMOMILE
• SPANGRAM: WILDFLOWER

• My rating: Easy
• My score: Perfect

When I saw the theme “Bee friendly” I automatically thought of nectar and flowers and quickly spotted the WILDFLOWER spangram.

Despite this being a relatively easy game I still struggled putting together a couple of the words.

I got DAISY and BLUEBELL easily enough, but it took me a while to see that the space between them was occupied by CHAMOMILE, which I associated with tea rather than flowers.

PRIMROSE I discovered in two parts – first “rose”, which even I know is not a wildflower, and then “prim”.

Yesterday's NYT Strands answers (Monday, August 25, game #540)

• LAPTOP
• NOTEBOOKS
• FOLDERS
• BACKPACK
• CALCULATOR
• SPANGRAM: SUPPLIES

What is NYT Strands?

Strands is the NYT's not-so-new-any-more word game, following Wordle and Connections. It's now a fully fledged member of the NYT's games stable that has been running for a year and which can be played on the NYT Games site on desktop or mobile.

I've got a full guide to how to play NYT Strands, complete with tips for solving it, so check that out if you're struggling to beat it each day.

Looking for a different day?

A new NYT Connections puzzle appears at midnight each day for your time zone – which means that some people are always playing 'today's game' while others are playing 'yesterday's'. If you're looking for Monday's puzzle instead then click here: NYT Connections hints and answers for Monday, August 25 (game #806).

Good morning! Let's play Connections, the NYT's clever word game that challenges you to group answers in various categories. It can be tough, so read on if you need Connections hints.

What should you do once you've finished? Why, play some more word games of course. I've also got daily Strands hints and answers and Quordle hints and answers articles if you need help for those too, while Marc's Wordle today page covers the original viral word game.

SPOILER WARNING: Information about NYT Connections today is below, so don't read on if you don't want to know the answers.

NYT Connections today (game #807) - today's words

(Image credit: New York Times)

Today's NYT Connections words are…

• TOAD
• RIDE
• TATTOO
• BAR
• MOLE
• FOOT
• HEEL
• BADGER
• RAT
• HARRY
• STEP
• SCAR
• CREEP
• PIERCING
• BUG
• SKUNK

NYT Connections today (game #807) - hint #1 - group hints

What are some clues for today's NYT Connections groups?

• YELLOW: Useful ID features
• GREEN: A contemptible person
• BLUE: Nag, nag, nag
• PURPLE: Add a sit down word that rhymes with “fool”

Need more clues?

We're firmly in spoiler territory now, but read on if you want to know what the four theme answers are for today's NYT Connections puzzles…

NYT Connections today (game #807) - hint #2 - group answers

What are the answers for today's NYT Connections groups?

• YELLOW: DISTINGUISHING CHARACTERISTICS
• GREEN: A REAL JERK
• BLUE: PESTER
• PURPLE: WORDS BEFORE "STOOL" 

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

NYT Connections today (game #807) - the answers

(Image credit: New York Times)

The answers to today's Connections, game #807, are…

• YELLOW: DISTINGUISHING CHARACTERISTICS MOLE, PIERCING, SCAR, TATTOO
• GREEN: A REAL JERK CREEP, HEEL, RAT, SKUNK
• BLUE: PESTER BADGER, BUG, HARRY, RIDE
• PURPLE: WORDS BEFORE "STOOL" BAR, FOOT, STEP, TOAD

• My rating: Hard
• My score: 2 mistakes

It took me two goes to get A REAL JERK, because I included MOLE instead of CREEP thinking this was a group made of criminal-related slang. And before that I foolishly lumped together HEEL, FOOT, and STEP (plus BAR) in the mistaken belief that there was some kind of podiatry connection.

After giving away these two mistakes I made progress with the green and blue groups, although I wasn’t sure how RIDE made it alongside BADGER, BUG, and HARRY.

DISTINGUISHING CHARACTERISTICS was probably a lot easier if you work in the emergency services or watch a lot of crime procedurals, but it took a lot of staring at the final eight tiles for the penny to drop for me. Arduous.

Yesterday's NYT Connections answers (Monday, August 25, game #806)

• YELLOW: PREDICAMENT FIX, JAM, MESS, PICKLE
• GREEN: SMALL SPHERICAL THINGS MOTHBALL, PEA, PEARL, POM-POM
• BLUE: PIPE-SMOKING ACCESSORIES FILTER, LIGHTER, PIPE CLEANER, TAMPER
• PURPLE: WHAT "DOWN" MIGHT MEAN FEATHERS, GUZZLE, SAD, WILLING

What is NYT Connections?

NYT Connections is one of several increasingly popular word games made by the New York Times. It challenges you to find groups of four items that share something in common, and each group has a different difficulty level: green is easy, yellow a little harder, blue often quite tough and purple usually very difficult.

On the plus side, you don't technically need to solve the final one, as you'll be able to answer that one by a process of elimination. What's more, you can make up to four mistakes, which gives you a little bit of breathing room.

It's a little more involved than something like Wordle, however, and there are plenty of opportunities for the game to trip you up with tricks. For instance, watch out for homophones and other word games that could disguise the answers.

It's playable for free via the NYT Games site on desktop or mobile.

Looking for a different day?

A new Quordle puzzle appears at midnight each day for your time zone – which means that some people are always playing 'today's game' while others are playing 'yesterday's'. If you're looking for Monday's puzzle instead then click here: Quordle hints and answers for Monday, August 25 (game #1309).

Quordle was one of the original Wordle alternatives and is still going strong now more than 1,100 games later. It offers a genuine challenge, though, so read on if you need some Quordle hints today – or scroll down further for the answers.

Enjoy playing word games? You can also check out my NYT Connections today and NYT Strands today pages for hints and answers for those puzzles, while Marc's Wordle today column covers the original viral word game.

SPOILER WARNING: Information about Quordle today is below, so don't read on if you don't want to know the answers.

Quordle today (game #1310) - hint #1 - VowelsHow many different vowels are in Quordle today?

• The number of different vowels in Quordle today is 3*.

* Note that by vowel we mean the five standard vowels (A, E, I, O, U), not Y (which is sometimes counted as a vowel too).

Quordle today (game #1310) - hint #2 - repeated lettersDo any of today's Quordle answers contain repeated letters?

• The number of Quordle answers containing a repeated letter today is 1.

Quordle today (game #1310) - hint #3 - uncommon lettersDo the letters Q, Z, X or J appear in Quordle today?

• Yes. One of Q, Z, X or J appears among today's Quordle answers.

Quordle today (game #1310) - hint #4 - starting letters (1)Do any of today's Quordle puzzles start with the same letter?

• The number of today's Quordle answers starting with the same letter is 0.

If you just want to know the answers at this stage, simply scroll down. If you're not ready yet then here's one more clue to make things a lot easier:

Quordle today (game #1310) - hint #5 - starting letters (2)What letters do today's Quordle answers start with?

• B

• W

• H

• I

Right, the answers are below, so DO NOT SCROLL ANY FURTHER IF YOU DON'T WANT TO SEE THEM.

Quordle today (game #1310) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle, game #1310, are…

• BEGET
• WRATH
• HEARD
• INDEX

I had a nice trouble-free run today until I hit my last word.

I had four letters with one in the correct position, but was struggling to think of anything that would fit. It was by moving the letters around that I finally got INDEX, although it always feels risky using a rare letter.

