Feed aggregator

In one of the oldest cities in Lebanon, once besieged by Alexander the Great, ordinary people struggle to survive Israeli attacks.

(Image credit: Diego Ibarra Sánchez for NPR)

Anthony Bailey's case is one of about a dozen that could be directly affected by a Supreme Court ruling limiting how prisoners can use the compassionate release program to get out early.

(Image credit: Via Anthony Bailey)

In 2028, President Trump will not be on the ballot, leaving Republicans to decide the future of the party. Utah — which has a complicated relationship with the president — could be a starting point.

(Image credit: Rick Egan)

An NPR analysis of more than a thousand Trump endorsements in House, Senate and governor races over the last decade finds the president now picks candidates earlier — and in safer races.

(Image credit: Win McNamee)

While workers, who were employees in government or public institutions, feel vindicated by how their lawsuits concluded, they are still grappling with the aftermath.

(Image credit: Kathryn Gamble for NPR)

If your doctor prescribes a GLP-1 medication for weight loss but your insurance won't cover it, you have options.

(Image credit: Oona Zenda)

For most of the last forty years, data center performance gains came from one place: smaller transistors. Moore's Law and Dennard scaling did the work.

Each new generation of silicon delivered more performance at the same or lower power, and thermal was a maintenance problem, not a performance limiter.

Cooling sat in the background. Operators measured it through PUE, optimized for it where convenient, and otherwise treated it as overhead.

That world is over.

Dennard scaling broke years ago, transistor efficiency gains are leveling off, and AI accelerator TDPs have climbed from 700 watts in the H100 generation to over 1,400 watts in current Blackwell deployments, with NVIDIA's upcoming Rubin platform expected to push further.

Thermal is no longer something that happens after the architectural decisions. It is now the binding constraint on how much performance a chip can sustain, and it is becoming one of the most strategic choices an AI data center operator can make.

Why this matters now

The macro numbers explain why this matters now. Data centers already consume up to 4.5 percent of total U.S. electricity production, a figure projected to reach 12 percent by 2028. McKinsey estimates global data center spending could approach $7 trillion by 2030, and that data center power demand will reach 220 gigawatts in the same window.

None of that capacity arrives quickly. New transmission lines and substations now take five to ten years to permit and build, which means operators cannot simply order more power when they need to scale.

The result is a hard pressure to extract maximum performance from the power they already have under contract. That pressure is what is reshaping how the industry thinks about cooling.

Cooling is no longer just an afterthought

For years, cooling was measured as an efficiency loss, captured through metrics like Power Usage Effectiveness (PUE) that quantified how much energy was burned on overhead before reaching the IT load. Today, the more meaningful metric is how much useful compute you extract per unit of power. NVIDIA's Jensen Huang now describes this as "performance per watt" or "tokens per watt" for AI workloads, and cooling plays a direct role in both halves of that equation.

Direct-to-chip liquid cooling has become the new baseline because it removes heat far more effectively than air. But even direct-to-chip is being pushed to its limit by 1,000+ watt accelerators, and most current deployments still require facility water around 30 degrees Celsius to stay within ASHRAE W2 and W3 envelopes, which means chillers running for much of the year in warm climates.

Better thermal management has effects on both sides of the tokens-per-watt equation. It reduces facility overhead, so more of the contracted power reaches the rack. And it allows chips to operate closer to their full thermal headroom, sustaining higher performance for longer.

Those gains compound. Recent UCLA study has shown that combining a 17 percent improvement in facility efficiency with a 15 percent gain in server-level performance per watt from better thermal management translates to roughly 35 percent more tokens per watt within the same power envelope. In a 10 megawatt facility, that is more than a megawatt of additional usable compute, with no additional grid commitment.

At GTC 2026, NVIDIA CEO Jensen Huang made this argument explicitly. He told the audience that beyond the silicon roadmap, infrastructure-level optimization across power and cooling represents another factor of two in performance still on the table. "There's no question in my mind there's a factor of two in here, and a factor of two at the scale we're talking about is gigantic," he said.

That gain does not come from a smaller transistor. It comes from rethinking how power and thermal energy move through the rack. Recent UCLA study suggests that at least one third of that infrastructure-level gain is attributable specifically to cooling. Cooling is no longer a support function. It is a primary lever for performance.

Water is becoming a hard constraint

Power is not the only pressure point. Water is emerging as an equally critical and often more immediate constraint on data center expansion. Traditional cooling architectures often rely on evaporative processes that consume vast amounts of water. According to the Environmental and Energy Study Institute, large data centers may use up to 5 million gallons per day, comparable to the daily water use of a town of 10,000 to 50,000 people.

