News
Mention high-frequency semiconductors, and the conversation often begins with gallium arsenide (GaAs). For decades, GaAs has been the material of choice for high-performance radio frequency (RF) devices that enable a wide range of applications, including satellite communications, radar, and mobile networks.
It is a mature, well-understood technology, and large, high-quality wafers can be grown economically, enabling consistent device performance and scalable manufacturing.
Yet a transition is underway. Increasingly, the industry is turning away from GaAs and towards gallium nitride (GaN), a material that offers clear performance advantages, particularly at higher frequencies, higher power levels, and more demanding operating conditions.
The promise of GaN is well established. The challenge lies in making it reliable, scalable, and commercially competitive.
And, as with many emerging semiconductor technologies, meeting that challenge is as much about measurement as it is about materials.
A material with advantages, and constraintsGaN enables the development of devices that can operate at higher voltages, higher temperatures, and greater power densities than their GaAs counterparts. This makes it particularly attractive for next-generation RF systems, including 5G and 6G communications, defense applications, and space technologies – where extreme environments are commonplace.
In principle, GaN can offer a straightforward upgrade path. It is sometimes thought of as a functional replacement for GaAs to deliver improved performance. Yet, in practice, the situation is more complicated, and a whole-system approach is required to redesign a module. Furthermore, the manufacturing approach for GaN is fundamentally different.
The hidden cost of heteroepitaxyUnlike GaAs, which can be grown as large, high-quality wafers using established methods, GaN presents a fundamental manufacturing challenge because producing large, defect-free GaN substrates remains difficult and expensive.
As a result, most GaN devices are produced using a process called heteroepitaxy, whereby a thin layer of GaN is grown on top of a different substrate, typically silicon or silicon carbide. This approach allows manufacturers to leverage existing wafer technologies. But it comes at a cost.
When GaN is grown on a dissimilar substrate, differences in lattice structure and thermal expansion introduce defects into the material. These defects can affect everything from electrical performance to long-term reliability.
This results in a difficult trade-off: GaN offers superior theoretical performance, but achieving that performance consistently across wafers and devices is far more challenging than with GaAs. The advantage of using GaN therefore increasingly depends on material quality, and on the ability to control and understand it.
That relies on measurement.
Measurement as a competitive toolFor companies developing GaN technologies, metrology plays three distinct and essential roles.
The first is in process development. Growing GaN through heteroepitaxy involves carefully balancing multiple parameters, including temperature, deposition rates, and substrate preparation. Small changes can improve or degrade material quality. Without reliable measurement, it is difficult to know whether a process adjustment has made the material better or worse. Metrology provides the feedback needed to refine growth techniques and reduce defect densities.
The second role is in demonstrating material quality. In a market where performance depends heavily on the underlying material, manufacturers must be able to show that their GaN is superior to that of competitors. This requires measurement methods that are not only accurate, but also comparable across organizations. Customers need confidence that a claim about material quality means the same thing, regardless of where it is measured.
The third role is in device performance validation. Ultimately, customers care about how a device behaves in real applications. For RF components, this includes metrics such as power output, efficiency, frequency response, and thermal stability. Linking these device-level characteristics back to material quality is essential. It allows manufacturers to demonstrate that improvements in material growth translate into tangible performance gains.
Across all three areas, measurement is not simply a supporting activity. It is a central part of how competitive advantage is created and communicated.
From materials to systemsThe challenges associated with GaN are part of a broader shift in the semiconductor industry.
As devices become more specialized and operate under more demanding conditions, performance is increasingly determined by subtle interactions between materials, structures, and processes.
This is particularly true in RF systems, where small imperfections can have outsized effects on signal integrity and efficiency.
It also connects to a wider trend seen in other areas of semiconductor technology, including photonics. There, heterogeneous integration is bringing together different materials and device types within a single system, creating similar measurement challenges.
In both cases, success depends on the ability to understand and control complexity at multiple scales.
A strategic inflection pointFor the RF semiconductor industry, the transition from GaAs to GaN represents more than a simple material substitution.
It marks a shift towards technologies where performance is less constrained by established manufacturing processes, and more dependent on how well new materials can be engineered and characterized. This creates an opportunity.
Countries with strong capabilities in materials science, process development, and measurement can play a defining role in shaping how GaN technologies evolve and are applied successfully in the semiconductor industry.
The ability to measure material quality, correlate it with device performance, and establish trusted benchmarks will influence how quickly GaN is adopted across global markets. This means the future of high-frequency RF semiconductors will not be determined by materials alone.
GaN may offer superior intrinsic properties, but those advantages must be realized in practice. That requires consistent, high-quality material growth, reliable device fabrication, and credible performance validation – all of which depend on effective measurement and standards. This is why metrology will become increasingly central to RF semiconductor innovation, not as a downstream check, but as an integral part of development, manufacturing, and market adoption.
In high-frequency semiconductors, as in photonics, the ability to measure well is no longer just a technical requirement. It is becoming a defining feature of competitiveness.
We've featured the best phone system for business.
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
We’ve been sold a comforting idea about artificial intelligence: that it’s making us dramatically more productive. Faster outputs, smarter tools, less effort. A quiet revolution in how we work.
But step back for a moment and ask yourself a simple question.
Do you actually feel more productive? Not faster. Not busier. Productive.
Because for most professionals I speak to, the answer is no.
Work feels quicker, yes. But also more fragmented, more reactive, and oddly more exhausting. The promise of efficiency is there on paper, but the true experience tells a different story.
That disconnect is worth paying attention to.
A typical working dayLook at how most of us spend a typical working day. We move between email, calendar, tasks, notes, messaging platforms, documents. Each tool holds a piece of the puzzle, none of them are the full picture. So, we become the system that stitches it together.
We check an email, then jump to our calendar to understand the context. We open a task list, then search our notes to remember why that task exists. We respond to a message, then dig through previous threads to find what was agreed.
This is not the work itself. It’s the management of work.
Now add AI into the mix.
We have tools that can summarize emails, draft responses, transcribe meetings, generate notes, and even suggest tasks. Each of these capabilities is impressive in isolation. They save minutes here, seconds there.
But they don’t remove the fundamental problem. In many cases, they amplify it.
Instead of switching between tools, we now switch between tools and their respective AI layers. An assistant in your inbox. Another in your document editor. Another in your meeting tool. Each one helpful, but none aware of the others.
So, we’re still managing everything ourselves. We’re just doing it faster.
This is where the narrative around AI productivity starts to unravel.
Defining successWe’ve defined success as speed. How quickly can a tool help you write, summarize, respond, or organize? And to be fair, AI has delivered on that front.
But speed without context is a blunt instrument.
If you’re responding faster but to the wrong priorities, you’re not more productive. If you’re generating more output but not moving meaningful work forward, you’re simply accelerating noise.
The real friction in modern work isn’t the execution of tasks. It’s the constant need to decide what matters, to reconstruct context, to align fragmented information across multiple systems.
AI, as it stands today, rarely addresses that layer.
At warpSpeed, we’ve approached this from a slightly different angle.
We didn’t start by asking how to make tasks faster. We started by asking why work feels so disjointed in the first place.
The answer was fairly obvious: everything is scattered. Email lives in one place, calendar in another, tasks somewhere else, notes somewhere else again. Every decision requires jumping between them.
So, we focused on bringing those elements together into a single, connected environment. a system where context flows naturally between them, facilitated by AI.
Not as a feature bolted onto individual tools, but as something that can see across them. Something that understands not just a single email or a single note, but the relationship between your communications, your commitments, and your priorities.
The difference, while subtle, is meaningful. This is how I like to think a successful assistant would function.
For example, when someone asks, “What should I focus on today?”, the answer isn’t generated in isolation. It draws on overdue tasks, unread emails that require responses, upcoming meetings, and previous commitments. It reflects the reality of that person’s day.
Small changesSimilarly, we’ve seen how small changes in interaction design can shift behavior. One example is email. By rethinking how users move through their inbox, we’ve seen people process large volumes of emails in a fraction of the time they previously spent. Not because they’re working harder, but because the system reduces friction and surfaces what matters.
These are not dramatic, headline-grabbing transformations. They’re incremental improvements grounded in real workflows. And importantly, they’re imperfect. We’re still learning, still refining, still discovering where the real value lies.
But they point to something broader.
If AI is to genuinely deliver on its promise of productivity, we need to rethink what we’re asking it to do.
Right now, most tools are designed to assist with tasks. Write this. Summarize that. Suggest a response. Create a list.
What’s missing is a deeper understanding of context and personalization.
Who is this for? Why does it matter? What else is happening around it? What should take priority?
Without that layer, AI remains reactive. It responds to prompts, but it doesn’t help you navigate your day.
Moving forwardTo move forward, the industry needs to shift in three ways.
First, from isolated tools to connected systems. The value of AI increases exponentially when it can operate across your entire workflow, not just within a single tool.
Second, from generic intelligence to personal context. The most useful AI will be shaped by how you work, what you care about, and how you make decisions.
Third, from output to outcome. It’s not enough to generate content or complete tasks. The goal should be to move work forward in a meaningful way.
None of this is easy. It requires rethinking product design, data architecture, and user experience at a fundamental level. It also requires a degree of restraint. Not every problem needs another feature. Sometimes it needs fewer moving parts.
Productivity at scaleThe irony is that the more powerful AI becomes, the more important simplicity becomes. Productivity at scale depends on removing the need to think through complexity, making intuition more valuable than ever.
Because ultimately, productivity isn’t about doing more things. It’s about doing the right things with less friction.
AI isn’t broken.
But the way we’re using it might be.
If we continue to layer intelligence on top of fragmented systems, we’ll keep getting the same result: faster work, but not better work.