Daily Sequence today (game #1310) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle Daily Sequence, game #1310, are…

• TWEET
• ADAPT
• OUGHT
• MYRRH

Quordle answers: The past 20

• Quordle #1309, Monday, 25 August: GEEKY, SHALT, CHIEF, JIFFY
• Quordle #1308, Sunday, 24 August: ROVER, GONER, ANTIC, OUTDO
• Quordle #1307, Saturday, 23 August: DEMON, GRATE, FLYER, SHEEP
• Quordle #1306, Friday, 22 August: TROOP, SCOPE, TORSO, BRINY
• Quordle #1305, Thursday, 21 August: QUEST, SPARK, WHITE, ACUTE
• Quordle #1304, Wednesday, 20 August: DOLLY, MERRY, BUGLE, WORST
• Quordle #1303, Tuesday, 19 August: KNAVE, SMART, CARRY, MAMMA
• Quordle #1302, Monday, 18 August: FIBER, TRADE, RAYON, TEASE
• Quordle #1301, Sunday, 17 August: FUNGI, AMITY, DRIER, CHECK
• Quordle #1300, Saturday, 16 August: OWING, QUAKE, SLIDE, ELITE
• Quordle #1299, Friday, 15 August: WHALE, PRISM, DRAKE, TEPEE
• Quordle #1298, Thursday, 14 August: LAPEL, IDIOM, RENEW, LIVER
• Quordle #1297, Wednesday, 13 August: CACTI, HOMER, EMAIL, ALBUM
• Quordle #1296, Tuesday, 12 August: SPOOL, TITLE, JAUNT, OVINE
• Quordle #1295, Monday, 11 August: ADULT, BROOM, PURER, CRUEL
• Quordle #1294, Sunday, 10 August: SCRUM, PIPER, TROLL, SPORE
• Quordle #1293, Saturday, 9 August: NOOSE, INLET, ELEGY, VIRUS
• Quordle #1292, Friday, 8 August: KNEEL, KINKY, RALPH, BOOZY
• Quordle #1291, Thursday, 7 August: PLUNK, PROXY, CURVY, PEARL
• Quordle #1290, Wednesday, 6 August: RISKY, APART, FAUNA, HANDY

• Gadhouse launched retro-inspired USB-C Wesley headphones: $68 / £59
• The launch comes days after Google called wires 'sloppy' and 'embarrassing'
• Wireless buds typically last 2 years; I own wired cans that are 15 years old

Sometimes a product just launches at a great time, doesn't it? A pertinent time; an important time.

Now, I could fashion a perfectly worthy news piece just about the Gadhouse Wesley. They are, after all, beautiful new retro-inspired headphones that wouldn't look out of place over Max's ears in Stranger Things (or, in fact, in any show set in the 80s).

And thanks to their lovely USB-C wire, they can offer 48kHz Lossless Audio – which is something even the best Bluetooth codecs within the very best wireless earbuds still can't truly do, but that the best wired earbuds and wired headphones certainly can.

You could also get that level of audio detail using your AirPods Max wired up with their USB-C wire, of course, but Apple's 2020 cans still cost a lot of money – $549 or £499 for quick reference. And the Gadhouse Wesley cost just $68 / £59 (or around AU$123). I love the retro-meets-now look of these headphones, and I'll certainly be buying them when they arrive in September, to partner with my Android.

But this is about something much bigger, for me. In fact, it's about Big Tech and what Google said at its Made By Google Pixel event on Wednesday, August 20. In case you missed it, I'll link the two-minute sponsored segment from Kareem Rahma's popular talk show and podcast Subway Takes, which Google cut away to during its event, here…

Wired is ‘sloppy’? Surely li-ion batteries in buds are sloppier

Where to start? One could take issue with the notion that wired headphones look "sloppy" or "embarrassing" (although the two wires from those mics to get decent voice pickup look absolutely fine, though, I'm guessing?) or the blunt statement that "they don't do anything that you want them to do". Actually, mine play hi-res music with virtually no latency, and that is exactly what I want, thanks.

Or there's the slightly more sinister notion that if your daughter started wearing wired headphones, she'd be "kicked out of the house… until she got some buds in".

I struggle to believe that, in 2025, we're advocating throwing away battery-free headphones that might last your child long into adulthood in favor of wireless earbuds that only last as long as the lith-ion battery squirrelled into them. Let me be clear: however good the audio quality and feature-set is, this battery will gradually deteriorate to the point of failure within around three years from the time of purchase, depending on use.

And you can state that earbuds are recyclable all you want (Apple absolutely does affirm this, with its AirPods), the truth is that many buds will still end up in landfill, because people don't always recycle – and with such tiny components, it is difficult to recycle every small component of a set of earbuds. It can also become expensive, since it requires actual human disassembly.

(Image credit: Gadhouse)

OK, so Google has announced that its new Pixel Buds 2a do have a replaceable battery in the case, but it's the batteries in the buds that tend to go first. And let's not forget, this is a company that has made a huge and much-touted pledge to achieve carbon neutrality by 2030. So this clip? Not a good look, Google.

I have been a full-time audio writer since 2019, and I still own three pairs of wired IEMs from that first, glorious year when I decided this was my future and forte. How many pairs of wireless earbuds do I still own and use from that time, six years ago? I'll give you one guess.

You may also like

• 4 reasons why the Google Pixel Buds 2a could be the $129 earbuds to beat in 2025 – and one key area in which they disappoint
• A gentle reminder to clean your earbuds and headphones: study reveals they may carry 10x more bacteria than a toilet seat
• See TechRadar's pick of the best over-ear headphones

• Ghost guns highlight the darker side of modern 3D printing technology
• Fingerprinting systems can embed timestamps, geolocations, and printer identification data
• SIDE offers break-resilient embedding methods for forensic data recovery

The rise of 3D printing has offered new ways to produce tools, spare parts, and even art. Yet its darker use has gained attention through “ghost guns.”

These are firearms produced outside regulatory systems, which are difficult to trace and have already been implicated in violent crime, including the 2024 killing of UnitedHealthcare CEO Brian Thompson.

Thankfully, researchers now believe that even if someone smashes such an object, investigators could still recover valuable forensic information.

Building a system of hidden fingerprints

The idea of fingerprinting in 3D printing is not entirely new. Various approaches embed unique information into printed items, often including details like timestamps, printer identification, and geolocation data.

Until now, a weakness has remained: what happens when a printed object is broken into pieces or deliberately tampered with?

A team led by Netanel Raviv at Washington University in St. Louis has developed a system designed to withstand such attempts.

Its framework, named Secure Information Embedding and Extraction (SIDE), uses mathematical methods that allow fragments to retain enough encoded detail for investigators to recover key information.

The SIDE system is built on earlier work presented at the IEEE International Symposium on Information Theory in 2024, which focused on encoding methods resilient enough to reconstruct data from partial remains.

By extending those ideas, the current study adds security mechanisms that can require 3D printers to include such fingerprints automatically.

According to the researchers, SIDE does not just add markings but uses loss-tolerant embedding techniques. This means that even a weapon deliberately broken into pieces may still yield identifiable data during forensic analysis.

On one hand, fingerprinting techniques could make it harder for illicit weapons to circulate undetected.

On the other hand, this raises questions about the scope of surveillance, control over personal manufacturing, and the balance between security and privacy.

A beginner 3D printer used for harmless home projects might soon be subject to the same requirements as industrial machines if regulators decide to mandate such systems.

While the debate will likely focus on ghost guns, SIDE has potential uses far beyond weapon tracing.

Any item produced with a 3D printer could, in theory, carry this invisible signature.

Combined with AI tools for pattern recognition and data recovery, investigators could reconstruct timelines and identities from fragments of plastic or resin.

"This work opens up new ways to protect the public from the harmful aspects of 3D printing through a combination of mathematical contributions and new security mechanisms," said Netanel Raviv, assistant professor of computer science & engineering in the McKelvey School of Engineering at Washington University in St. Louis.

"While SIDE has limitations in defending against resourceful attackers with strong expertise in 3D printing, it raises the level of sophistication, prior knowledge, and expertise required from the adversary to remain undetected after committing the crime."

Via Techxplore

You might also like

• These are the best small business printers to keep your SMB going
• Here are the best printers available
• We've also listed the best business VPNs to use now

• The AirPods Pro 3 are predicted to launch this year
• They could come with heart rate monitors
• It's been three years since the Apple AirPods Pro 2 launched

There have been a lot of conflicting reports around whether Apple would get around to launching the AirPods Pro 3 this year, but one of the more reliable reporters on the Apple beat says the wireless earbuds are indeed coming before the end of 2025.