This is drawing notice from regulators and communities in already water-stressed areas. The result is longer permitting cycles, higher project risk, and in some cases new developments paused entirely. States and municipalities are also implementing stricter reporting requirements and adjusting electricity rate structures specifically for data centers.

Operators now have to factor water alongside power into site selection. Facilities that minimize energy waste and reduce or eliminate water consumption are better positioned to navigate this environment.

The shift toward next-generation cooling

In response, the industry is entering a new phase of cooling innovation. Air cooling is no longer sufficient for high-density AI workloads. Liquid cooling has become the baseline, but within liquid cooling, not all approaches deliver the same efficiency or scalability.

The next wave of innovation focuses on improving heat transfer at the source: removing thermal energy more effectively at the chip level while reducing system-wide overhead. Some of these approaches draw on heat transfer techniques refined in other high-density power industries such as nuclear power generation, where the challenge of moving large amounts of thermal energy from a constrained physical space has been studied for decades.

The goal is straightforward. Better cooling enables higher rack densities, allows operation at higher facility water temperatures, and reduces or eliminates reliance on water-intensive heat rejection. Just as importantly, the next generation of cooling architectures is being designed to integrate with existing data center footprints, so operators can evolve their infrastructure rather than rebuild it from scratch.

NVIDIA's Vera Rubin platform, announced at CES 2026, was a clear signal of where this is heading. Vera Rubin is designed for 45 degree Celsius supply water, which means dry coolers can do most of the heat rejection year-round and mechanical chillers become optional in most climates. That is a fundamental shift in how cooling infrastructure will be designed for the next decade.

A defining moment for data center design

The data center industry is at an inflection point. AI compute demand is accelerating, and every resource needed to support it, power, water, physical space, is becoming harder to secure. Cooling sits at the intersection of all three.

It determines how efficiently power is used, how much water is consumed, and ultimately, where infrastructure can be deployed. The operators that recognize this now will have a sustained advantage. How to keep data centers cool under AI workload pressure has become one of the most strategic decisions in modern infrastructure.

We feature the best web hosting services: tested and reviewed.

This article was produced as part of TechRadar Pro Perspectives, our channel to 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/pro/perspectives-how-to-submit

U.S. lifts oil sanctions on Iran, Trump visits Pennsylvania to tout economic wins, judge blocks DOJ subpoenas targeting Minnesota officials.

France is experiencing scorching heat, with most of the population exposed to extreme temperatures. The heat is expected to last until at least the end of the week.

(Image credit: Michel Euler)

When the Sony WH-1000XM6 dropped last year, they had some big expectations to live up to.

They followed up a pair of headphones that had proven somewhat divisive, in the Sony WH-1000XM5. This model left some wearers conflicted owing to the lack of foldability, limited new features, and smaller 30mm drivers, in spite of stepping up in terms of looks and still offering great sound. For me, the XM5 just didn’t quite feel like the statement headphones that the legendary Sony WH-1000XM4 were — and in my view, the real task of the XM6 was to live up to the brilliance of that model.

And what do you know, Sony really did it — the company finally delivered a pair of XM4-beating headphones. The XM6 blended the beauty of the XM5 with the practical design of the XM4, all while delivering substantially better audio than their predecessors, alongside far greater noise cancelling capabilities and a more fleshed-out feature set.

And even though it’s been a year now, the XM6 remain my go-to headphones — even after hearing other flagship kids on the block, including the Bose QuietComfort Ultra Headphones Gen 2, Apple AirPods Max 2, and even Sony’s own 1000X The Collexion cans. But why do I keep reaching for the XM6? And what really sets them apart from the competition? Well, I’ll get to all of that and more as we take a look at my year of listening with the XM6.

What’s so special about the XM6?

(Image credit: Future)

For the uninitiated, allow me to give you the run down on the Sony WH-1000XM6. These are Sony’s best noise-cancelling cans to date, and even on flights, they can easily dispatch low-end rumble, high-pitched clamors, and sudden noises like doors closing. If you’re at the office, things like colleagues chatting, typing sounds, and vehicles passing outside the window will be drastically dulled too — if not totally inaudible.

That’s thanks to Sony’s QN3 noise-cancelling processor, which harnesses a system of 12 microphones to deliver some of the best ANC on the market right now.

The QN3 processor also features a noise-shaper, which pre-empts sudden sound changes, resulting in a more controlled listening experience. Pair that with LDAC for ‘hi-res’ Bluetooth listening as well as a more balanced sound signature than the XM5 and XM4 offered — in part thanks to Sony’s collaboration with leading mastering engineers — and you’re getting amazing audio from the XM6.