The real opportunity lies in something quieter, but far more impactful. Using AI to remove the need to manage work in the first place.
Not to help you keep up.
But to help you stay focused on what actually matters.
We've reviewed, rated, and ranked the best business laptops.
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
Quantum risk has moved from theoretical to operational in 2026, and this shift is now impossible for enterprises to ignore. Google has committed to completing its post-quantum migration by 2029, the NCSC 2035, and the G7 says 2034. NIST has finalized its first full suite of post‑quantum cryptographic standards, triggering mandatory migration planning across regulated industries.
The timelines might not be perfectly aligned, but the message they send is: the risk is real, and the time to act is now. Breakthroughs in quantum hardware and AI‑accelerated quantum optimization are bringing “Q‑Day”ever closer. This is elevating quantum‑safe visibility, cryptographic discovery, and encrypted‑traffic intelligence to a now‑priority for every enterprise.
This is not the first time the industry has faced a widely anticipated but imperfectly understood threat. The parallels between quantum computing and Y2K are difficult to ignore. What began as stories of a supermarket system rejecting food as 80 years out of date, or a 104-year-old being invited to school because a computer registered her as four, soon evolved into fear as the scale of the problem became clear.
By 1995 the New York Stock Exchange had spent over $30 million remediating its systems. Y2k evolved into a defining moment for risk management and the key lesson was that organizations must understand and respond to exposure quickly and decisively.
Successfully addressing the threat quantum computing poses to encryption demands the same mindset, although this time applied across a far more complex and interconnected digital landscape.
The invisible threat already underwayA key distinction between Y2K and quantum computing is that the latter is not anchored to a single moment in time. The phrase “harvest now, decrypt later” describes the practice of adversaries collecting encrypted data today with the expectation that quantum capabilities will allow them to decrypt it in the future.
The implications of this tactic are significant, with 87 percent of organizations expressing concern about such scenarios as quantum computing advances.
Financial records, personal data, and intellectual property retain their value long past creation and the information encrypted today may still be sensitive well into the future. Decisions made now about cryptographic resilience will directly shape an organization's future security and reputation.
Why PQC is fundamentally more complex than Y2KAt its core, the Y2K challenge was a remediation problem with a relatively well-defined scope. Migrating to post-quantum cryptography is fundamentally different.
Cryptographic controls are deeply embedded across modern digital infrastructure, underpinning applications, APIs, cloud services, IoT devices, operational technology, and a growing web of third-party integrations. In many cases, they operate invisibly and are poorly documented. Organizations are not simply upgrading known systems, but first discovering what encryption levels have been used and where.
Addressing this means building a comprehensive inventory of cryptographic assets. All weak cipher suites, expired certificates, and non-compliant encryption methods must be identified.
Once found, organizations should standardize on stronger protocols such as Transport Layer Security (TLS) 1.3. This faster, more streamlined and longer protocol is ultimately more secure than older versions like TLS 1.1 and TLS 1.2, which will be broken by quantum computers in a matter of hours, minutes, or even seconds.
You cannot secure what you cannot seeIn addressing both Y2K and today’s cybersecurity challenges, one principle consistently determines success: visibility. As organizations plan for a post-quantum future, 91 percent report that visibility into encrypted traffic is critical for PQC readiness.
Network-derived telemetry provides a scalable way to gain this. By analyzing traffic flows and metadata, organizations can build a comprehensive picture of cryptographic usage across both managed and unmanaged assets. This outside-in perspective complements internal inventories and helps uncover dependencies that might otherwise remain hidden.
With improved visibility comes the ability to assess risk more accurately, prioritize remediation, and ensure that the adoption of quantum-resistant approaches does not introduce unintended vulnerabilities.
Avoiding a repeat of historyThe response to Y2K ultimately succeeded because organizations acknowledged the threat and took action. It was a forcing function that led to massive technology infrastructure upgrades and tech stack modernization.
Like Y2K, today’s quantum challenge isn’t just the potential event itself, it’s the scale of the remediation effort required across systems, applications, and embedded technologies, which makes early action critical.
The transition to post-quantum cryptography will not be achieved overnight. It will require a coordinated effort across security, infrastructure, development, and compliance functions, alongside close collaboration with vendors and strategic partners.
More importantly, it requires a shift in how the challenge is framed. Organizations that take proactive steps now by establishing a comprehensive inventory of cryptographic assets, improving visibility across their environments, and developing structured transition plans will be far better positioned to navigate the shift.
They will retain control over their timelines and reduce the likelihood of disruptive, last-minute change.
Those that delay may find themselves in a position that feels uncomfortably familiar. A known problem with a shrinking window in which to respond.
We've featured the best endpoint protection 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
Spare a thought for the Samsung Galaxy Z Fold 8 Ultra, which isn't having the easiest time of it right now. Yes, it's a fantastic foldable phone, but it was launched in the shadow of its upstart sibling, the Galaxy Z Fold 8, with Samsung taking the OG folding phones' name and applying it to a fresh, dare I say cute, design.
The Galaxy Z Fold 8 Ultra, by contrast, is almost indistinguishable from its predecessor, the Galaxy Z Fold 7. That's hardly a criticism — I loved that Android device — but it's hard to stand out when you've launched alongside to the phone world's shiny new object.
If we can put the 4:3 aspect-ratio Galaxy Z Fold 8 aside for a moment — you can read our Z Fold 8 review if you're so inclined — I'll focus here on the Galaxy Z Fold 8 Ultra, a fantastic folding device in its own right.
If you're unfamiliar with the Galaxy Z Fold line, here are the highlights for the Z Fold 8 Ultra. 'Ultra', but the way, does not mean it somehow more closely resembles the Galaxy S26 Ultra, though the camera and CPU enhancements do bring it more in line with the performance of its flagship cousin.
Folded, the Galaxy Z Fold 8 Ultra looks like a standard, if slightly narrower than normal smartphone (the S26 Ultra is a few millimeters wider). Unfolded, it's an 8.0-inch flexible display — and the chief upgrade on the display side is a mechanical and material engineering marvel.
Under the protective plastic cover, ultra-thin glass and a flexible AMOLED screen are a pair of new titanium panels. One is pure titanium, with an all-new perforation design for better bending performance, and the other is an alloy that, while thinner than the previous polymer sheet, is stronger and more reliable.
The Galaxy Z Fold 8 Ultra is on the right here — note the reduced crease compared to the Z Fold 7Lance Ulanoff / FutureThe flatter Galaxy Z Fold 8 Ultra is on the right.Lance Ulanoff / FutureThe quite noticeable result is an essentially flat main screen with a far less visible crease, with the promise that over the years the bendy display won't suffer noticeable deformations (i.e., a more noticeable crease).
There are still five (yes, five) cameras, and while the 200MP main camera is essentially unchanged, the ultra-wide is now a more powerful 50MP camera (up from 10MP). The 3x optical zoom is still 10MP, and I do wish it were a 5x optical zoom (or at least 4x like the iPhone 17 Pro Max's). You can go far further with digital zoom, but then you have to wonder how much of an assist your photos are getting from the on-board AI (assume a decent amount). All of the photography is further improved by the upgraded image processing.
Under the hood is the latest Qualcomm silicon, which supports an impressive array of AI features. As with other Galaxy devices, there's almost too much AI here, but features like Photo Assist and Now Brief are standouts.
All in all, the Galaxy Z Fold 8 Ultra is an excellent follow-up to the Galaxy Z Fold 7, and is set to easily take over from its predecessor at the top of our rankings of the best foldable phones.
Samsung Galaxy Z Fold 8 Ultra review: Price and availability- Even more expensive than the Z Fold 7
- Still a good value for a two-in-one device
I guess it was inevitable that with the price of RAM and components rising, the latest Galaxy Z Fold would be more expensive than its predecessor.
The new Galaxy Z Fold 8 Ultra starts at $2,099.99 / £1,899.00 / AU$2,999. If anything, the $100 bump in the US is less than I thought it might be; double that would not have been surprising.
Released alongside the smaller Galaxy Z Fold 8 ($1,899.99 / £1,699 / AU$2,699 ), the Z Fold 8 Ultra has an inarguably premium price, though it seems more reasonable when you consider that this is a two-in-one device.
The Z Fold 8 Ultra is available in Graphite, Cream, Green Shadow, and my test unit's Violet Shadow.
- Value score: 4 / 5
Here's a breakdown of the Samsung Galaxy Z Fold 8 Ultra's key specs:
Dimensions:
Closed: 158.4 x 72.8 x 8.9mm
Open: 158.4 x 143.2 x 4.1mm
Weight:
215g
Display:
Cover display: 6.5-inch AMOLED
Main display: 8.0-inch AMOLED
Resolution:
Cover display: 1080 x 2520
Main display: 2504 x 2256
Refresh rate:
Cover display: adaptive 1-to-120Hz
Main display: adaptive 1-to-120Hz
Peak brightness:
Cover display: 2,400 nits
Main display: 3,000 nits
CPU:
Qualcomm Snapdragon 8 Elite Gen 5 for Galaxy
RAM:
12GB / 16GB
Storage:
256GB / 512GB / 1TB
OS:
Android 17 / OneUI 9
Cameras:
200MP F/1.7 main, 50MP F/1.9 ultra-wide
Selfie Camera:
Main screen: 10MP F/2.2
Cover screen: 10MP F/2.2
Telephoto
10MP F/2.4 3x optical
Battery:
5,000mAh
Charging:
45W wired, 20W wireless, Wireless Powershare
Colors:
Graphite, Cream, Violet Shadow, Green Shadow
Samsung Galaxy Z Fold 8 Ultra review: Design- Folded, it looks virtually identical to the last model
- Easier to unfold
- Unfolds flat... finally
- The crease is almost gone
There's little argument that the Samsung Galaxy Z Fold 8 Ultra is, despite the new superlative, a visually iterative update. Folded, it's hard to distinguish it from the Z Fold 7.; it's virtually the same size and weight, meaning it's still lighter than the Galaxy S26 Ultra. It's still an Armor Aluminum Frame, though the Gorilla Glass Ceramic 2 from the last Fold 6.5-inch cover display is now Gorilla Glass Ceramic 3. On the back is Gorilla Glass Victus 2.