That reporter is Mark Gurman at Bloomberg, who says the AirPods Pro 3 will also come with a significant upgrade: built-in heart rate monitoring. The same tech has previously been integrated into the Powerbeats Pro 2 which launched in February.

With the Powerbeats Pro 2, LED optical sensors pulse at over 100 times per second, measuring heart rate via blood flow. Presumably the AirPods Pro 3 will work in a similar way, and feed the stats back to the Apple Health app.

Gurman also reiterates his previous assertion that Apple is working on a paid Health+ subscription option for users, which may include the services of an AI fitness coach. The idea is you would get personalized advice on diet, exercise, and more.

Three years later

The AirPods 4 launched in September 2024 (Image credit: Future/Lance Ulanoff)

We've been waiting a long time for the AirPods Pro 3 to arrive, considering the Apple AirPods Pro 2 made their debut back in September 2022. The wireless earbuds market has changed quite a bit in that time – and we've since had the Apple AirPods 4, of course.

Most of the leaks and rumors we've heard around these earbuds are concerned with when they'll actually see the light of day. There have been hints in iOS 26 that the AirPods Pro 3 would arrive in 2025, though one reliable source has suggested a 2026 launch.

It seems clear that Apple wants health tracking to be a primary focus of the AirPods Pro 3, besides getting high-quality audio into your ears, and we may see the wireless earbuds able to track other vital statistics besides your heart rate.

The wireless earbuds are one of several devices "in the pipeline" for Apple "as early as this fall", which is typically September, October, and November. We're also of course expecting the iPhone 17 and the Apple Watch 11 sometime next month.

You might also like

• 5 features the AirPods Pro 3 should steal
• Here are the best wireless earbuds
• Apple code hints at AirPods Pro 3 launch

• A Microsoft designer imagined a new Copilot wearable without a display and HUD
• The Copilot Veja concept transmits information entirely through audio feedback
• Dual cameras give the wearable stereoscopic vision and potential depth awareness

Microsoft designer Braz de Pina has created an unofficial concept for a Copilot-powered wearable that reimagines what a successor to the Hololens headset could be.

Called the Copilot Veja, the device is not a headset but instead takes the form of ear-worn stems equipped with cameras, microphones, and buttons for interaction.

Although the name Veja comes from the Portuguese word for “see” (making the absence of any eye cameras appear especially odd), not having a heads up display was a deliberate decision. Most people already have screens in their pockets and on their wrists, and adding another was considered unnecessary by de Pina.

(Image credit: Braz de Pina )Audio-first

De Pina describes the Copilot Veja as a “fan-made concept”, designed around the idea of making Copilot portable, conversational, and practical without the need for a head-up display.

Instead of delivering information visually, the Veja uses audio as its main channel.

“I keep thinking about how AI has changed the course of the metaverse hype and the AR/VR universe. I know Meta is heavily investing in smart glasses and similar devices, not to mention all the exciting VR and AR headsets out there. But my question is: with capable agentic AI, do I really need to see what the AI tells me? Or is it enough to just hear it,” de Pina asked in a post on Instagram.

"In that case, glasses wouldn’t be necessary," he continued. "We’d simply be fixing one of AI’s flaws: the inability to read context in real time. An AI that can see what you see and provide instant feedback about your surroundings would allow for quick, natural communication. That’s what Copilot Veja is about.”

The ear stems are designed with comfort in mind. Each one carries a set of controls, including a power button, a volume knob, a Copilot button, and a camera trigger.

De Pina imagined the dual cameras working together to replicate stereoscopic vision, giving the AI the ability to analyze the world in depth while also allowing the user to capture content.

Although this isn't something that Microsoft would ever likely make, it's a great idea and one we'd love to see go beyond being a mere concept.

You might also like

• Farewell HoloLens - Microsoft officially quits wearable headset technology
• No escape from AI now - Microsoft is shoving Copilot into every Excel cell
• Hands on: Copilot Pro for Excel review

The third-party controller market shows no sign of slowing down, with innovative new models continuing to release at a rapid pace, promising more features and improved performance with each subsequent release.

Mobapad and GameSir are well-known in the space for their mid-range offerings, but both brands have recently released high-end pads that could vie for the best PC controller crown.

Both occupy the same price point, and both are compatible with multiple platforms (albeit with different consoles supported). They also offer similar levels of performance – but after testing both side-by-side, there’s one clear winner for me.

Design and features

(Image credit: Future)

The Huben 2 made a great first impression before I’d even seen it: its hard-shell case was sturdier and more premium than I was expecting, with its insides perfectly modeled to the contours of the controller, spare D-pad hat, and USB dongle. It also has a netted area to store the USB cable, complete with an integrated hook-and-loop fastener to keep it coiled in place.

The controller itself lives up to its high quality packaging. It looks very sleek and smart with its transparent faceplate revealing the brilliant white colorway of my review unit, which is nicely accented by the red sticks and face buttons.

All of its materials look and feel premium, and the overall build quality is excellent, with the moving parts being smooth to operate yet solidly fitted. The material behind the handles is textured to offer grip, although it’s a slight shame it’s still made from a hard plastic material.

The G7 Pro has a slightly smaller form than the Huben 2, so those with smaller hands might find it preferable to hold. Build quality is similar to that of Huben 2, although it’s perhaps not quite as sharp looking or as premium.

Nevertheless, it still feels very nice in the hand, and it has better grips, thanks to the soft-touch material on their undersides that provide more grip than the Huben 2’s. The texture and even the color are very similar on both pads, though.

There’s no included case for the G7 Pro, but it does come with a charging dock, which is made to a similarly high standard and fits the pad seamlessly, offering plenty of security while being easy to set and release. It’s also pleasingly slight, so it takes up minimal space.

Both controllers adopt the familiar Xbox controller style, although there are some differences in their features. The G7 Pro has a headset jack, the Huben 2 doesn’t. But the Huben 2 has four back buttons, while the G7 Pro only has two. However, the G7 Pro has a further two extra buttons nestled between the triggers and bumpers.

(Image credit: Future)

This is a choice location for them, as they’re very easy to use – easier than using back buttons. There’s also less chance of accidentally triggering them when gripping the pad, as I often did with the Huben 2’s back buttons.

What’s more, the two back buttons of the G7 Pro have an adjacent slider to disable them, which is far more convenient than having to use its accompanying software, as is the case with the Huben 2.

Speaking of which, the software for both gamepads offers plenty of customization options, beyond those accessible on-board. Both have mobile and Windows apps, although in the case of GameSir, its app for the latter system (GameSir Nexus) is far more capable than its app for the former.

The Mobapad mobile app, on the other hand, does feature plenty of tweaking potential to please serious tinkerers. As with Nexus, it allows buttons to be remapped and stick deadzones and curve profiles to be altered.

Other adjustments for motion controls and Turbo functions are also possible in both Nexus and the Moboapad app, although the latter offers slightly more depth. In reality, though, you won’t find many other software that offer as many customizations as these two.

Both are also easy to use, although it’s a shame the Huben 2 itself can’t be used to navigate the Mobapad app – even though it can be used to navigate your phone’s UI.

While both controllers have Bluetooth and USB dongles for multiple wireless connections, there are some differences in the platforms they support. Both work with PC and Android devices, but the Huben 2 has added iOS compatibility. And as for consoles, the G7 Pro is compatible with the Xbox Series X and Series S and the Xbox One, while the Huben 2 works with Switch.

Performance

(Image credit: Future)

In action, the Huben 2 really shines. The face buttons are clicky, lacking the heavily damped feel of many other controllers, but still just as tactile; they’re more akin to the clicks of the best gaming mice. They’re also very snappy and responsive as a result.