As you’d expect, these headphones aren’t producing true studio-grade neutrality or fidelity, but they sound more composed right across the frequency range. Still, bass is phenomenally dynamic and punchy, treble is vibrant and articulate, and there’s definitely an exciting listen to be had. Mids are also incredibly rich and well-weighted, and the XM6 offers a wider soundstage than its predecessor, all while supplying a tight and cohesive listen.

Other XM6 headlines include a new, more comfortable headband, the welcome return of foldability, and enhanced call quality. There are six beamforming mics in the XM6, and I’m yet to try any headphones that beat them for clarity and precision in the call quality department.

You also get the features that made the XM6’s predecessors great, including DSEE Extreme upscaling (which boosts the quality of lower-res audio files), scene-based listening, transparency mode, and the most responsive and accurate touch controls around.

How do they compare to the competition?

Here I am, holding up the Sony WH-1000XM6 and a pair of the Bose QuietComfort Ultra Headphones (Image credit: Future)

So, the XM6 sound pretty great so far, right? But you might be wondering how they compare against the competition. So, let’s see how they compare against some of their main rivals: the Bose QuietComfort Ultra Headphones Gen 2, the AirPods Max 2, and Sony’s own The Collexion headphones.

Let’s start with the QC Ultra, which are in my view the main competitor to the XM6 — they’re similarly priced, and each have a strong focus on ANC performance. But interestingly, I think this contest is a relatively straightforward one.

When it comes to ANC, these two are on very similar footing. I’d perhaps argue that the QC Ultra can deal with higher-pitched sounds ever-so-slightly better, but I’m talking very fine margins. Conversely, I’d consider the XM6 slightly ahead when it comes to those deeper, darker sounds like engines roaring.

But elsewhere, I think the Sony XM6 are just the better buy. For me, the Bose have a less revealing sound that lacks the refinement and attention to detail Sony can deliver. Sony also supplies better balance across the frequency range, and even more customizable sound, making them my clear favorite sound-wise.

The XM6 also have a more cohesive, sleek, and suave look in my opinion — I’m also more partial to the magnet-lock case that Sony bundles in. Both headphones are well-matched elsewhere, though, offering fully foldable designs, solid security and comfortability in-use, and 30 hours of battery life — essentially the standard for headphones in this price bracket.

Sony WH-1000XM6 next to the Apple AirPods Max 2 (Image credit: Future)

Moving onto the AirPods Max 2 now, and this is where things get interesting. Of course, these headphones have some features that are suited to Apple devices only — such as Spatial Audio and instant pairing — meaning brand loyalists may prefer these. But as someone with an Android phone, this is hardly something I’m interested in.

I’d also argue that the XM6 are generally better all-rounders than the AirPods Max 2. For instance, Apple’s flagship headphones only muster up a measly 20 hours of playtime with ANC on — something that really should’ve been improved from the first generation AirPods Max.

I also find the XM6’s use of precise touch controls to be more intuitive and user-friendly, and their more accessible price is an undeniable plus. The AirPods Max 2 come in at $549 / £499 / AU$999 — quite the increase over the $449 / £349 / AU$699 you’ll pay for the XM6.

The AirPods Max 2 still impressed me though — they offer significantly improved ANC that blocks external noise on a similar level to the XM6, and they performed very well in the audio department. They have a more spacious and expansive soundstage than the XM6, and the bass is full-sounding and beautifully controlled. While the XM6 sound pretty balanced, they undoubtedly retain a slight preference towards bass and treble from previous generations, but I personally prefer their tighter, punchier approach to the low-end.

Sony 1000X The Collexion beside the Sony WH-1000XM6 (Image credit: Future)

It’s an incredibly similar story for Sony’s own 1000X The Collexion headphones, which took aim at the AirPods Max with an incredibly broad, almost three-dimensional soundstage that results in a highly immersive listen.

The Collexion undoubtedly provide stellar detailing and incredible instrument separation, but again, I always reach for the XM6 — I find their dynamism and punch to add a bit more excitement, making funky tracks or hard-hitting anthems come through with more bite.

In addition, The Collexion add very little in the way of features that the XM6 won’t already provide. Sure, they have more 360 Upmix modes, but these sound pretty bad anyway. They also have weaker ANC and lower battery life, largely due to their slimmer build and larger ear cavities.

Don’t get me wrong, The Collexion look great, and their metallic details contrast the faux-leather casing nicely, but for me, the XM6 are stronger overall.

One year later: the verdict

The magnet-lock case for the XM6 is truly excellent (Image credit: Future)

So, even after testing some of the XM6’s biggest rivals, I still reach for them every time. Their combination of class-leading ANC, excellent features, and dynamic, punchy, yet well-balanced audio makes them my go-to time after time.