It's not until you unfold the handset, or rather attempt to unfold it, that the design changes, er... unfold.
First, there's a new chamfered edge along the fold inseam, which makes it far easier to insert the tips of your fingers and unfold the phone. Samsung told us it made this change because some users complained that the 4.2mm thick Z Fold 7 was difficult to open. The even thinner 4.1mm Z Fold 8 Ultra is, in fact, easier to open. I might even call it a pleasure.
Speaking of pleasure, there's another tiny little change that I really love. The power/sleep/fingerprint/Gemini button is no longer recessed, and is now so much easier to find by touch.
The biggest design upgrade is almost purely mechanical, and Samsung pulled it off by swapping out various materials and hinge designs to finally achieve a virtually flat and almost creaseless unfolded body and display. That is to say, the screen and chassis are flat, although if you lay it on a table, that ample camera bump will keep the whole thing rocking.
Lance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureInside the 4.1mm body — it's so thin it only just accomodates the USB-C port — Samsung has re-engineered the rigid titanium panel along the hinge strip, transforming the perforations into an almost 'A' shape, which allows for easier and more reliable bending. That panel is paired with a new titanium alloy sheet (it's thinner and stronger than the old polymer one). Together, they make the folding and unfolding processes better, and, when the phone is fully unfolded, allow for a virtually flat frame and the least noticeable crease yet. I'll have more to say about this design in the Display section below.
(Image credit: Lance Ulanoff / Future)As I've been carrying around the Z Fold 8 Ultra and opening and closing it, people have consistently asked how long the phone might last, and whether it fatigues at the bend. I reply that, in the course of using and testing Samsung's folding phones over the last seven years, I've found them surprisingly resilient.
This model is actually water-resistant (IP48; yes, I ran it under water, yes, it survived, and perhaps enjoyed it), though I would not take it to the beach if it's not in a sealed case. When I traveled to South Korea last month. Samsung showed me how it torture-tests these devices, including folding and unfolding them thousands of times. I'm quite confident that with normal use the Z Fold 8 Ultra will last for years — at least through all seven years of Samsung's promised OS and security updates.
- Design score: 5 / 5
- Main screen is almost perfectly flat
- Responsive, colorful, main screen has sunlight-beating brightness
- Cover screen is wide enough to be useful as a regular phone
- I wish the cover screen was a little brighter
Samsung now has enough experience making folding phones to be well into the refinement phase. The Galaxy Z Fold 8 Ultra is already thin and light enough, and the last model, the Z Fold 7, folded the screen so tightly that there was barely any gap between the two halves of the display. Now, though, the new display technology allows the screen to unfold virtually flat and, thanks to both the hinge engineering and new anti-reflective coating, make the crease very close to invisible.
The crease is not entirely gone. When I run my finger over the center of the 8-inch display I can feel it, and if I view the screen at the right angle, I can just make out the slight half-inch-wide indent.
(Image credit: Lance Ulanoff / Future)Still, everything looks fantastic on the tablet-size display, which now sports a 2504 x 2256 resolution, a slight upgrade from the Z Fold 7's 2184 x 1968 resolution. I played games, watched video, viewed countless photos, used it for full-screen turn-by-turn navigation in maps (it's easy to get lost in London, which I was visiting for the Galaxy Unpacked event at which this phone was launched), and played more games.
(Image credit: Lance Ulanoff / Future)If I had one real criticism of the main screen, it's that it no longer has a digitizing display — but this is not an issue new to the Z Fold 8 Ultra. Samsung dropped this feature on the Galaxy Z Fold 7, and there were rumors that the S Pen-supporting digitizing layer would return with the 8, but that didn't happen. Without it, and a special S Pen, to protect the phone’s flexible display, I really don't feel comfortable drawing on it with a universal rubber stylus (which you can buy from third-party providers, but which lack the S Pen’s unique features.
(Image credit: Lance Ulanoff / Future)The cover screen reads as a 6.5-inch display, but that's measured diagonally, and the reality is that it's not as wide as the display on the Samsung Galaxy S26 Ultra. The Z Fold 8 Ultra cover display is about 2.5 inches wide, and the S26 Ultra is about 2.75 inches.
I still find the Z Fold 8 Ultra's cover display more than usable. The virtual keyboard is large enough for me to comfortably type on it. Websites, texts, and social media look good on it. The only issue I found is that its max brightness is not as strong as the inner display (3,000 nits, max), and it was occasionally overwhelmed by direct sunlight.
It's also a good-looking screen across multiple content types and activities thanks to the variable refresh rate that can sit anywhere from a power-sipping 1Hz up to a silky smooth 120Hz.
- Display score: 5 / 5
- Excellent collection of cameras
- Best image processing I've seen from Samsung to date
- Finally, a 50MP ultra-wide camera
- Wish the telephoto was 5x optical
As with most of the Galaxy Z Fold 8 Ultra, Samsung didn't completely reinvent the photography suite. Instead, we have a strong collection of cameras with some key improvements, chief among them a much more powerful 50MP ultrawide and the overall updates to the ProVisual engine for some of the best photography I've ever achieved with a Samsung smartphone.
Yankee StadiumLance Ulanoff / Future3x zoom moon shotLance Ulanoff / FutureUsing digital zooom. Not the detail on the moon, which is likely added by AI.Lance Ulanoff / FutureI've spent more than a week photographing everything from the iconic landmarks of London to a soccer match between Wrexham and Liverpool at New York's Yankee Stadium. In virtually every scenario, the Galaxy Z Fold 8 Ultra's cameras did not disappoint. Samsung's ProVisual engine is finally focused on color fidelity, and while I can use Galaxy AI Photo Assist to alter mu images to my heart's content, I applaud Samsung's focus on photographic reality.
To recap, here are all the cameras (five of them!) you get:
- 200MP wide
- 50MP ultra-wide
- 10MP telephoto (3x optical)
- 10MP selfie (main screen)
- 10MP selfie (cover screen)
Like most high-megapixel main cameras on phones, the Galaxy Z Fold 8 Ultra's 200MP and 50MP cameras default to a binned mode, squeezing multiple pixels into one (the main 48MP camera on iPhone 17 Pro, for instance, similarly defaults to 24MP). The result with the Z Fold 8 Ultra is 12MP photos, unless you choose 50MP or 200MP. The majority of my photos were shot with the 12MP default.
Lance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffEven though Apple defaults to a higher resolution, Samsung's overall higher megapixel count means it's theoretically squeezing even more information into each photo. In anecdotal side-by-side shots, Samsung's photography is now as good as, and sometimes better than, what I can get with the iPhone 17 Pro Max.
Portrait-mode photography is excellent, with the processing effectively separating wisps of hair from the bokeh background.Lance UlanoffLance Ulanoff / FutureLance UlanoffPortrait-mode photos have become a real highlight in Samsung's photographic arsenal. From the very first photo I took of my colleague Axel Metz (above) to a shot of my son and his girlfriend, I've been impressed with the phone's uncanny ability to capture every detail and strand of hair while warmly blurring the background. My favorite shot is this one, of one of the Beefeaters who guard the Tower of London guards. This gentleman was kind enough to pose for me, and I think even he was surprised at the result.
(Image credit: Lance Ulanoff)While I was generally pleased with the telephoto performance, I'm not thrilled with the underpowered, 10MP 3x optical zoom lens. It obviously pales in comparison to the iPhone 17 Pro Max's 48MP 4x optical zoom. Those extra megapixels also help the iPhone produce unprocessed 8x zoom sensor crops. After 3x on the Z Fold 8 Ultra, everything else gets some digital enhancement. Most of the images I've shared are optical-strength only.
Lance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffOne lens that's gotten a big upgrade with the Z Fold 8 Ultra is the ultra-wide, which is now a 50MP lens. This helps improve the quality of those wide-angle shots, and macro images, as well. There's some real drama in these ultra-wide images, and I'm impressed with how even the edges remain sharp.
Lance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffNight photography is also strong, and I was often pleased with how much light, color, and detail the Z Fold 8 Ultra could find in even the darkest environments.
Lance UlanoffLance UlanoffLance UlanoffLance UlanoffOn the video side, the Z Fold 8 Ultra lives up to its name, letting you shoot and edit 8K video. I waded into my pair of hydrangea bushes and dozens of buzzing bees to capture some cool, up-close 8K footage that I was then able to fully edit inside the phone's camera app, including adjusting brightness, and even using the AI-powered audio scrubber to bring down the wind and sound of leaves brushing against the microphones.
(Image credit: Future)One of the other highlights is the Dual Camera shooting mode, which, while using a front and back camera, produces a rather distinct result. Instead of the 10MP selfie-camera image appearing as a picture-in-picture window on the front-facing camera video — as it does when using the similar Dual Capture mode on Apple’s latest iPhones — the Galaxy Z Fold 8 Ultra creates a split-screen view.
A post shared by Lance Ulanoff (@lanceulanoff)
A photo posted by on
I used this mode a few times in London, most memorably at the Royal Palace and at Borough Market, where I shot the famous strawberry stand. It's a great effect, and creates some really engaging videos.