Even better is the D-pad: it feels very similar to the face buttons, and is just as quick and easy to use. Rapid, multiple inputs were a cinch in Tekken 8, while maintaining high levels of accuracy. There’s a small amount of dampening in each direction, which helps to make presses more controlled.

The default hat also helped in this regard. It adopts a very similar design to that of the latest Xbox controller, with its prominent middle dip and clearly squared-off segments, which provide plenty of feedback and stability.

The D-pad on the G7 Pro is, if anything, even easier and more clicky in its actuation than the Huben 2’s. However, it lacks any discernable dampening, which can result in less control over presses. I also found the hat too small and flat to provide sufficient grip, resulting in my thumb frequently falling off its left edge.

Both controllers have similar shoulder buttons, adopting the familiar shape and placement of those adorning the best Xbox controllers. I did find the bumpers of the Huben 2 a tad too high for my liking, but this is a minor gripe.

Both sets of triggers offer sufficient travel and feel, although those on the G7 Pro offer more resistance and have greater dampening. You might think this would make them easier to control than the Huben 2’s, but the latter’s more rounded shape compensates for their light actuation.

(Image credit: Future)

This is because a larger portion of your fingers is in contact with the physical buttons, which in turn allows you to control their depression more easily, at least in my experience. They’re also more comfortable as a result of this shape, allowing me to play with the Huben 2 for longer periods without fatigue.

I certainly welcomed this aspect when playing games requiring constant use of the triggers, such Assetto Corsa and titles requiring vehicle control. However, all of this isn’t to say that the triggers of the G7 Pro are poor, as they’re still perfectly enjoyable to use.

Each trigger also has its own travel lock switch next to it, allowing you to switch it from analog to digital modes, which is useful in the case of shooters where quick taps are required.

Both controllers also feature motion controls. The Huben 2’s work excellently on Switch: when I aimed bows in The Legend of Zelda: Breath of the Wild, their tracking and precision rivalled that of the best Nintendo Switch controllers. I experienced no drop outs or lag, either.

Since the G7 Pro isn’t compatible with the console, its motion controls can only be used on PC as a substitute for the sticks or even mouse inputs (something the Huben 2 is capable of, too). While they’re accurate enough, few PC games were designed with these controls in mind, so they don’t work as seamlessly. This isn’t exactly the G7 Pro’s fault, but it does beg the question as to why it has gyro capabilities if it’s not compatible with the platform that makes best use of them.

When it comes to sticks, I prefer the smoother, softer material on the Huben 2’s, as well as their deeper indentations which offer more purchase, and overall feel more satisfying to use than the G7 Pro’s.

It’s not that their hall effect technology seems more accurate than the TMR sensors in the G7 Pro’s sticks, but they feel more responsive as well as being easier to wrangle, moving in precisely the way you want them to.

As for connectivity, both controllers worked with their respective devices without issue during my testing. Switching between devices is also easy, since both feature sliders on the rear for changing platforms, although it can take a few moments and press of the power button to connect to the new device.

Verdict

(Image credit: Future)

Both the Huben 2 and the G7 Pro are great controllers. They’re expensive – and there are plenty of great budget alternatives out there (including other models from the two brands) – but if you’re looking for something a little more upmarket, neither is likely to disappoint.

However, I think the Huben 2 has the slight edge, feeling just that little more premium and satisfying to use, with the D-pad deserving special mention. A large part of your decision may come down to the specific platform support, but if you’re playing on PC, Android, or Switch – or all three – the Huben 2 is the superior choice as far as I’m concerned.

You might also like

• Don't buy a DualSense Edge if you're after a premium PS5 controller; the Victrix Pro BFG Reloaded is better in almost every way
• This year has been phenomenal for controllers, and these five are the best I've personally tested and reviewed in 2025 so far
• I’ve played EA Sports FC 26, and I’m ashamed to say it’s not the gameplay that’s got me excited

• Windows 11 has a Resume feature in test builds
• For now it's just for Spotify, but more developers will get on board in time
• If you're listening to a track on your Android phone, you can resume that playback on your Windows 11 PC seamlessly

Windows 11 is getting a new ability whereby the desktop operating system will allow you to resume what you were previously doing in an app on your Android phone.

This is only available in testing for now - in Windows 11 preview builds in the Dev and Beta channels - and while it will eventually come to a range of apps, to begin with it only works with Spotify.

As Microsoft explains in a blog post, the way the feature works is as follows: say you're listening to a playlist or podcast on Spotify on your smartphone, and you stop it, then head to your desktop PC.

When you fire up Windows 11, you'll see a ‘Resume alert’ pop up from your taskbar at the bottom of the screen, letting you know that you can continue your Spotify listening on the PC. Just click on that alert, and the Spotify app will be opened, continuing playback exactly where you left off on your Android smartphone.

Microsoft reminds us that we'll need to have the Spotify app installed on the PC and be logged into the same account as the one running on the smartphone, as you'd expect. If you haven't got Spotify installed in Windows 11, a button to install the software from the Microsoft Store will pop up - then you just have to log in to your Spotify account.

Also, you must have the Link to Windows app running on your smartphone, and have access to your mobile enabled on your Windows 11 PC, as per Microsoft's instructions in its blog post.

(Image credit: Microsoft)Analysis: A great addition - but Microsoft's really taken its time

This is essentially Microsoft's take on Apple's Handoff feature, and it looks to be a streamlined and useful ability that I'm keen to see debut on Windows 11. Its revelation is no surprise, though, seeing as Microsoft accidentally gave us a glimpse of the functionality in a pre-recorded Build session earlier this year, the video of which was swiftly edited to remove this. It worked pretty much exactly the same as we see in testing now (with Spotify also being used in this early demo).

The technical name for this trick is Cross Device Resume and hopefully a number of software developers will be getting on board with bringing in support for this - those who have apps available on both Windows 11 and Android, of course. Microsoft will doubtless be prodding some of the bigger names out there, and besides Spotify we've also heard that WhatsApp might be in line for this treatment.

Clearly, the Resume feature is going to try to push installations from the Microsoft Store as well, which won't hurt traffic there - something the company has been working on improving of late.

While resuming Android apps on Windows 11 is clearly a great feature, and one that it's good to see inbound, the main question here is why has it taken so long for Microsoft to implement this?

You might also like

• Latest Windows 11 update packs a great new feature - but I hope I never have to use it
• Microsoft promises to crack one of the biggest problems with Windows 11: slow performance
• Can't reset your Windows 11 or 10 PC? Microsoft rushes out fix after breaking vital recovery features

• Some key details of the folding iPhone have leaked
• Expect four cameras, two colors, and Touch ID
• We'll probably see the handset launch in 2026

At this point it seems almost certain that Apple is going to bring out a foldable iPhone in September 2026, and we just got a few more details about what to expect from an Apple tipster who is usually more reliable than most.

That tipster is Mark Gurman at Bloomberg, who says that the folding Apple phone is going to use Touch ID rather than Face ID. While all the current iPhones now use Face ID, the thinking is that the required sensors would add too much thickness to the foldable.

Apple is therefore going to go with Touch ID to keep the device as thin as possible, Gurman says, and there'll be no SIM card slot for the same reason – this is going to be eSIM only. Previous leaks have predicted the phone will be 4.8 mm thick when unfolded, compared to the 4.2 mm of the recently launched Samsung Galaxy Z Fold 7.

This latest report also states we're going to get four cameras on the folding iPhone: one on the outer screen, one on the inner screen, and two on the back (which will both be 48MP cameras, according to an earlier rumor).

Creasing up

The Apple iPhone 16 is about to get a successor (Image credit: Future)

If past reports are to be believed, one of the reasons a foldable iPhone has taken so long is because Apple is very keen to make the screen crease as invisible as possible, and it seems the company now has a display setup it's happy with.