It’s also worth flagging that after a full year of use, the headphones have held up incredibly well. I can count a total of zero times where I experienced a fault or bug, they don’t have a single scratch (even after being folded and thrown in my bag countless times), and the earcups feel incredibly comfortable, with no signs of degradation whatsoever.

What’s more, I’ve accidentally kept the headphones on during rainy days from time to time, and despite lacking an IP rating, they seemed to have weathered the storm without issue. Would I recommend using them in harsher weather conditions? No. But the fact they’ve stuck it out a couple of times only hammers home their durability and high build quality.

I won’t pretend that the XM6 are cheap by any means. But they genuinely earn their price in every way imaginable. And they do go on sale quite a bit, so if you can find them going for less, I’d strongly suggest snapping them up, and treating yourself to superior sound.

I don't know if you've seen the price of a PlayStation 5 SSD lately, but they're absolutely cooked. Right now prices are easily triple what they were just a few years ago thanks to the pressure caused by AI demand, so I'd recommend other options instead.

Luckily Amazon Prime Day Is bringing some serious value, with the 5TB Seagate Game Drive PS5 external hard drive discounted to just £143.48 (was £207.99) right now.

• Browse the full Amazon Prime Day sale

This is an external HDD, which means it functions a little differently to a regular SSD, but it's still designed for storing games. It's also officially licensed for PS5, and has a design that matches the console itself to keep your setup looking super slick.

Today's best PS5 storage deal

This officially licenced external PS5 HDD from Seagate is the way to go if you want to upgrade your console storage right now. A quality 4TB PS5 SSD currently costs about £470, making this drive roughly a third of the price for more storage.View Deal

So what makes a PS5 external hard drive different to a PS5 SSD? The main differentiator is that you can't play the PS5 games stored on an external hard drive directly. Instead, they need to be copied back to your internal storage for each use.

It's perfect if you have slow internet speeds and want to avoid lengthy downloads when coming back to large games, which is mainly how I use the Seagate Game Drive hooked up to my own console.

Backwards-compatible PS4 games are playable with no barriers, though, making this a very good choice for those with large libraries of older games they want to keep installed.

More Prime Day deals in the UK

• Amazon Prime: sign up for a 30-day free trial
• Amazon Devices: Fire TV, Ring & Blink from £24.99
• Amazon Haul: up to 30% off
• Apple: AirPods Pro 3 for a record-low price
• Appliances: up to 45% off Ninja, Tefal & Sage
• Beauty: up to 60% off Philips & Oral-B
• Essentials: household goods from under £10
• Fashion: up to 50% off
• Headphones: up to 50% off
• Laptops: from £149.99
• Tablets: Samsung & Lenovo from £125
• Toys: up to 25% off Lego and Tonies
• TVs: from £129.99
• Vacuums: up to 40% off Eufy & Roborock
• Wearables: Garmin & Huawei from £36.99

Sony's premium Mini LED TV brings 4K resolution, deep contrast and PS5-ready features to the big screen.

Get up to 70 minutes of runtime and almost 40% off the price.

Generative AI has moved quickly from experimentation into early production use in many enterprises. However, very few can confidently forecast what it’s going to cost them in six months.

For a technology that has consumed so much board-level attention and capital, that reflects a lack of certainty, and one that some technology leaders may privately recognize as true of their own organizations.

The spend is real and the direction is clear, but the number at the end of the year can remain genuinely uncertain.

To capture a glimmer of the confidence driving the infrastructure race, Amazon’s CEO has indicated it expects to spend heavily on IT infrastructure to support AI, with an estimated $200 billion in AI capital spending, arguing it is “not going to be conservative” in how it invests in the tech.

In practice, what makes AI different from the infrastructure investments that came before it is not the scale of the commitment but the nature of the consumption.

Cloud computing was unpredictable when it arrived too, but it eventually settled into patterns that finance teams could learn to model. AI hasn’t settled in the same way yet, and much of the reason comes down to how it is being used.

A great deal of enterprise AI use remains exploratory, which is part of what makes forecasting harder. And unlike cloud, which stayed largely within technical teams for years before spreading, AI is moving across the whole organisation almost immediately. That changes everything about how you try to govern it.

The limits of financial visibility

On the surface, some forms of AI appear to offer what earlier infrastructure lacked: clean, granular, real-time data and what it costs. But across the rapidly growing landscape of technology providers leveraging AI in some way, many do not.

In some cases, token-based pricing is precise in a way that early cloud billing never was, and for finance teams accustomed to working with far less, it can feel like a step in the right direction for solving the visibility problem.