Lance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffLance UlanoffThe other most notable video feature is a Galaxy AI enhancement called My FanCam. As the name implies, it lets you focus on your favorite video subject. The example we saw involved video of a soccer game. On playback in the camera app, you select the Galaxy AI icon and then the My FanCam option. The system identifies all the moving people (it doesn't work with animals or anything other than humans). You select the one you want, and the video tracks that subject. But the best part is that it can auto-crop around the subject in the aspect ratio of your choice — perfect for sharing on social media.
I ran a couple of tests with My FanCam, and it worked perfectly every time.
@techradar ♬ original sound - TechRadar- Camera score: 4.5 / 5
- One UI 9 is clean, elegant, and intuitive
- Lots of AI, but none of it gets in the way
The Samsung Galaxy Z Fold 8 Ultra arrives with Android 17 and OneUI 9. It's a great combination. Overall, OneUI is a great Android overlay, because it's useful without getting in the way of Android. There are still some odd choices, like the existence of a Galaxy web browser that I never use (does anyone?), but some of the native Galaxy features are actually quite useful.
It's also a system with multiple AI platforms: Bixby, Galaxy AI, and Gemini. Of the three, Bixby, which can answer general questions and step into controlling key phone features, got the least amount of use. I do like though for it for how it lets you dig down to settings like brightness control without ever digging into settings. I could simply talk to Bixby and tell it what I wanted, like when I asked, "Can you make sure the brightness is set to 100%?" Bixby opened settings, brightness control, and pushed it to 100%.
Galaxy AI is a far richer collection of native AI tools, ranging from on-screen awareness to voice-driven photo editing.
All I said was, "remove the old man"FutureFutureI had some fun editing photos with my voice — and I was shocked by how good the results were. With the above photo pair, I simply asked Galaxy AI to "Remove the old man". There was no hesitation, and in about 20 seconds I had the second image. I was gone, and Galaxy AI filled in the rest of my friend's arm, turned a small satchel he was carrying into a shoulder bag, and fleshed out the guy behind me. It looked like reality.
There's also a whole collection of Now features, including the "your world on a single screen" Now Brief, which now adds custom cards, and Now Nudge, which pops up contextual guidance for things like multitasking and automatically noticing that you might want to add an appointment to your calendar. I couldn't get Now Nudge working for me in messages in RCS — I suspect it may take some time for the system to learn about me and start making these suggestions.
Lance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureOn the quality-of-life front, the QuickShare updates that make sharing between this Android Phone and iPhones (and Macs) possible have been a revelation. I feel as if the world's been thrown open, and I can only say it's about time. Okay, I know that's a bit much, but this feature update is bigger than you might think.
Gemini is especially powerful if you give it access to personal details and data inside apps like calendar and mail. When I asked Gemini, without preamble, "Where am I going today?" it looked in my emails and gave me my schedule and location for my next meeting. Impressive.
Features like Circle to Search can be summoned by holding down the virtual home button and then circling on-screen objects. The results are quick and accurate
I can also hold down the power/sleep button to summon Gemini. Six months of access to the more powerful Pro version comes with the phone; after that, you have the most basic Google AI, which is still pretty good.
- Software score: 4.5 / 5
- Powerful Qualcomm Snapdragon 8 Elite Gen 5 for Galaxy
- Performance is peerless
- Back can get a little hot
This is a powerful and smoothly performing smartphone. Qualcomm's latest silicon virtually matches the iPhone 17 Pro Max's A19 Pro Geekbench 7 single- and multi-core scores. The numbers, by the way, appear lower than phones tested last year with Geekbench 6 because, in the upgrade to 7, Geekbench completely recalibrated how it measures performance.
In the real world, I have yet to run into a task or situation the phone couldn't handle. Games like Asphalt Legends looked and ran perfectly. The Galaxy Z Fold 8 Ultra can handle multitasking with ease — you can have up to four apps open simultaneously on the main screen — and editing 8K video is seamless.
Lance Ulanoff / FutureLance Ulanoff / FutureLance Ulanoff / FutureThere were times when I could feel the back of the phone getting warm, but not so much that it became uncomfortable to hold.
In other performance areas, like audio, the speakers can get loud but with virtually no bass; they can also sound a bit tinny. Call quality is excellent, though my first few call tests were with phone scammers who somehow instantly found the new number.
5G speeds and WiFi 7 connectivity were strong and, with enough signal, fast. I've also enjoyed a solid Bluetooth 6.0 connection to my Samsung Galaxy Buds Pro 2 and Galaxy Watch Ultra.
Samsung Galaxy Z Fold 8 Ultra benchmark scoresSamsung Galaxy Z Fold 8 Ultra
Geekbench 7 single-core
1835
Geekbench 7 multi-core
4476
3D Mark Wildlife Extreme
6632
3D Mark Wildlife Extreme Low
1660
3D Mark Solar Bay Extreme Unlimited
1366
- Performance score: 4.5 / 5
- 5,000mAh battery
- 45W wired charging
- All-day battery life
Even though this is a thinner Z Fold than last year, the battery is now a larger 5,000mAh, a feat made possible by the thinner titanium alloy, which replaced the thicker polymer layer and the new silicon-carbon battery technology (more power in less space).
That battery gave me, with constant, all-day, mixed use in bright sunlight, 12.5 hours of battery life. In our lab tests, we got a solid 14 hours. It's an all-day champ, to be sure.
The phone can also charge quickly with a 45W fast charger — I almost 50% in 15 minutes and 76% in half an hour — while wireless charging supports 20W charging. You can also put another Galaxy phone on the back and share a little power via Wireless PowerShare.
- Battery score: 5 / 5
Attributes
Notes
Rating
Value
The price of Samsung's best foldable just keeps going up
4 / 5
Design
Thin, flat, almost creaseless and just beautiful. I do wish the camera bump didn't mess up the on-desk flatness
5 / 5
Displays
Excellent displays and the upgrades to the main screen (flat, less of a crease, antireflection) make it a great content and productivity device.
5 / 5
Camera
Excellent collection of cameras. The ultrawide update is especially welcome. I do wish the telephoto was a bit stronger (optically). Powerful features like Photo Assist and My FanCam are strong additions.
4.5 / 5
Software
Clean, well-organized platform with lots of useful, baked in, AI features like Now Brief. Lots of AI.
4.5 / 5
Performance
Qualcomm's best mobile processor (tweaked for Samsung) does not disappoint.
4.5 / 5
Battery
5,000mAh battery. All day battery life is virtually guaranteed.
5 / 5
Buy it if...You want the ultimate foldable phone
Sure, the Galaxy Z Fold 8 is adorable, but no other Galaxy folding phone has this collection of cameras, vast screen size, and thinness. It's the full package.
You want all-day battery life
The cutting-edge silicon-carbon 5,000mAh battery is a true workhorse.
You want the best photos
The Galaxy Z Fold 8 Ultra is unquestionably the camera leader among its Galaxy Z8 cousins, but it also has one of the best camera phone arrays I've used.
You're on a bugdget
This is an exciting folding phone, but you will pay dearly. If cost is an issue, you may want to look to the Z Fold 8 or Z Flip 8.
You want more zoom
As much as I like these cameras, I still want a 4x optical zoom, and would love a 5x. Maybe next year.
Samsung Galaxy Z Fold 8
The Galaxy Z Fold 8 promises the same level of performance and AI as the Z Fold 8 Ultra, but in a cuter form factor and at a better price.
Read our Samsung Galaxy Z Fold 8 review
Motorola Razr Fold 2026
This foldable manages to match the Z Fold 8 Ultra in power and battery technology (but has a larger battery). The design is questionable, but it's generally a very good foldable that will save you money over the Z Fold 8 Ultra.
Read our full Motorola Razr Fold 2026 review
How I tested the Samsung Galaxy Z Fold 8 UltraI've been working, playing, and traveling with the Samsung Galaxy Z Fold 8 Ultra for almost two weeks. I've tried to use it as I would any smartphone, but with the added benefit of the large screen.
This is my seventh year testing foldable phones, and I am constantly astonished by the progress that's being made with the form factor, especially by Samsung. The Z Fold 8 Ultra is the full expression of its smartphone ideas and, as I carried it with me and pulled it out to take photos in London, Manhattan, and at home, I was constantly surprised and pleased with its performance.
More broadly, I've been testing phones for over 22 years and technology for 40. I use my wealth of experience to put products like the Z Fold 8 Ultra in context and perspective. As I see it, this is the best folding phone on the market right now.
Read more about how we test
First reviewed August 2026
Lioness season 3 is only one episode down, but it's already off to a very chaotic start.
We now know that its eight episodes will comprise two simultaneous timelines: one, Joe being kidnapped by unknown Russian operatives and, two, the six months leading up to her capture.
The Lionesses were tasked with extracting a target out of a Ukrainian war space when foreign operatives "offered" him up to the CIA. However, it's now looking like this was a trap for something much bigger.
Will we find out what's actually going on this week? Probably not. Even so, when does Lioness season 3 episode 2 arrive on Paramount+?
What time can I watch Lioness season 3 episode 2 on Paramount+?Lioness season 3 episode 2 will drop on one of the world's best streaming services in the US and Canada on Sunday, August 9 at 12am PT / 3am ET.
Here's when it will be released in other nations globally:
- US – 12am PT / 3am ET
- Canada – 12am PT / 3am ET
- UK – 8am BST
- India – 1:30pm IST
- Singapore – 4pm SGT
- Australia – 7pm AEDT
- New Zealand – 9pm NZDT
As mentioned, Lioness season 3 will have a total of eight episodes, with new entries airing weekly. That gives us the following schedule:
- Episode 1: out now
- Episode 2: August 9
- Episode 3: August 16
- Episode 4: August 23
- Episode 5: August 30
- Episode 6: September 6
- Episode 7: September 13
- Episode 8: September 20
- Microsoft aims to cut down on "tokenmaxxing" by employees
- New guidelines will look to control AI token use to focus on ROI
- This is despite Microsoft reporting record financial results recently
Microsoft has apparently been forced to introduce limits on how much AI usage its engineers are allowed followed reports that some have been taking things to extreme.