As per Gurman, the folding iPhone is going to use what's known as an in-cell touchscreen, with an integrated display and digitizer layer. That would make the screen tech similar to what we have on the current iPhone models, ensuring a "less obvious" crease and improving touch accuracy.

Other tidbits from the report: the folding iPhone will be available in black or white, and will include Apple's own C2 modem chip that's expected to debut in the iPhone 17 (and which should lead to improved efficiency and battery life).

Gurman says the foldable iPhone will be a "game changer" for Apple, and his belief is that anyone who makes the switch to a folding iPhone won't want to go back. First though, we're got this year's iPhone 17 launch coming up next month.

You might also like

• Apple is reportedly working on a folding iPad too
• These are the best foldables you can buy right now
• We may see a camera tweak on the folding iPhone

• A new phishing scheme successfully bypasses most security tools
• It abuses ads and Microsoft's Active Directory Federation Services tool
• It is designed to steal login credentials, so users should take care

Cybercriminals have found a clever way to make phishing sites look like legitimate login pages, successfully stealing Microsoft credentials, experts have warned.

Cybersecurity researchers at Push Security recently published an in-depth report on how the scam works, outlining how the attackers created fake login pages that mimicked authentic Microsoft 365 sign-in screens.

Then, instead of sending victims directly to the site, which would probably get flagged by security solutions and quickly blocked, they used a Microsoft feature called Active Directory Federation Services (ADFS). Companies normally use it to connect their internal systems to Microsoft services.

How to stay safe

By setting up their own Microsoft account, and configuring it with ADFS, Microsoft’s service is tricked to redirect users to the phishing site, while making the link look legitimate because it starts with something like ‘outlook.office.com’.

Furthermore, the phishing link was not being distributed by email, but rather - malvertising. Victims were searching for “Office 265” which was presumably a typo, and were then taken to an Office login page. The ad also used a fake travel blog - bluegraintours[.]com - as a middle step to hide the attack.

The way the entire campaign was set up made it particularly dangerous. With the link looking like it was coming from Microsoft, and it successfully bypassing many security tools checking for bad links - its success rate was probably higher compared to “traditional” phishing.

Furthermore, since it doesn’t rely on email, the usual email filters couldn’t catch it. Finally, the landing page could even bypass multi-factor authentication (MFA), which made it even more dangerous.

In order to prevent such scams from causing any real harm, IT teams should block ads, or at least monitor ad traffic, and watch for redirects from MIcrosoft login pages to unknown domains.

Finally, users should be careful when typing in search terms - a simple typo can lead to a fake ad that can result in device compromise and account takeover.

Via BleepingComputer

You might also like

• The growing threat of device code phishing and how to defend against It
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

• Elipson launches Horus 6B bookshelf and Horus 11F active floorstander speakers
• 50W stereo amplification, aptX HD, USB Audio and HDMI ARC
• £399 (about $530 / AU$840) and £899 (about $1,200 / AU$1,880)

French luxury audio brand Elipson has announced two new active speakers in its Horus range: the Horus 6B Active BT bookshelf speaker and the Horus 11F Active BT floorstander.

Both sets of speakers have built-in 50W stereo amplification, a "sophisticated" bass reflex system, 130mm glass fiber-reinforced cellulose pulp bass/mid-range drivers, and 25mm silk dome neodymium tweeters, and they're available in three finishes: light wood / beige, walnut / dark gray, and black / carbon.

(Image credit: Elipson)Elipson Horus Active BT speakers: key specs and pricing

The Horus 6B Active speakers have a USB Audio connection for PC or Mac offering plug-and-play audio at up to 24-bit/96kHz, and they also have a moving magnet (MM) phono input for your turntable so you can connect it without also needing a pre-amp.

Both sets of speakers also have HDMI ARC for a TV, coaxial and optical inputs (24-bit/192kHz) and Bluetooth 5.0 with aptX HD.

Frequency response is 40Hz to 20,000Hz for the floorstanding Horus 11F Active BT and 55Hz to 22,000Hz for the bookshelf Horus 6B Active BT.

The bookshelf speakers are available now and the floorstanders will go on sale in early October 2025. The recommended retail price is £399 (about $530 / AU$840) for the Horus 6B Active BT and £899 (about $1,200 / AU$1,880) for the Horus 11F Active BT.

You might also like

• Elipson's new compact streaming amp looks like a brick, but kicks like a horse
• Best stereo speakers 2025: stereo systems for all budgets
• Audio-Technica's cheap active bookshelf speakers look ideal for new vinyl fans

We are at a tipping point for quantum computing, which is on the verge of becoming a reality. While its potential is tantalizing, it also represents an unprecedented threat to the traditional data security infrastructure and the cryptographic algorithms that protect it. Post-quantum cryptography – algorithms designed to be secure against classical and quantum computer attacks – is the response.

Quantum computers exploit the principles of quantum mechanics to solve complex problems that classic computers cannot feasibly tackle. Quantum computers use the principles of quantum mechanics to process information in a way that uses qubits, which can exist in multiple states, as opposed to normal computers, which only use "zeros" and "ones". This creates an exponential scale, which is what gives them their computational power.

Of particular concern is their ability to crack widely used public key encryption algorithms such as RSA and ECC (elliptic curve cryptography). By the time a sufficiently powerful quantum computer becomes available, these encryption methods, which protect virtually all current digital communications, will be obsolete.

The date when cryptographically-relevant quantum computers will appear remains uncertain: estimates range from five to 10 years. However, the risk is immediate due to the "harvest now, decrypt later" attacks that are already taking place, especially for data with a longer lifetime.

If an organization retains sensitive data for the long term, such as financial information, personal data or even trade secrets, this represents a significant and growing risk.

What is at stake is nothing less than the most valuable digital assets: intellectual property, private and sensitive data, authentication systems and secure communications.

The financial, operational and reputational damage from such exposures could be catastrophic, and unavoidable without proactive measures.

What is post-quantum cryptography?

Post-quantum cryptography (PQC) refers to cryptographic algorithms designed to be secure against attacks by classical and quantum computers. These algorithms are based on mathematical problems that remain difficult to solve even for quantum computers.

In 2024, the National Institute of Standards and Technology (NIST) published its first set of standardized post-quantum cryptographic algorithms, including CRYSTALS-Kyber, CRYSTALS-Dilithium, SPHINCS+ and FALCON. In March 2025, NIST selected a new algorithm, Hamming Quasi-Cyclic (HQC), which will serve as a backup to the existing Module-Lattice-Based Key-Encapsulation Mechanism (ML-KEM) algorithms recommended by FIPS 203 to protect against quantum attacks.

HQC is based on error-correcting codes, a concept that has been fundamental to information security for decades. Unlike ML-KEM, which relies on structured networks, HQC's unique mathematical basis offers a robust alternative that can help combat the potential threats posed by future quantum computers. This shift in mathematical approaches is crucial for maintaining the integrity of encrypted data.

The time for change is now

The time when currently encrypted data can be decrypted using quantum technology is closer than many people think. However, while most organizations are actively working on cyber resilience strategies, including their core IT infrastructure and components of the supply chain, the risk to quantum computing is not as widely considered.

Changing cryptography in a complex IT environment is not something that can be done overnight. It can take years, especially for large organizations with complex IT environments. Historical precedent shows that major cryptographic transitions typically take 5-10 years to complete.

Beginning the transition

To begin a transition to post-quantum cryptography, a number of steps must be followed:

1. Cryptographic inventory: Not all data is equally important, and not all data needs to be encrypted in the same way. It is therefore necessary to identify where cryptography should be used in the digital heritage. This should include the most sensitive data, applications, networks, identity systems, and third-party connections.

2. Risk assessment: Given the cost of post-quantum cryptography, it makes sense to prioritize protecting the most sensitive data rather than trying to protect everything. Evaluate your data in terms of its sensitivity and longevity. Information that must remain confidential for more than five years should receive immediate attention. For less sensitive data, standard encryption methods will suffice in keeping it secure.