We unfortunately still have a long way to go, since simply understanding what was spent last month tells you very little about what will be spent next quarter, particularly once adoption moves beyond the teams who originally shaped the business case.

One must consider that teams across legal, HR, and customer operations are not thinking about token economics (tokenomics). They’re only thinking about whether the tool works.

Cost exposure builds not through any single decision but through dozens of small expansions, each reason in isolation, none of them reflected in a comprehensive forecast. By the time anyone joins the dots, the demand curve has already moved.

Extending the disciplines that already exist

The organizations who are doing a better job managing AI spend have tenured experience managing consumption-based technology. IT asset management (ITAM) teams for example often have more experience dealing with more fixed constructs like users or seats, which makes the consumption-based nature of AI far more challenging.

FinOps teams on the other hand have grounded experience in managing consumption that originated in public cloud. FinOps teams may therefore better positioned to deal with the new tsunami of AI consumption and spending, ensuring that it is governed as adoption scales.

FinOps has also been broadening its scope beyond the initial roots in public cloud, with AI cost management now sitting firmly within that remit for many, a shift reflected in how the FinOps Foundation is increasingly incorporating AI into its guidance. Part of that expansion is about forecasting demand that behaves differently from conventional workloads.

There is also growing interest in whether AI itself can support FinOps practices, particularly in anomaly detection, optimization and, over time, forecasting, as consumption patterns become harder to model.

The challenge is applying FinOps practices early enough so that governance shapes how AI scales, rather than scrambling to restore control once spend has already outpaced oversight.

The compounding difficulty of legacy environments

For organizations whose technology estates were built around consistency, extending governance into AI is harder than it sounds.

AI-first organizations design with cost in mind from the beginning, treating inference the way they would any other product input, with economic constraints shaping architecture decisions before commitments are made.

Retrofitting AI into legacy infrastructure means something different. Existing commercial commitments and operating models do not adapt quickly to a consumption model that is inherently variable, and that friction has a direct bearing on cost.

The difficulty is often that AI is being introduced into environments built around very different assumptions about how demand behaves, and that is part of what makes forecasting harder.

The challenge is not simply new spend, but expenditure ballooning in environments where oversight and control are already difficult to maintain.

Organizations navigating this will tend to run controlled experiments before broad rollout and are deliberate about how adoption spreads. In practice, that is often about containing unmanaged adoption early, before usage patterns, costs and dependencies become harder to unwind.

That same exposure increasingly carries beyond internal governance. As AI appears more often in customer procurement conversations, questions that were once largely internal are starting to be probed externally too.

For organizations whose governance has not kept pace, those questions can force a level of clarity they may not yet be prepared to provide.

From activity metrics to business outcomes

Beyond governance and cost control, there remains a harder question, which is whether AI investment is producing meaningful business value. Most leadership teams are not yet in a position to answer that with confidence, and the metrics currently reaching the board are not making it easier.

Model usage, inference volumes and compute consumed describe activity without explaining value. It is easy to build a compelling board update from consumption data without addressing whether any of it is moving the business.

What gets closer to an answer is understanding whether individual inferences are delivering something a customer would pay for, or something that meaningfully reduces cost or risk.

Incremental business outcome per pound or dollar of AI spend is a harder measure to produce, but it is closer to the economics that matter because it requires a clearer position on what AI is actually delivering.

That is precisely where many organizations are still finding the work harder than it looks, particularly as AI deployment moves ahead of the models used to understand cost and value.

That disconnect matters more as the market expands, because where those economics remain unclear, cost exposure can build in ways that are harder to recognize early and harder to contain later.

For many enterprises, the challenge ahead is scaling AI without allowing spend to outrun the value it is meant to create.

The best cloud storage: tested, reviewed and rated by experts.

This article was produced as part of TechRadar Pro Perspectives, our channel to 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/pro/perspectives-how-to-submit

I've been trying out Amazon's latest — and possibly greatest — e-reader, the Kindle Scribe Colorsoft, and while I love it for reading comics in full colour and note-taking, I've been waiting patiently for Amazon Prime Day to see how the device fares in the megasale.

As expected, it's seen a sizeable discount; 20% off, knocking it from £629.99 to £504.99, which is, to be honest, much closer to the value proposition I think the slate offers. Still, though, it's pricey for what it is; and the fact that it's actually cheaper right now to buy a Kindle, Kindle Scribe and a Kindle Colorsoft separately leaves me perplexed.

Right now, the Kindle (in graphite) is £79.99 (was £94.99) at Amazon, the 32GB Kindle Scribe (in tungsten grey) is £244.99 (was £399.99) and the Kindle Colorsoft is £154.99 (was £239.99). That totals £479.99; £26 cheaper than just one Kindle Scribe Colorsoft. Make it make sense.