The software giant is looking to cut back on "tokenmaxxing" within the company - where employees use far more AI tokens that may be necessary.
In an email seen by 404 Media, Microsoft warned employees that new limits on token usage would be introduced as it looked to focus on getting the most out of its AI platforms.
New guidelines“As we accelerate our use of GitHub Copilot to deliver on our goals, we all need to be aware of how we consume tokens,” the email to employees from Jay Parikh, an executive vice president at Microsoft said.
“Tokenmaxxing is not what we are optimizing for," he continued. "I want all of us focused on maximizing outcomes that move the needle for our customers and our business.”
“As such, we are updating our internal guidance and managing token spend with the same discipline we apply to every other critical resource."
In a bid to achieve “get greater value from our token investment”, Parikh went on to say Microsoft is making access to the cheaper OpenAI GPT-5.6 model the default model for internal use.
Employees were also reminded (via a link to updated internal Copilot guidelines) that as of July 2026 Microsoft divisions have an “AI token budget target,” with employees also able to track their individual AI spending.
“While there is no target spend value being shared at this time. The data shows that many engineers spend in the range of hundreds of dollars a month to a few thousand dollars in tokens,” 404 Media reported the guidelines as saying.
Parikh noted that Microsoft does not want to impair the company’s progress towards becoming “AI-first,” and that it will keep learning and adjusting its AI policies as models and products evolve.
“We are not optimizing for fewer tokens,” he said. “We are optimizing for more impact per token."
The restrictions may come as a surprise to some Microsoft employees, given that the company recently reported yet another bumper financial quarter, and in that respect should have money to splash out on AI usage.
However it is the latest step by Microsoft as it looks to focus internal AI usage. In May 2026, it was reported the company was reportedly canceling most of the Claude Code license it uses internally, with engineers being told to use GitHub Copilot CLI, with users given a June 30 2026 deadline to remove Claude Code from their workflows.
However tokenmaxxing has proven to be an issue at other tech giants - perhaps most notoriously at Uber, which was forced to admit it had exhausted its entire annual 2026 AI coding token budget in just four months due to massive employee adoption of agentic tools like Anthropic PBC's Claude Code and Cursor.
Amazon also recently revealed it had spent $1.8 million on an internal Claude Sonnet deployment intended for matching author details with product listings, after it ballooned far beyond its planned budget and was ultimately given menial tasks to do.
With back-to-school season in the air, you might be looking to stock up on a new Mac for college. But doing so can be a confusing affair, with many different models and configurations to choose from, and you can waste a lot of time trying to work out what’s best for your needs.
To skip all the bother, just give our guide a read. We’ve put together everything you need to know, from which Mac is best for you to how to get those all-important student discounts. You’ll be kitted out and ready to go before you know it.
How to choose the right specs for your Mac(Image credit: Apple)When you buy a Mac, there are a few specifications you need to choose, including the chip, memory and storage. You’ll also need to pick a color and, with some Macs, extra options like the display coating and power adapter.
The first three factors — chip, memory and storage — are the most important, and the ones where there’s the most to learn.
We’ll start with the chip, which determines how powerful your MacBook is. More powerful and more recent chips will generally be better, and the higher the chip’s number, the more recent the generation it belongs to.
The chip usually comes in a few different tiers. There’s the baseline option, such as the M5, M4, M3 and so on, then Pro chips like the M5 Pro, M4 Pro and M3 Pro. Then there are Max chips, including the M5 Max and M4 Max. The most powerful level is the Ultra, and it’s limited to desktop Macs like the Mac Studio. The latest version there is the M3 Ultra.
Apple’s chips combine CPU, GPU and memory onto a single system on a chip (SoC). The CPU and GPU will have their own core counts, and the more cores you get, the better able the chip is to handle difficult workloads.
So, which do you need? The baseline chips will be more than enough for essay writing and basic research. On the other hand, you might want a Pro or even Max chip if you’re doing anything that involves complex apps, data processing or graphics work.
As for memory, this is built into the SoC, but you can increase or decrease it separately from the chip selection when configuring your Mac (although it’s worth noting that some memory amounts require upgrading to a more powerful chip). More memory helps speed up multitasking and is also useful for demanding workloads.
The amount of storage you choose depends on how long you want to keep the Mac, how much of a digital hoarder you are, and the type of things you’ll be storing. Run lots of large apps or work with high-resolution video files? You’ll need lots of storage.
Here, it’s best to err on the side of caution, as it’s very difficult (although not impossible) to upgrade your storage once you’ve bought your Mac. While it can be done, it’s not for the faint of heart.
One alternative is to plug in an external storage drive to give yourself extra space. This isn’t optimal, though, as it’s one more thing to carry around, and if it becomes accidentally disconnected, you could suffer data loss.
Make sure to regularly back up your Mac, either using an app like Time Machine or with a cloud service like Backblaze.
How to get Apple’s student discount(Image credit: Apple)Apple has a special education store for students, teachers and more. Here, there are various savings available, depending on the product. For example, the MacBook Neo is priced at $599, down from the regular $699 price (bringing it back to the price it had when it launched, before Apple’s recent price rises).
Apple says that student discounts are available to “current and newly accepted college students and their parents, as well as faculty, staff, and homeschool teachers of all grade levels.” You’ll need to verify your student status, which you can do online or at an Apple Store.
Student savings can be had on more than just Macs. In total, the discounts apply to:
- Mac
- iPad and select accessories
- Apple Watch
- Displays
- AppleCare+ for Mac
- AppleCare+ for iPad
- AppleCare+ for Display
Apple says to look out for the mortarboard icon, which indicates a product bought through the education store is discounted.
You get all the same perks as if you were to buy through Apple’s regular store. You can pick components and colors, trade in an old device, personalize your device with engraved monograms and emojis, pay in full or over time, and more.
The three best Macs for back to school(Image credit: Apple)Here, we’ve picked three Macs that we recommend the most for students. All three are Apple laptops, which will be perfect for carrying to lectures and around campus. Note that if you don’t have many lecture hours or are studying from home, you might want to consider the Mac mini, which starts at $699 in Apple’s education store. It’s a brilliant computer with an affordable price tag, but as it requires an external display, it’s not ideal for lectures and seminars.
MacBook Neo(Image credit: Future)The MacBook Neo is the perfect option for students on a budget. The default model comes with 256GB of storage, an Apple A18 Pro chip and 8GB of memory. It costs $599 from Apple’s education store, marking a saving of $100 over the regular price.
If your school or college work mostly involves writing essays and researching online, the MacBook Neo is exactly what you need and will be able to handle everything you do. The only question is whether you should upgrade to 512GB of storage — doing so also adds a Touch ID button to the keyboard, which is a handy way to log in and verify purchases.
MacBook Air(Image credit: Lance Ulanoff / Future)Stepping up, a solid mid-range option would be the 13-inch MacBook Air with M5 chip. You can choose between M5 chips featuring an eight-core GPU or a 10-core GPU, but we’d stick to the former — if you need more graphical power, the MacBook Pro will be a better option, which we’ll come to later.
Next, you can choose between 16GB, 24GB and 32GB of memory. The option you choose here depends on your workloads, but generally 16GB will be enough for most students. Again, if you need more, go for the MacBook Pro.
As for storage, the entry-level M5 chip is locked to 512GB of storage. To get more, you’ll need to upgrade to the higher-end M5 chip, which ups the price. Consider how much storage you need before making this move — if you have an existing computer, see how much you’re using with that. Bear in mind that you can’t change the internal storage after purchase, so if in doubt, opt for more.
Finally, there’s the matter of the charger. The M5 MacBook Air comes with a 40W charger that can handle loads up to 60W. If you want fast charging, which Apple says can power your laptop up to 50% in about 30 minutes, you’ll need the 70W option. This adds $20 to the cost.
MacBook Pro(Image credit: Future)For the most demanding student workloads, you’ll want to look at the MacBook Pro. Our recommendation is to start with the 14-inch model with M5 Pro chip. Skip the nano-texture display for now — it’s a nice addition, but unnecessary for students on a budget.
Do upgrade to the M5 Pro chip, however, or the M5 Max if you know you’ll be doing really heavy work that requires a lot of computing power. Unlike the MacBook Neo and MacBook Air, the MacBook Pro has built-in fans that mean it has a far superior cooling system that is much better equipped for demanding jobs. With that at your disposal, you can pick a beefy chip if you need it.
With the M5 Pro, your choices for memory are 24GB or 48GB. The former should be enough for most, but if you’re unsure, step up to the higher amount. While there are other options, including 36GB, 64GB and 128GB, these require swapping the M5 Pro for an M5 Max and paying a much higher price.
The configuration we’ve picked so far comes with 1TB of storage included, which is enough for most people. That said, anyone working with large photos and videos, high-grade apps or machine-learning models will likely need more space.
Finally, there’s the issue of the charger. The default option is 70W, which isn’t enough for fast charging (the MacBook Pro uses much more power than the MacBook Air, where 70W is enough for fast charging). Seeing as the 96W upgrade costs just $20 more and enables both fast charging (empty to 50% in about 30 minutes) and High Power Mode, we think it’s worth it.
Thanks to rapid developments around AI, cybersecurity pros seem due for a reminder that the threat landscape has fundamentally changed. This time it's OpenClaw ringing the alarm bells, but in truth we see the same cycle repeating itself time and time again, just with new technology.
Within days of OpenClaw’s open-source, researchers discovered exposed management interfaces and malicious "skills" packages designed to trick users into installing compromised functionality.