3. Crypto-agility implementation: Being crypto-agile – having the ability to switch between different cryptographic algorithms in response to new threats – will be essential in the post-quantum era. Develop frameworks that allow you to quickly replace cryptographic algorithms without the need for extensive system redesign. Crypto-agility also requires employee training, so invest time and resources to bring your employees on this journey with you.

4. Prioritized migration: Start with your most sensitive systems and data, particularly those that protect intellectual property or personally identifiable information.

5. Supplier engagement: Confirm that all suppliers in your ecosystem are aligned with emerging standards to ensure end-to-end protection and agility.

In summary

By starting your post-quantum transition today, you can help protect your organization's most valuable data as we enter the quantum era. The alternative is to wait for quantum computers to break existing encryption – by then, it will be too late for data that has already been compromised. The future is quantum and the time to future-proof your data is now.

We list the best performance 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

• Surfshark has conducted a review of location-hungry social media apps, including X, Instagram, and Snapchat
• Of the 10 social media apps analyzed, X demands the most granular location data
• Data was collected in August 2025 from the Apple App Store

Think you’ve got your privacy covered? It could be time to give that a rethink, especially if you’re a regular user of X and other social media apps.

A study from leading VPN provider Surfshark has analyzed the default location permissions required by the top 10 social media apps in the Apple App Store and found that most of them are very data-hungry – if not stalker-y.

Elon Musk’s X and Mark Zuckerberg’s Instagram, Threads, and Facebook especially know a lot about where you are… even if you’re using one of the best VPN apps.

Why do social media want your location data?

Your smartphone records two types of location data: precise and coarse.

Coarse data is a zoomed-out, low-resolution option that shares your general location in the world. The precise option relies on GPS, Bluetooth, and even IP addresses to determine your exact location based on nearby networks. It is the latter data that X and other networks are using, to varying degrees.

As Surfshark Chief Technology Officer, Donatas Budvytis, explains, precise location data falls under the most sensitive user data category because it can be linked to highly personal behavior and routines.

"This all can lead to building your profile, predicting behavior, or can be used for targeted manipulation and discrimination," said Budvytis.

(Image credit: Surfshark)

Location data might be helpful at times. For example, it adds some useful context to photos (and helps you find a specific image years later). It can also help you locate interesting sights and businesses nearby. Social networks, however, seem to be enjoying just a little bit too much of it.

Surfshark has discovered that while eight popular social networks collect precise location data, X is especially keen on it. Instagram, Threads, and Facebook all need a lot, as does Pinterest. Snapchat needs less; LinkedIn and YouTube require almost none.

Collecting user information in this way allows social networks to offer advertisers a better reach, ensuring targeted ads based on your location. They might even send you marketing outreach messages using this sort of data.

According to researchers, X and Pinterest may even use location data to track users. There is no clarity as to how this information is used or whether it is available to data brokers. In such a scenario, various information from other apps could be used in tandem to establish more specific and valuable data about your actions and interests.

Can’t you just use a VPN?

A virtual private network (VPN) is security software that encrypts your internet connections, while spoofing your real IP address location. (Image credit: Getty Images)

VPNs are great for specific privacy uses, but when it comes to location data collected using GPS and other networks, a VPN is less effective.

Because the VPN encrypts your internet access and routes it via a secure VPN server, only your IP address can be obfuscated. Social networks that rely on GPS and proximity to other networks – perhaps railway stations, or city centre networks – can overlook the notion that your IP address is set to Berlin if your GPS says London.

While a VPN is definitely recommended for public Wi-Fi, it’s only going to prevent bad actors with access to the network. Elon Musk and co can still record your location.

So, can you stop X from tracking you?

All social media apps have settings that you can use to tweak what information is collected. So, while you might be okay sharing photos and (potentially embarrassing) thoughts and opinions, you can at least control some of these privacy options.

In the case of X, you can open Settings > Privacy and safety > Location information and disable all three options (tap the third, Explore settings, to toggle it off) to reduce the social network’s hunger for your activities.

Other apps should have similar options. You can, of course, also disable your phone’s GPS when it isn’t required, and review your phone’s location services settings.

You might also like

• International VPN Day: a privacy solution or "a loophole that needs closing" – here's what's at stake
• Meta halts phone and browser tracking tools after researchers expose user snooping
• Is 10,000 steps a day worth your personal data? How 80% of fitness apps are selling your privacy

The Chief Information Security Officer (CISO) has rapidly evolved from a technical specialist into a strategic business leader. While still responsible for defending against cyber threats, today’s CISOs are also tasked with aligning cybersecurity initiatives to broader organizational goals. This shift is driven by a more complex threat landscape, advances in technology such as AI, and increasing regulatory pressure that holds security leaders to a higher standard of accountability.

As a result, CISOs must bridge the gap between security operations and business strategy. They are expected to translate technical risks into business impacts, engage with executive leadership, and help shape enterprise resilience.

For security and business leaders alike, this evolution highlights the critical need to embed cybersecurity into the fabric of strategic planning and decision-making for the business.

The Changing Threat and Regulatory Landscape

Cyber threats are growing in scale and sophistication. Attackers can and do use AI tools to accelerate reconnaissance, craft convincing phishing schemes, and execute ransomware at unprecedented speed. Deepfakes, autonomous malware, and AI-driven social engineering are pushing traditional defenses to their limits.

At the same time, regulatory scrutiny has intensified. Frameworks like the US Security and Exchange Commission's cybersecurity disclosure rules and the European Union’s Network and Information Security Directive 2 (NIS2) demand greater transparency, rapid incident reporting, and clearer communication of cybersecurity’s business impact. As a result, CISOs are more visible than ever—expected to brief boards, own cyber risk posture, and help ensure regulatory compliance.

Expanding Responsibilities of the Modern CISO

The role of the CISO has significantly broadened beyond traditional technical oversight. Today’s CISOs are now pivotal in securing expansive business ecosystems, encompassing cloud infrastructures, third-party vendor relationships, operational technologies, and software-based products. This expansion necessitates a unified approach to security and risk management, integrating these functions at the enterprise level. ​

In addition to these responsibilities, CISOs are instrumental in driving business efficiency and facilitating the adoption of emerging technologies. Today, the integration of AI tools and automation into security operations is transforming the cybersecurity landscape.

For instance, automation streamlines processes such as vulnerability management workflows, enabling quicker remediation, and AI enhances threat detection and response capabilities by rapidly analyzing vast datasets to identify potential risks, both of which allow security teams to focus on strategic initiatives.

As new technologies that have significant implications for business and cyber security continue to come to market - such as quantum computing and its impact on encryption - CISOs will need to stay on top of operational, business and strategic implications.

By embracing these advancing technologies, CISOs can scale their operations effectively, ensuring that human expertise is applied where it delivers the most significant impact. This evolution underscores the CISO's role as both a protector of information assets and a strategic enabler within the organization.

Bridging Security and Business Strategy

To succeed in this expanded role, CISOs must communicate risk in the language of the business. That means framing threats in terms of financial exposure, operational disruption, and reputational damage—helping executive teams make informed, risk-based decisions.

Regular board of directors engagement is now essential. CISOs must earn trust and ensure that cybersecurity is viewed not just as a compliance obligation, but as a strategic priority.

When embedded into business initiatives from the start, cybersecurity becomes a driver of innovation. It accelerates digital transformation, enables secure cloud adoption, and builds customer trust. Forward-thinking CISOs position security not as a cost center, but as a competitive advantage.

Challenges Facing CISOs in 2025 and Beyond

Despite progress, CISOs still face major roadblocks. Budgets are increasing modestly – often by 10% or less - while responsibilities grow rapidly. Security teams remain understaffed, contributing to burnout and resource constraints.

Culturally, many organizations still treat cybersecurity as a siloed IT function rather than a shared business responsibility. This lack of integration slows decision-making and limits effectiveness.