While we scored it a respectable 4 stars in our review, a major sticking point for the Scribe Colorsoft is its value proposition. There's nowehere near enough going for it to justify this e-reader costing more than a MacBook Neo. View Deal

For the pared-back basic experience, the most recent Kindle (2024) scored 4 stars in our review as an affordable and capable e-reader. It's compact enough to stash in your hand luggage while travelling, performance is solid, and it taps into Amazon's vast library of books. View Deal

A notetaker's best friend, the Kindle Scribe is great for students, avid readers and productivity users alike, offering a satisfying and slick screen on which to doodle and annotate. We scored it 4.5 stars in our review.View Deal

If you mostly want the colour screen, you'll be just fine with the Kindle Coloursoft; as a comic reader, this is really all I want and need from my Kindle. While you can't ever match the vibrancy of an LCD or OLED screen in e-ink, it's a valiant effort that earned the device 4.5 stars in our review.View Deal

While not a factor in the Kindle Scribe Colorsoft pricing hoo-ha, it's worth noting the Kindle Paperwhite is also on sale with a neat £35 discount. We scored this model 4-stars in our review, praising its bright, white screen and long battery life.View Deal

The Kindle Scribe Colorsoft isn't without its selling points entirely; if you really do want to read comics and take notes on one device, it's the only Kindle that offers both functions... just not in tandem, so you can't annotate comics or manga.

That means the main benefit is access to colour pens and highlighters, which isn't a big enough sell for me to drop half a grand on an e-reader.

More Prime Day deals in the UK

• Amazon Prime: sign up for a 30-day free trial
• Amazon Devices: Fire TV, Ring & Blink from £24.99
• Amazon Haul: up to 30% off
• Apple: AirPods Pro 3 for a record-low price
• Appliances: up to 45% off Ninja, Tefal & Sage
• Beauty: up to 60% off Philips & Oral-B
• Essentials: household goods from under £10
• Fashion: up to 50% off
• Father's Day: up to 50% off gifts & drinks
• Headphones: up to 50% off
• Laptops: from £149.99
• Tablets: Samsung & Lenovo from £125
• Toys: up to 25% off Lego and Tonies
• TVs: from £129.99
• Vacuums: up to 40% off Eufy & Roborock
• Wearables: Garmin & Huawei from £36.99

If you were looking to get your hands on a PlayStation 5 ahead of the Grand Theft Auto 6 pre-order date this week, then this Amazon Prime Day deal is for you. Right now you can grab the console at a 16% discount over at Amazon, which takes its price down to just £479 (was £569.99).

• Browse the full Amazon Prime Day sale

It's not the cheapest PS5 we've ever seen, but with the recent price increases, it's likely going to be the lowest it's available for some time. This is the Slim 1TB version of the console as well, with a disc reader fitted so you can buy the physical version of GTA 6 to keep on your shelf.

The deal is only on for a limited time, and stock is already flying off the shelves - with just under half of the available units sold already. This is a deal I'd recommend snapping up quickly!

Today's best PS5 deal

A chunky discount on the PS5 Slim here, with 1TB storage and the disc reader attachment included out of the box. With recent price hikes, this is likely the cheapest the console will be ahead of GTA 6 pre-orders.View Deal

More Prime Day deals in the UK

• Amazon Prime: sign up for a 30-day free trial
• Amazon Devices: Fire TV, Ring & Blink from £24.99
• Amazon Haul: up to 30% off
• Apple: AirPods Pro 3 for a record-low price
• Appliances: up to 45% off Ninja, Tefal & Sage
• Beauty: up to 60% off Philips & Oral-B
• Essentials: household goods from under £10
• Fashion: up to 50% off
• Headphones: up to 50% off
• Laptops: from £149.99
• Tablets: Samsung & Lenovo from £125
• Toys: up to 25% off Lego and Tonies
• TVs: from £129.99
• Vacuums: up to 40% off Eufy & Roborock
• Wearables: Garmin & Huawei from £36.99

Amazon's Prime Day sale is here (running from June 23 to 26, and exclusive to Prime members), and if a 360 camera has been on your wish list, it's the perfect time to act. Two of Insta360's best models are getting solid discounts in the sale — but the one I'd actually put in my basket might not be the one you're expecting.

I reviewed the Insta360 X5 last year, and its low-light performance was the standout feature for me. It remains, in my view, the best all-rounder Insta360 makes. At 21% off ($434.99, down from $549.99 for the Standard Bundle, a $115 saving), it's a pretty tempting buy in the Prime Day Sale.