While the legitimate security community did what it always does, innovate rapidly, attackers moved just as fast.
That reality should force organizations that haven’t already done so to rethink a dangerous assumption: the belief that if something malicious appears inside our environment, internet security products will detect it quickly enough to stop serious damage.
That assumption is unrealistic in the world of AI-driven automation and open-source agent ecosystems.
The better question is: ‘Why should unknown software be allowed to run in the first place?’
Uncontrolled AI adoption is the core issueOpenClaw highlights the much larger issue of Shadow AI running across organizations. Employees are downloading local AI agents, experimenting with open-source models, installing community-developed skills and connecting all of these tools directly to corporate resources.
Unlike traditional SaaS applications, many of these agentic platforms execute directly on endpoints. They request filesystem access, interact with browsers, connect to cloud services and automate business workflows.
Every new skill, extension or plugin expands the attack surface. While AI usage is an important progression of technology, the problem that’s emerging is that organizations frequently have little visibility or control over which AI tools are entering their environment, let alone what they're allowed to do once they arrive.
Detection can't keep up with automated attacksThe industry continues investing enormous resources into detecting threats faster and there is absolutely value in that; but OpenClaw illustrates why detection alone cannot be the primary strategy.
By the time a detection platform identifies suspicious behavior, an AI agent may already have accessed sensitive files, authenticated them to cloud services, downloaded additional components or exposed confidential information. Instead of asking how quickly we can detect something, organizations need to first ask whether it should have been able to execute at all.
Control AI without disrupting the businessTo effectively mitigate AI-driven cybersecurity threats, the industry must embrace a Zero Trust approach that moves from an allow-by-default to a deny-by-default posture.
In the context of Shadow AI, application allowlisting ensures only approved agents run inside your environment, while application containment further limits what those trusted agents are allowed to do.
This way, if an agent is compromised, any unnecessary access to files, memory, scripting engines, networking functions or other applications is blocked entirely.
Together, these controls enforce the boundaries your business already intended to have.
One of the biggest misconceptions surrounding Zero Trust is that it requires organizations to lock everything down overnight. This isn’t true.
When done correctly, application control allows organizations to understand what's already running, establish normal operating behavior and gradually enforce policies without disrupting users.
Define what good looks likeCybersecurity has traditionally focused on identifying bad behavior, yet attackers are constantly inventing new forms of it. A more sustainable model is to define what good looks like for your organization.
If a script or application like OpenClaw is not explicitly approved inside your environment, it shouldn’t be allowed to execute.
The bottom line is that if an AI agent doesn't require access to sensitive directories, cloud resources, PowerShell or credential stores, those interactions shouldn't be possible. This deny-by-default approach dramatically reduces the opportunities available to both attackers and compromised applications.
Rather than chasing an endless stream of new threats, you're enforcing known business requirements.
The leadership lessonOpen-source innovation has enormous value and will continue driving technological progress, but every major cyber incident teaches a lesson.
The Open Claw incident shows that organizations can no longer afford environments where any new tool is free to operate with minimal oversight. Leadership teams need to stop measuring success by how quickly they respond to breaches and start measuring how effectively they've reduced the opportunity for one to occur in the first place.
Open-source AI ecosystems will continue evolving at remarkable speed, new capabilities will continue to emerge. In tandem, attackers will continue to innovate their own methods. Fortunately, your security strategy doesn't need to change every time a new threat emerges.
The principles behind hardening your environment remain the same: know what belongs in your environment, and allow only what you've explicitly approved. Restrict what trusted applications can do, and treat every request for access as untrusted until proven otherwise.
Organizations that embrace that philosophy won't just be better prepared for OpenClaw, they’ll be ready for whatever comes next.
We've reviewed, rated, and ranked the best firewall 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
Not too long ago, one of the world's most data-driven companies admitted that its AI spending was becoming “harder to justify”. Uber's President and COO, Andrew Macdonald, told the Rapid Response podcast that the company had blown through its entire 2026 AI budget in roughly four months, with around 5,000 engineers leaning on Anthropic's Claude Code.
Uber isn’t alone. Forrester research found that enterprises are deferring around 25% of planned AI spend to 2027, as CFO scrutiny over ROI intensifies. And McKinsey's State of AI report summarized that while 62% of companies are experimenting with AI agents, only 23% have scaled them in even a single business function.
While these may look like the statistics of a technology that isn't working, they’re actually the statistics of a technology being used in the wrong way.
If we rewind back to 1987, Nobel laureate economist, Robert Solow, observed something that many at the time really resonated with: “You can see the computer age everywhere but in the productivity statistics.” Computers were everywhere across the innovative businesses that had invested heavily in them, but the productivity numbers didn’t move.
The returns only materialized years later, once organizations stopped bolting computers onto old processes and started fundamentally redesigning how they worked.
Many executives, alongside Uber’s COO, are at exactly that inflection point with AI - people are using it, running out of budgets to maintain usage, and at the same time, not really seeing the productivity boost they were hoping for.
This is the Solow Paradox repeating its course, which begs the question: will enterprises learn from previous mistakes?
The flatline behind the hypeThere’s a critical distinction that most enterprise leaders are still failing to make: AI activity is not the same as AI maturity. You can run 40 pilots, adopt six platforms and report impressive usage statistics, and still be no closer to measurable business value.
Uber found this out in painful, public fashion, and has since openly questioned whether the rising cost of AI token usage is translating into proportional productivity gains. What makes Uber's situation instructive is not just the financial exposure, but how the organization approached adoption. Internal leaderboards were introduced to rank teams by AI tool usage, with the incentive being to use more tools.
The outcome was more usage, but the business impact slowly became harder to justify.
This is what happens when you gamify adoption without redesigning the workflows underneath it. You optimize the tool usage metric, not the business outcome. Uber has now joined several other top organizations, including Microsoft, Meta and Amazon, in capping AI usage to tackle the issue.
What’s interesting here is that when AI token usage is unconstrained, activity becomes the proxy for progress, but when it’s capped, organizations are forced to confront a harder question: what is each token actually producing?
In that sense, token spend behaves like an economic mirror. It scales immediately with adoption, while productivity only improves when workflows are redesigned. The gap between the two is where most AI ROI disappears.
The real culprit: Individual task optimizationWhen AI tools are deployed at the individual level, they tend to optimize the task, not the workflow.
A developer writes code faster, a marketer drafts copy in a fraction of the time, or a data analyst generates a summary report in minutes rather than hours.
All of these examples are real gains, but if the code still sits in a review queue for four days, if the draft still passes through three rounds of manual approval, or if the report still requires someone to manually transfer it into a decision-making dashboard - the time saved will pool at the next bottleneck.
Individual productivity gains that don’t translate into workflow redesign don’t compound. They stagnate, and this is the core of the maturity gap. AI maturity isn't about how many tools you've adopted, or how many pilots you've launched. It's about whether you've moved consistently from opportunity to outcome.
That shift requires a fundamentally different way of working.
Now, the term ‘production-ready AI’ gets used loosely. It’s worth being precise about what it actually means in practice, because most enterprise AI deployments fall short of the bar.
Production-ready AI has four characteristics. First, the output feeds directly into a downstream decision or action without manual transfer. It’s embedded in the workflow, not adjacent to it.
Second, the system has clearly defined failure modes, so the organization knows exactly what happens when the AI gets something wrong, and who’s accountable for remedying it. Third, there’s a named owner responsible for performance, adoption and iteration. Finally, and most critically, the surrounding process has been redesigned - not merely augmented.
The framework that closes the gapSo, how can businesses make this shift in practice? The answer is maintaining disciplined execution.
One way to build that discipline is a structured cadence we call the 3-3-3 framework: three days to prioritize, three weeks to prove value, and three months to launch a first release. The logic is deceptively simple and deliberately structured.
In the prioritization phase, the question is not “what can AI do?”, it’s “which specific opportunity, tied to a specific business outcome, has the right combination of value, feasibility, data readiness, and organizational sponsorship to pursue right now?”
That focus alone eliminates a significant proportion of AI initiatives that consume resources without clear purpose. It’s the antidote to the open-ended experimentation that left Uber burning through its budget before April was out.
In the proof phase, the focus is validation - not in technical terms, but in commercial ones. Can this solution create measurable value for users and for the business? A proof of concept that demonstrates technical possibility without demonstrating business value isn’t a proof of concept, it’s a prototype without a destination.
In the launch phase, the solution moves into a real environment. It integrates with existing systems, gets adopted by the people it was built for and gets measured against the outcome it was designed to improve. Not against token usage and not against adoption rates.
This rhythm isn’t a rigid formula, however. The shape of the work always depends on the business problem, the data environment, the technical complexity and the organization's appetite for change. What the framework provides is momentum and the discipline to keep that momentum anchored in value.
Shifting from adoption metrics to outcome metricsThe most consequential change enterprise leaders can make right now is a measurement decision.
Enterprises that are serious about closing the gap between AI activity and business impact need to make three shifts. The first is from tool deployment to operating model redesign. Rolling out AI tools is table stakes but building a repeatable operating model - a structured path from idea to proof to scale - is the competitive differentiator.
The second shift is from adoption metrics to outcome metrics. Usage rates, logins and token volumes tell you whether people are using the tools, but they don’t tell you whether the tools are working. Define success in business terms from the outset: cost reduction, time-to-decision, revenue impact, customer satisfaction, and build your measurement framework around those outcomes.
The third shift is from centralized experimentation to distributed accountability. The AI factory model - where a central operating model connects business priorities with delivery, adoption and value measurement - works precisely because it distributes accountability across the organization. Every initiative starts with a clear owner, a clear problem and a clear definition of what success looks like.