Meanwhile, legal and regulatory accountability is rising. CISOs are being named in lawsuits and enforcement actions, most notably in the SolarWinds breach, raising concerns about personal liability.

Risk-sharing mechanisms such as indemnification and Directors and Officers (D&O) liability insurance - a specialized form of coverage to protect individuals from personal losses if they are sued - are becoming essential safeguards.

As the CISO role continues to evolve, several trends will define the path forward:

AI as Standard: AI-driven detection and response are becoming baseline capabilities.

Quantum-Ready Security: Forward-looking CISOs need to begin preparing for post-quantum threats.

Board-Level Strategic Advisor: More CISOs will report directly to CEOs or boards.

Workforce Development: Organizations must invest in training to address persistent talent shortages.

Integrated Security Culture: Cybersecurity will be embedded into all business functions.

Legal Protections: More formal structures to protect CISOs from undue liability.

To stay ahead, CISOs must focus on strategic alignment, talent development, automation, and executive communication, ensuring that security becomes a shared, organization-wide priority.

From Defender to Business Enabler

The CISO’s transformation into a strategic advisor reflects a broader realization: cybersecurity is business-critical. As technology advances and threats intensify, organizations need leaders who can secure operations while enabling innovation and growth.

Security and business leaders must work in tandem investing in collaboration, communication, and continuous improvement to build resilient, forward-looking enterprises equipped for evolving risks and opportunities.

We list the best identity management solutions.

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

CRM software systems were once the digital backbone of business growth, but what worked in the past doesn’t always fit the pace and pressure of today. In 2025, companies aren’t just seeking digital tools, they’re demanding agility, speed, and autonomy. And that’s exactly where legacy CRM systems are falling short.

A new wave of AI-native, no-code platforms is reshaping the industry, moving CRM from a rigid, IT-owned tool into a dynamic, business-controlled engine.

According to a recent Nucleus report, organizations shifting away from legacy vendors such as Salesforce in favor of modern no-code platforms are seeing significant improvements, including implementation times reduced by up to 70%, a 37% drop in total cost of ownership, and lead response times cut by more than 60%.

Here are five reasons why legacy CRMs are fast becoming obsolete, and why no-code is proving to be a promising, modern alternative.

1. Legacy CRMs can’t keep up with today’s speed of business

Businesses today operate in a state of perpetual change – new markets, new regulations, and new customer expectations. Yet many traditional CRMs, born in a different era, demand lengthy implementation cycles and code-heavy customization just to support basic adaptations.

In contrast, modern systems like no-code platforms are built on composable architectures that prioritize speed, flexibility, and business user autonomy. This allows organizations to configure and scale solutions rapidly without getting held back by complex, developer-led customization.

In fact, we’re seeing companies across core industries like manufacturing and financial services increasingly pivot to no-code models to avoid delays and spiraling costs.

It’s a clear signal that businesses are no longer willing to tolerate multi-year timelines to achieve meaningful outcomes. Adaptability and speed are no longer nice-to-haves, they’re essential for staying competitive.

2. AI and automation are now table stakes

Businesses now rely on AI to deliver personalized experiences, make smarter decisions, and streamline workflows. Yet many legacy CRM systems treat AI as an afterthought – a feature layered on top rather than embedded into the core of the platform.

Modern platforms, by contrast, are AI-native from the ground up. They use machine learning to automate lead routing, predict customer behavior, and optimize campaigns in real time. Organizations embracing these capabilities are seeing substantial improvements, such as a 61% reduction in lead generation response times and higher overall conversion rates.

Automating repetitive and manual tasks assist in giving teams back valuable time to focus on what truly matters, such as strategic thinking, creative problem-solving, and driving business growth.

3. The cost of complexity is unsustainable

Legacy CRMs often come with hidden costs: high developer overheads, third-party consulting fees, and expensive integrations that require constant maintenance. The burden on IT is immense, and the total cost of ownership (TCO) continues to climb long after deployment.

No-code platforms significantly lower these expenses, with firms reporting up to a 70% reduction in development costs and average savings of over $300,000 on external consultancy fees.

Newer no-code platforms remove many of these barriers by enabling configuration and updates without specialized technical knowledge. As a result, organizations are seeing a meaningful reduction in total cost of ownership. ROI is scrutinized more than ever, and the economic case for moving away from legacy tools is increasingly compelling.

4. Business teams demand control

Legacy CRM systems were designed with IT departments in mind, often requiring developer intervention for even simple changes. This creates bottlenecks and slows down innovation, particularly for departments like sales and marketing that need to move quickly and iterate on processes in real time.

No-code platforms are reversing this dynamic by giving control directly to business users. In fact, the number of citizen developers - non-technical employees empowered by no-code tools to build applications - is expected to rise by 50% this year.

With drag-and-drop interfaces, visual workflow designers, and intuitive configuration tools, non-technical staff can build and refine processes without needing to go through IT. This decentralized control fosters agility and allows organizations to respond to market shifts much faster.

5. Unified platforms drive real business agility

Legacy CRMs often operate in silos - disconnected systems for sales, marketing, service, and operations that struggle to communicate with one another. This fragmented approach creates inconsistent experiences and significant operational drag.

Modern no-code CRMs break down these barriers by unifying all functions within a single, cohesive platform. Shared data models, integrated workflows, and real-time visibility empower teams to collaborate seamlessly, respond faster to customer needs, and drive consistent outcomes.

With AI embedded throughout, this unification is key to enabling true business agility - not just reacting quickly, but aligning seamlessly across departments to deliver smarter, more cohesive customer experiences.

The no-code future is already here

The rise of no-code platforms marks a turning point in enterprise software. Rather than relying on rigid, IT-managed systems that require months of development and a deep bench of engineers, businesses now have access to tools that are fast, flexible, and accessible to all.

For organizations still tied to legacy CRM systems, the question is no longer if change is coming, it’s how quickly they can catch up. No-code isn’t just a trend. It’s a response to the urgent need for speed, adaptability, and user empowerment in the business environment.

We list the best sales CRM.

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

A function that was once buried deep within IT departments, cyber security is now firmly making its way up the boardroom agenda. 72% of UK businesses now classify cyber security as a high priority, with that extending to 96% of large businesses.

As recent high-profile breaches at M&S, Co-op and Harrods have shown, cyber resilience is now central not only to operational integrity but also to brand value, regulatory compliance and investor confidence.

Greater awareness has emerged as businesses shift from short-term solutions adopted during the pandemic to long-term, strategic partnerships with specialist cyber security providers. Increasingly, organizations recognize that cyber security requires an integrated approach involving continuous monitoring and proactive risk management.

The growing complexity and specificity of cyber threats mean that a bespoke, tailored approach is necessary, driving demand for advisory-led solutions delivered by experienced, security-cleared professionals.

That shift in perception is now being reflected in dealmaking. In the second quarter of 2025, the UK cyber market saw a flurry of M&A activity, much of it led by private equity platforms executing bolt-on acquisitions. These transactions may not always grab headlines, but they are sending a clear signal - cyber security is a strategic growth priority.

A new kind of risk

The cyber threat landscape has evolved. Today’s attacks are more frequent, more sophisticated and more damaging. The recent incidents involving the “Scattered Spider” group are just the latest reminder of the long-term impact these attacks can have, beyond the legal and financial consequences, but to customer trust and brand reputation. That’s why boardrooms are starting to reassess their cyber readiness.

In sectors such as public services, infrastructure and education - where the risks of failure are especially high - strong cyber defenses are no longer optional.

In addition, the rapid advancement of AI will accelerate cybersecurity risks, as it lowers the barrier for executing sophisticated attacks and enables threat actors to automate, scale, and personalize their tactics with unprecedented precision.

Regulation is raising the stakes

At the same time, government regulation is putting company directors firmly on the hook. The UK’s proposed Cyber Security and Resilience Bill will make senior executives directly accountable for managing cyber risks and ensuring operational resilience, bringing the UK closer to European frameworks like the NIS2 Directive and DORA.