But my colleague Peter Fenech has since put the newer Insta360 X4 Air through its paces, and with Prime Day discounts of 25% on the Standard Bundle ($299.99 at Amazon, down from $399.99) and 26% on the Starter Bundle ($324.99, down from $439.99, also a $115 saving), I think it's the smarter buy for most people.

The Insta360 X5 and X4 Air are both superb 360 cameras, and they're currently both on sale for Prime Day. (Image credit: Future / Peter Fenech)A baby X5, in the best sense

The X4 Air inherits a lot of what made the X5 so appealing. It uses the same easy-to-swap lens replacement system (which can be a real lifesaver if you scrape a lens on the ground), and it's waterproof to the same 15m depth without a housing.

At 165g, it's also noticeably lighter than the X5's 200g. It's not a huge number on paper, but you'll feel the difference after an hour of holding a selfie stick overhead.

The real divergence is in sensor size. The X5 uses larger 1/1.28-inch sensors, while the X4 Air steps down to 1/1.8-inch (still bigger than the original X4's sensors, for context). In daylight, you'd struggle to tell the two apart. It's only once the light fades that the gap opens up, with the X5's PureVideo mode giving it around a two-stop low-light advantage over the X4 Air.

That sounds like a point in the X5's favor, and it is. But I think that the vast majority of 360 footage gets shot in daylight — think travel clips, bike rides, beach days, family get-togethers — and the X4 Air handles all of that with ease. Peter was impressed by its image quality overall, and found its low-light performance better than expected even without PureVideo on board.

On the battery life front, the X5's larger 2,400mAh battery wins out for marathon sessions: its Endurance Mode can run for over three hours at 5.7K, versus around 105 minutes for the X4 Air with the same settings. But at 8K/30fps (the resolution most people will actually shoot in), both cameras land at roughly 90 minutes.

The X4 Air is slightly lighter than the X5, which will feel like a benefit on a long day of shooting. (Image credit: Future)

If you're a professional creator, you regularly shoot in low light, or you simply want the best 360 camera Insta360 currently makes, the X5 at 21% off remains an easy recommendation.

But if this is your first 360 camera, or you want something that genuinely disappears into a pocket, the X4 Air gives you almost everything that made the X5 special, for $100 to $135 less. That's where my money would be going this Prime Day.

Today's best Insta360 X4 Air and X5 deals (US)

Taking many of the X5 features but squeezing them into a smaller body, there's plenty to like about the X4 Air, especially with its first major discounts for Prime Day, now 25% off. View Deal

For 360-degree action cameras, it doesn't get any better than the X5 from Insta360. With huge sensors, the X5 delivers 8K video and excels in low light. It's also feature-rich with replaceable lenses to boot. Save $85 on the best action camera for shooting in 360.View Deal

Today's best Insta360 X4 Air and X5 deals (UK)

There's also 25% off the X4 Air for shoppers in the UK, taking the base package price down to a record-low £269. Not bad for a 'baby X5'.View Deal

My favorite 360 camera, with a healthy 25% off — it doesn't get any better! With huge sensors, the X5 delivers 8K video and excels in low light. It's also feature-rich with replaceable lenses to boot. Save $85 on the best action camera for shooting in 360.View Deal

More of today's best Insta360 dealsMore Prime Day deals in the US

• Amazon Devices: Fire Sticks & Echo from $18
• Amazon Haul: viral gadgets, tech & appliances from $1.99
• Apple: MacBooks, AirPods & AirTags from $29
• Beauty: 50% off toothbrushes & hair tools
• Cheap TVs: smart TVs from $69.99
• Garden: tools, mowers, planters from $24.99
• Headphones: 50% off Beats, Bose & Samsung
• Laptops: Apple, HP & Dell from $199
• Mattresses: Sealy, Serta & more from $186
• Patio: outdoor furniture, rugs & decor from $19.99
• Sports: 50% off fitness gear, treadmills & clothing
• Vacuums: Dyson, Shark & Bissell from $34

More Prime Day deals in the UK

• Amazon Prime: sign up for a 30-day free trial
• Amazon Devices: Fire TV, Ring & Blink from £24.99
• Amazon Haul: up to 30% off
• Apple: AirPods Pro 3 for a record-low price
• Appliances: up to 45% off Ninja, Tefal & Sage
• Beauty: up to 60% off Philips & Oral-B
• Essentials: household goods from under £10
• Fashion: up to 50% off
• Father's Day: up to 50% off gifts & drinks
• Headphones: up to 50% off
• Laptops: from £149.99
• Tablets: Samsung & Lenovo from £125
• Toys: up to 25% off Lego and Tonies
• TVs: from £129.99
• Vacuums: up to 40% off Eufy & Roborock
• Wearables: Garmin & Huawei from £36.99

Agentic AI is moving rapidly from boardroom ambition to enterprise reality.