Winning the AI dayThe productivity paradox Solow identified in 1987 eventually resolved itself. Not because computers got better - though they did - but because organizations learned to reorganize work around the technology, rather than fitting the technology around old ways of working.
The same resolution is available to enterprises deploying AI today. But it requires leaders to make a deliberate choice - to stop measuring success in terms of how many tools have been adopted and start measuring it in terms of how many outcomes have been delivered.
The organizations that win the AI era will not necessarily be those with the largest number of pilots, but rather the ones that build the maturity to turn the right ideas into value and then scale that value with confidence.
That kind of maturity doesn’t happen by accident. It requires structure, discipline and the willingness to ask harder questions about what AI is actually delivering, and what it’s not.
We've featured the best AI chatbot for business.
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
- Aikido researchers uncovers ChainDrop, a Shai‑Hulud variant infecting 1,300+ npm packages with an infostealer
- Attackers compromised GitHub accounts tied to popular libraries (Keyv, Cacheable, flat‑cache, file‑entry‑cache) and pushed tainted releases with 2B monthly downloads
- Malware exfiltrates developer/cloud credentials and secrets to a public GitHub repo; admins should treat affected systems as compromised even after removal
Another Shai-Hulud variant has been discovered in the wild, infecting more than 1,300 npm packages with an infostealer.
Security researchers Aikido reported finding “at least 868 packages (across 1381 versions) that have been compromised by the worm.”
Shai-Hulud is a self-propagating supply chain malware that targets software developers by compromising open-source packages and CI/CD pipelines. It steals credentials, API keys, and access tokens and then uses those stolen secrets to publish additional malicious packages.
What to do in case of an infectionIn May 2026, actors claiming to be associated with the TeamPCP group publicly released the Shai-Hulud worm's source code, saying they were “open sourcing the carnage” and inviting other threat actors to adopt and modify the code. Since then, there were multiple copycat campaigns and variants, including this one which Aikido dubbed ‘ChainDrop’.
Aikido said the attackers compromised the GitHub account of the person maintaining Keyv and Cacheable, widely used open source JavaScript libraries for caching data in Node.js applications. From there, they were able to move into other popular utilities such as flat-cache and file-entry-cache, as well as packages associated with organizations such as Deliveroo, Ornikar, OneReach, Picsart, Qlik, and ServiceTitan.
The malware was pushed directly into the projects’ main branches, and then generated additional package releases. The compromised packages have a combined 2 billion monthly downloads.
Aikido says the infostealer grabs developer and cloud credentials, encrypts them, and then sends them to a public GitHub repository called “Shai-Hulud: Here We Go Again.”
It also steals local configuration files, GitHub PATs, workflow tokens, and other ghp_, gho_, and ghs_ tokens, certain npm tokens, GitHub Actions secrets, AWS credentials, Kubernetes secrets, and more.
The researchers are saying system admins who installed a tainted package should treat their developer workstation or CI/CD runner as compromised, even if they removed the package.
- We're saying goodbye to Google Assistant on more devices
- September 4 is the cut-off point for Android and Wear OS
- The process of switching over to Gemini is pretty seamless
Google Assistant had a good run — it made its debut in May 2016 as part of the Allo app (remember that?) — but Gemini is very much the future when it comes to Google-made AI helpers, and now we finally know when Google Assistant will stop working on Android phones and Wear OS wearables.
As per emails sent out to some Google Assistant users (via 9to5Google), the Android and Wear OS shutdown will begin on Friday, September 4. From that point, it may take a few weeks to reach every user and every device.
It means you won't be able to use Google Assistant at all on Android phones, tablets, and paired wearables. This will apply to Android Auto too, though vehicles with Google built-in, Google Assistant will carry on functioning for a little longer.
The cut-off point was originally scheduled for sometime last year, before being extended, but it now looks as though this is final. Google has been busy making Gemini the default choice on a multitude of gadgets in recent months, including smart home devices and Google TV streamers.
What to do nextThe future is Gemini (Image credit: Google)If you're still using Google Assistant on Android or Wear OS, you need to move over to Gemini at your earliest opportunity. If it didn't already come preinstalled on your device, you can get it from the Google Play Store. A whole new world of AI awaits, covering everything from basic phone commands to image creation.
When you're setting up Gemini, it will guide you through the process of replacing Google Assistant as the default on-board AI. As you're using the same Google account for both apps, the switch should be pretty seamless.
Google Assistant doesn't really keep a chat archive, so there's nothing to sync over in terms of conversations. If you want to see your old Google Assistant interactions, they'll still be available via the Web & App Activity page in your Google account.
You can carry on using the old Google Assistant commands with Gemini, including the "hey Google" alert phrase that gets the AI to listen for instructions. Everything should work as before — though there have been some issues along the way — and your Gemini conversations will be synced between devices by default.
- Some leaked Pixel 11 marketing materials suggest that Pixel Glow could instead be called HiLight
- This new feature might work with both Gemini and notifications
- We've rated the likelihood of this and various other Pixel 11 series rumors
The Google Pixel 11 series is just a week away now, with the phones set for an August 12 launch. But as ever with phone launches, plenty of details have leaked in advance of that date, so we already have some idea of what to expect.
The latest of those leaks details a brand-new feature, reportedly dubbed HiLight, but we’ve heard all sorts of other things recently too about the Pixel 11, Pixel 11 Pro, Pixel 11 Pro XL, and Pixel 11 Pro Fold.
So, below, we've listed the most interesting of the latest Pixel 11 leaks, and rated how accurate we think they are.
1. New HiLight featurePixel Glow is officially called HiLight, I'd say.
— @rquandt.bsky.social (@rquandt.bsky.social.bsky.social) 2026-08-05T15:55:13.608ZFor a while now, we’ve been hearing that the Pixel 11 series might have lights on its rear panel, seemingly to power a feature previously thought to be called Pixel Glow. But now it seems that Pixel Glow might actually be called HiLight.
In a post on Bluesky, reputable leaker Roland Quandt has shared what appears to be an official French marketing image for the Pixel 11 Pro Fold, and it refers to this feature as HiLight, explaining (once translated) that “when your smartphone is face down, HiLight lights up discreetly when your favorite contacts call or when you chat with Gemini.”
So, it looks to be an evolution of the old-school notification light, though we still have questions about it — images show it displaying multiple colors, so we wonder whether specific colors or patterns can be set for specific kinds of notifications.
In any case, Google itself has officially teased this feature too, so while we can’t be 100% sure of the name, we’re certain that at least some Pixel 11 models will have lights on the back.
- Likelihood: 10/10
Not all of the Pixel 11 rumors are good ones, as we’ve also heard that these phones could have higher prices than their predecessors.
Specifically, according to Android Headlines, the Pixel 11 will start at $899, the Pixel 11 Pro at $1,099, the Pixel 11 Pro XL at $1,299, and the Pixel 11 Pro Fold at $1,899.
That, in all cases, represents a $100 increase on last year's models, though in the case of the Pixel 11 and the Pixel 11 Pro, you’re also apparently getting a storage boost, with 256GB said to be the starting capacity rather than 128GB. The other two models, though, come with the same starting capacity as their predecessors according to this leak, so it would just be a straight price rise.
We don’t have UK or Australian price leaks, but we’d expect Google to apply increases globally. For reference, the Pixel 10 starts at £799 / AU$1,349 in those regions, the Pixel 10 Pro at £999 / AU $1,699, the Pixel 10 Pro XL at £1,199 / AU $1,999, and the Pixel 10 Pro Fold at £1,749 / AU $2,699.
So, if this leak is right, you can expect higher prices than that this year, and there’s a good chance this leak is correct, given the ongoing RAM price crisis.
- Likelihood: 8/10
New models usually mean new colors, and the Google Pixel 11 series looks to be no exception, with a whole selection of shades having leaked on Amazon.
These include Midnight (black), Fuchsia (pink), and Moss (green) for the Pixel 11, Dune (a pale pink), Sterling (a grayish lavender), Pine (green), and Light Fog (a pale blue) for the Pixel 11 Pro and Pro XL, and Midnight and Pine for the Pixel 11 Pro Fold.
Those — at least in terms of their names — are all different to what you’ll find with the Pixel 10 series, though in practice some of them will probably be similar.
Still, it’s a largely appealing bunch going by the pictures, with a mix of bright and understated shades. The leaked images of them look fairly official too, so there’s a good chance these are what we’ll get.
- Likelihood: 7/10
The one spec upgrade you can be almost certain of in a new phone is that it will have an improved chipset, and so we fully expect the Pixel 11 series to replace the Pixel 10’s Tensor G5 chipset with a Tensor G6, and leaks agree.
From what we’ve heard across various leaks, this chipset could pair a significant upgrade in CPU performance with a rather small improvement in GPU performance. So, it might not be a complete home run for Google, but it should still deliver plenty of power overall.
- Likelihood: 9/10
As well as costing more, the Pixel 11 could have at least one spec downgrade compared to its predecessor, as Android Headlines reports that it will have just two cameras on the back — a wide and an ultra-wide, with no telephoto.
The Google Pixel 10 had these and a telephoto camera, so this would be quite a reduction in specs if it’s correct.
We would therefore take this rumor with a pinch of salt, since usually we expect upgrades from new models. But it might help Google keep the costs down in the face of rising component prices, and that would still keep it in line with a key rival in the iPhone 17, which also lacks a telephoto camera.
- Likelihood: 5/10
That’s not the end of the possible downgrades, as we’re also hearing from the same Android Headlines report that the Pixel 11 Pro and Pixel 11 Pro XL could come in configurations with just 12GB of RAM, whereas the Pixel 10 Pro and Pixel 10 Pro XL came with 16GB.
There will apparently still be 16GB models, but these will of course cost more, pushing prices up even higher.