This is changing how cyber security is viewed at the top. It’s not just about ticking boxes or passing audits. It is now a central part of good governance. For investors, strong cyber capabilities are becoming a mark of well-run companies. For acquirers, it’s becoming a critical filter for M&A, particularly when dealing with businesses that hold sensitive data or operate critical systems.

This regulatory push is part of a broader global shift towards greater accountability. In response, businesses are increasingly adopting governance models that embed cyber risk management into their strategic decision-making processes. Boards that fail to adapt not only risk regulatory penalties but also stand to lose investor confidence and market competitiveness.

Private Equity steps in

While overall deal values are still below long-term trends, deal volumes are rising. In Q2 alone, there were 114 cyber-related deals across Europe and North America, well above average. In the UK, activity is particularly strong in the small to mid-sized market, with private equity firms at the forefront.

Cyber security is a highly fragmented mission critical sector with strong recurring revenues, sticky customer relationships and a compelling margin profile. In an environment where investors are increasingly focused on resilience over growth, these are attractive attributes.

From product to partnership

The post-Covid shakeout is also playing a role. Many companies quickly adopted off-the-shelf solutions during the pandemic to meet urgent needs. Today, with greater familiarity and a clearer understanding of risk, boards are opting for more tailored, enterprise-grade services.

This is not just about technology, there is a growing premium on advisory-led solutions. Highly qualified, security-cleared professionals providing bespoke assessments and continuous monitoring. In other words, clients want expertise and service, not just software.

From a valuation perspective, this matters. While public market multiples continue to fluctuate, exposed to macro shifts such as US tariff announcements earlier this year, premium valuations continue to cluster around providers with diversified offerings and deep client integration. As PE buyers weigh bolt-ons and platforms, these traits are driving acquisition rationale.

Players need to stay ahead

The forces pushing cyber up the corporate agenda aren’t going away. Threat actors are growing bolder, regulators are getting tougher and the risks remain high.

The result is a market in transition. What began as a compliance arms race is evolving into a sophisticated, services-led ecosystem. For dealmakers, this creates opportunity but also demands discernment. Not every cyber asset will command a premium.

The winners will be those with deep expertise, defensible margins and client relationships that extend beyond the server room.

We list the best client 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

Hyper-personalized AI is transforming the workplace. Unlike standard automation, it works by learning from individual user behaviors, allowing businesses to tailor interactions on a much more human level. Not only does it help businesses to streamline their operations, it also drives efficiency while enhancing the user experience.

For employees, AI can suggest ways to improve productivity, automate repetitive tasks and provide real-time insights based on their work habits. In contact centers, for example, it can escalate intuitively from an AI Agent to a call with a human if it's a nuanced or complex problem.

While in retail, AI-powered assistants make personalized recommendations and offer timely discounts based on past interactions, purchasing history and market trends, which can result in impromptu purchases. It allows customers to feel like they are being listened to – and that their previous purchases are appreciated.

And it is this increasingly natural and ‘human’ experience that is helping to redefine what a digital brand interaction looks and feels like. The impact cannot be ignored.

According to Gartner, businesses investing in hyper-personalization are experiencing an uptick of 16% in commercial outcomes. The ability of AI to refine and improve interactions in real-time makes it a valuable tool for business growth. But as AI becomes more embedded – as it becomes smarter and more human – there are concerns over the impact this might have on privacy and security.

Ensuring privacy and security

After all, the very nature of hyper-personalized AI presents a paradox. The more an AI system knows about an individual, the better its recommendations. But this has led to some people raising concerns over issues such as surveillance, consent and the potential misuse of data.

It’s one of the biggest question marks still hovering over the issue of continuous learning. And without proper governance, AI may retain sensitive information, increasing the risk of unauthorized access, data breaches and regulatory non-compliance.

Thankfully, high-profile regulations such as GDPR and California Consumer Privacy Act (CCPA) frameworks already impose strict rules on data handling, and businesses failing to comply face not just legal repercussions but also reputational damage. But, that hasn’t stopped some people from pursuing AI-specific legislation – such as the AI Act in the EU – to provide specific protection.

There’s also an ethical dimension. Poorly designed AI can inadvertently reinforce biases or expose personal details that should remain confidential. If employees and customers lose trust in AI systems, companies will struggle to gain the full benefits of hyper-personalization. To thrive, organizations need to harness the power of AI while ensuring that privacy isn’t compromised.

How businesses can balance AI innovation with privacy

The good news is that businesses don’t have to choose between AI-driven efficiency and data privacy – they can have both. The solution lies in embedding privacy-first, responsible AI principles into frameworks and strategies from the outset. Here’s how:

Anchor core, long-lasting principles: Build ethical, trustworthy AI systems by prioritizing transparency, inclusiveness, and ongoing monitoring to foster trust and drive lasting value.

Establish robust governance: Define clear policies, conduct risk assessments, and assign dedicated roles to ensure compliance and ethical AI practices.

Ensure data integrity: Use high-quality, unbiased data to deliver fair and accurate AI outcomes across all user groups.

Adhere to compliance needs: Proactively address tightening regulations with strong governance and data protection to mitigate legal risks.

Test and monitor consistently: Conduct regular testing and continuous monitoring to align AI with ethical standards and performance goals.

Optimize tools effectively: Leverage advanced platform features like retrieval mechanisms and feedback loops to enhance transparency and ethical behavior.

Human oversight and participation is also part of the puzzle in optimizing and building trust in AI. At key points in AI workflows, humans help ensure accuracy and reliability. For example, in Agentic Workflows, AI breaks tasks into smaller steps and handles repetitive work, while humans review important decisions before final actions are taken.

AI then continuously learns from human input, improving over time. This approach combines the speed and efficiency of AI with the judgment and experience of human workers to create a system that is not only faster but also more reliable and adaptable.

In other words, with the right safeguards – and the right leadership and employee engagement strategy to ensure these protocols are followed – businesses can unlock the full potential of hyper-personalized AI without compromising security or trust.

The future of AI and privacy: staying ahead of the curve

What’s more, businesses that integrate privacy-first thinking into their AI strategies are likely to be the ones that thrive in the long run. The key is to build a governance framework that not only meets regulatory standards but also fosters trust among employees and customers.

One of the first steps is ensuring AI tools are rigorously assessed before deployment. That means evaluating how data is being processed, stored and used. It would also help to run pilot programs to help iron out privacy concerns and identify risks before full-scale implementation.

From the beginning, you need to work with a trusted provider to define and configure algorithms to prevent unintended biases and ensure fairness across different user groups at every level, whether it's an LLM, agent or App. AI should also be designed to evolve responsibly, integrating smoothly into workflows while maintaining strong privacy protections.

Clear visibility

Observability and traceability are crucial. Not only should people have clear visibility into how AI makes decisions, they should also be able to challenge or verify AI-generated outputs with real-time tracing, explainable AI decision paths and thought streaming.

Organizations should also be actively monitoring and optimizing AI agent performance with comprehensive analytics that track the likes of latency, workflow success, and operational efficiency. Adopting such an approach would help to build confidence while reducing the risk of AI being perceived as a black box.

Finally, we must never forget who has the final word. AI should be an enabler, not a replacement for human expertise. Organizations that combine AI’s analytical capabilities with human judgment will be better positioned to innovate while maintaining ethical and privacy standards.

There’s a lot to take in. But one thing is clear. Hyper-personalized AI presents enormous opportunities for businesses. But it comes with responsibilities. Work with vendors with a responsible AI framework and platform that ensure robust tools and embedded ethical considerations for meticulous data curation, rigorous model testing, ongoing transparency, and continuous monitoring and adaptation of AI systems.

The result then becomes not just compliance but also enhanced user trust and exceptional experiences, setting the stage for pioneering a future where AI is a trusted ally.

We feature the best employee 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