Gartner forecasts that roughly 40% of enterprise applications will incorporate task-specific AI agents this year, up from just 5% last year.

This surge forces every CIO, CISO, and technology leader to consider: What should AI be allowed to access, and how should it operate once inside the enterprise?

Many organizations begin by embedding AI agents directly into legacy systems, connecting them to backend databases, APIs, and workflows in the name of speed.

While this inline approach can work in modern, well-governed environments, it often bypasses the approval workflows and controls that legacy systems were built around. Agents can access restricted data, skip approvals, or execute transactions without a complete, attributable record.

The result is a growing governance gap. Decisions tied to sensitive data can’t be reliably reconstructed or defended with the same confidence as human-driven work. Even advanced models stall in pilots because organizations can’t prove how outcomes were produced.

The solution is not to slow AI adoption. It’s to change how AI interacts with the systems that already run the business.

When AI bypasses the system, it breaks it

Consider a finance workflow in an ERP software system. An agent updates vendor bank details and pushes a payment through a fast-track path, bypassing a required approval step and segregation-of-duties check. Later, when the transaction is questioned, the organization can’t prove who approved the change, why it was made, or whether proper controls were followed.

That’s where accountability breaks down. Changes are made inside core systems, but the evidence is incomplete, inconsistent, or disconnected from the system of record.

Emulated human behavior offers a more secure and practical path. These agents operate exactly as a human employee would: logging in with standard credentials, navigating the existing user interface, reading screens in context, following established workflows, and executing tasks while remaining fully subject to every control already in place.

No new APIs. No raw backend data exposure. No rewriting of decades-old business logic or security rules. The guardrails designed to protect against human error or misuse — validations, permissions, approvals, and audit logging — remain 100% intact.

This UI-first approach is especially effective for organizations running mission-critical processes on older platforms. Building secure, governed APIs for legacy systems is expensive and time-consuming, often leaving out protections built into the interface layer.

While emulated human agents may not match the speed of direct backend calls, they provide far more valuable enterprise advantages: immediate deployability, ironclad accountability, and zero disruption to proven controls. Secure operation doesn’t require avoiding AI. It requires rethinking how it fits into the systems around it.

Preparing for emulated human in the enterprise

Three priorities can help organizations prepare for the emulated human approach as AI scales into critical workflows.

1. Place AI at the points where work happens

Most enterprise AI strategies assume deeper backend integration creates better automation. In environments shaped by legacy systems, it often does the opposite: introducing new complexity while bypassing the workflows and controls already built into the interface layer.

Instead, focus AI at the points where it can operate without requiring systems to be rebuilt. This approach dramatically reduces integration overhead, limits exposure of core systems, and allows AI to scale within existing operating models rather than forcing costly modernization.

2. Align AI accountability with human accountability

Agents should operate under named identities and the same policies as employees. They preserve approval workflows, follow role-based permissions, and generate the same audit artifacts — including log entries, change histories, tickets, and recorded approvals — that organizations already rely on to review human activity.

This removes the dangerous two-tier governance model where AI operates under different standards than employees. Organizations can maintain visibility, accountability, and established compliance and risk management controls as AI takes on greater responsibility.

3. Design for adaptability rather than brittle automation

Traditional robotic process automation (RPA) relied on rigid, click-by-click scripts that broke the moment screens changed or exceptions appeared. Emulated human agents interpret context in real time, adjust to variation, and continue operating, just as skilled employees do.

That adaptability is essential in dynamic enterprise environments where policies change, exceptions are common, and systems are rarely static. Instead of constant break/fix maintenance, organizations gain AI that can operate more resiliently inside real-world workflows.

Scaling AI with the systems already in place

As agentic AI scales, enterprises will be judged not only by the intelligence of their systems but by their ability to govern them. The pressure to balance innovation with control will only intensify.

The most durable strategies will be those that embed AI safely within the systems already in place, rather than racing around them. When an agent’s actions can be audited and justified with the same rigor applied to a human colleague, it’s finally ready for production.

That’s how secure, scalable AI will be defined in the enterprise.

We feature the best small business software.

This article was produced as part of TechRadar Pro Perspectives, our channel to 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/pro/perspectives-how-to-submit

Brexit fractured the European Union, and broke British politics. A decade on, millions of voters have deserted the two big parties for alternatives.

(Image credit: Kirsty Wigglesworth)

Firefighters are still battling a blaze at a massive frozen-food storage facility near downtown Los Angeles six days after the fire started.

(Image credit: Jae C. Hong)