A price increase combined with a RAM reduction could be a tough pill to swallow, but this is somewhat believable given that RAM pricing is increasingly out of control.
- Likelihood: 6/10
Thankfully, there is still another major potential upgrade to talk about, as some Pixel 11 Pro marketing images shared by Evan Blass — a leaker with a superb track record — mention 120x zoom. And if the Pixel 11 Pro has this feature, the Google Pixel 11 Pro XL almost certainly will as well.
Now, this wouldn’t be an optical zoom, of course; instead, you can expect it to be some kind of AI-boosted digital zoom. And it’s a believable claim, since not only does the source have a great track record, but the Pixel 10 Pro series already offers 100x zoom.
Still, that extra 20x would mean these phones could zoom further than almost any other handset.
- Likelihood: 8/10
The Lavazza A Modo Mio Jolie Evo is an exceptionally small capsule coffee maker, which brews espresso using the company's own A Modo Mio pods. These vacuum-packed capsules are available in a wide variety of coffee types, roasts, and blends, but unlike Nespresso capsules, there are no elaborate brews with added flavors to choose from.
That's no bad thing in my opinion, because the Jolie Evo produces some of the best espresso I've had from a pod machine, and flavored coffees often tend to taste artificial.
The Jolie Evo is very reasonably priced too (considerably less than the A Modo Mio Smeg and A Modo Mio Deséa), though you'll need to pick up a standalone milk frother too (I recommend either the Smeg Mini Milk Frother or Philips Baristina Milk Frother) if you want to make lattes and cappuccinos.
The controls are extremely simple, with just a single button and a large lever. However, bear in mind that there's no timer, so you'll need to manually tap the button to stop the flow once the machine has dispensed the right amount of coffee.
Overall, it's an excellent compact espresso machine. For me, the only real drawback is that the plastic pods aren't particularly easy to recycle. They can only be dropped off at designated collection points, which can be scarce outside big cities.
(Image credit: Future)Lavazza A Modo Mio Jolie Evo: specificationsName
Lavazza A Modo Mio Jolie Evo
Type
Capsule coffee maker
Dimensions (W x H x D)
4.9 x 8.3 x 13 inches / 124 x 210 x 330mm
Weight
5.7lbs / 2.6kg
Water tank capacity
20.3oz / 600ml
Lavazza A Modo Mio Jolie Evo: price and availability- List price £99 / AU$129 (about $190)
- Widely available in UK and Australia
- Listed on Lavazza's US site, but not available at the time of writing
The Lavazza A Modo Mio Jolie Evo has a list price of £99 / AU$129, which is about $133, but at the time of writing the machine doesn't seem to be available to actually buy in the US — even though it's listed on Lavazza's site. So it may become available in the future.
If you're in the UK, you can get the coffee maker for £49 if you subscribe to Lavazza's coffee capsule delivery service.
Even at full price, that's very reasonable for a good quality capsule coffee maker. For comparison, the Lavazza A Modo Mio Smeg is £249 / AU$349 (about $350), while the Lavazza A Modo Mio Deséa is £249 / AU$349 (about $330) — though the latter machine also has a built-in milk-frothing system.
(Image credit: Future)If you want to make lattes and cappuccinos with the Jolie Evo, you'll need a standalone milk frother like the Lavazza MilkEasy Plus, which is priced at £79 / AU$119 (about $150) — or the Smeg Mini Milk Frother or Philips Baristina Milk Frother I mentioned earlier.
- Value score: 4.5/5
- Ideal for small kitchens
- Only two controls (lever and button)
- Lever is ergonomic and easy to use
The Lavazza A Modo Mio Jolie Evo is easily one of the smallest coffee makers I’ve tested, measuring a mere 4.9 x 8.3 x 13 inches / 124 x 210 x 330mm (W x H x D). That’s even tinier than the diminutive Nespresso Vertuo Pop, which is 5.3 x 9.8 x 16.7 inches / 136 x 250 x 426mm. It’s ideal for slotting into small spaces, and you can even route the power cable right or left underneath the machine if you need it to sit against a wall.
The Lavazza machine looks classier than its Nespresso counterpart as well, with a smart matte finish and a large chrome handle on top. Pulling this handle up opens a slot on top of the machine. so you can slide a coffee pod inside.
(Image credit: Future)Pushing it back down closes the drawer and pierces the pod, and pulling it up a second time will drop the used capsule into a small bin behind the Jolie Evo’s drip tray. The bin has a grille at the bottom that allows the pods to drain so they’re dry when you remove them for recycling, which is a thoughtful touch.
The large lever is easy to grip and requires barely any force to move, which makes this a particularly good option for people with limited strength or mobility in their hands.
The drip tray has a plastic grate that can be removed for easy cleaning, and is just the right height for an espresso cup. A standard size cup or mug will fit neatly under the dispenser with the drip tray detached.
(Image credit: Future)Other than the handle, there is just one control on the Jolie Evi: a button on top of the machine, with a white light to indicate the machine’s current status: blinking means heating, and white means ready to brew.
It’s available in black and white in Australia, with a gray option also offered in the UK. It was sold in green and red too when it launched in the US.
- Design score: 5/5
- Extremely simple to use, but there's no timer for the coffee dispenser
- It makes delicious espresso in moments
- Its pods can be hard to recycle compared to Nespresso aluminum ones
What makes the Jolie Evo different from all other capsule coffee makers I’ve tested to date is its lack of a timer for brewing different sized drinks. Once you’ve pressed its single button, the little machine will continue dispensing coffee (or hot water, if there’s no pod inside) until you either tap the button again or the tank is empty.
This isn’t ideal if you want to leave your coffee maker to handle itself while you’re preparing the rest of your breakfast, but making a double espresso takes less than a minute so it’s no real hardship.
I would recommend using a measuring cup as your recepticle, though, at least until you’ve got your eye in, because it’s easy to under or over-estimate the output. You could use a coffee scale to a perfect measure, but it would be overkill; this entry-level machine makes a very good espresso, but it’s not intended for real aficionados.
(Image credit: Future)It uses Lavazza’s own A Modo Mio capsules, which are excellent value compared to most coffee pods. Different types command different prices, but you can pick up 80 Qualità Rossa pods direct from Lavazza for £24.50 (about $30 / AU$50). That’s just over 30p per drink, compared to 49p for the cheapest Nespresso Vertuo pods.
Unlike Nespresso, the Modo Mio selection doesn’t include elaborate flavored drinks; there are various beans, blends and roasts to choose from in Lavazza’s store, but each pod contains only ground coffee with no extras. That's no bad thing, in my opinion — coffee capsules with added flavors can taste unpleasantly artificial.
(Image credit: Future)I tested the Jolie Evo with three of Lavazza’s coffees: Qualità Rossa, Passionale, and Qualità Oro. Each comes in a box of 16, and is marked with an intensity score out of 12 to give you an idea of what to expect. The first is a blend of robusta and arabica beans, while the other two are 100% arabica.
My personal favorite was the Qualità Oro, which produced one of the best-tasting espressos I’ve had from a capsule machine and is described as “Sweet and delicate taste, with an intense and fragrant aroma”.
It goes particularly well with milk in a white americano or (if you also pick up a milk frother) a creamy latte.
The problem with these pods, delicious though they are, is that they can be difficult to recycle. Nespresso pods are made entirely from aluminum, which is relatively easy to process once the coffee grounds are removed, and there’s almost certain to be a location nearby where you can drop off the used capsules. If not, you can mail them off for recycling with a pre-paid bag.
(Image credit: Future)By contrast, Lavazza A Modo Mio pods are mostly plastic and can be recycled in the UK through the company’s partnership with Terracycle, but collection points are few and far between.
If there’s nowhere to drop off your pods, your only option is to start your own collection point, accepting pods from other people and sending them for processing yourself. Individuals can’t send their pods in the mail.
South England is relatively well-served, but at the time of writing there are only four drop-off points in Scotland, and none at all in Wales or Northern Ireland.
- Performance score: 4/5
Attribute
Notes
Score
Value
The most affordable A Modo Mio coffee maker, though not currently available in the US.
4.5/5
Design
Stylish and super compact, with ergonomically designed controls.
5/5
Performance
Brews delicious espresso, but lacks a timer to stop flow, and pods can be hard to recycle.
4/5
Buy it ifYou have a small kitchen
This is the most compact coffee maker I've tested to date, and the fact its cable can be routed to the right or left means you can push it right up against a wall if you need to.
You have limited dexterity in your hands
The large lever handle requires virtually no effort to push and pull, and there's just one button to worry about. All removable pieces are easy to detach for cleaning as well.
You want elaborate flavored coffees
A Modo Mio capsules are full of nothing but unadulterated ground coffee, so if you want something fancier you'll need to add your own syrup.
You want easily recyclable coffee pods
A Modo Mio capsules are plastic and can only be recycled if you drop them at a designated collection point. You can't send them in the mail for processing.
If you're not sure whether the Jolie Evo is the right espresso machine for you, here are two others to think about.
Lavazza A Modo Mio Deséa
This is Lavazza's most advanced A Modo Mio machine, with its own built-in milk frother, though it's only compatible with Lavazza's own glass mug.
Read our full Lavazza A Modo Mio Deséa review
Bosch Tassimo Finesse Friendly
Just as easy to use as the Jolie Evo, and affordably priced too. There's a huge range of Tassimo pods to choose from, though they can get pricey — particularly ones for milky drinks that require two capsules each.
Read our full Bosch Tassimo Finesse Friendly review
How I tested the Lavazza A Modo Mio Jolie EvoI used the Lavazza A Modo Mio Jolie Evo in place of my usual espresso machine for two weeks. I used three different types of A Modo Mio capsules (Qualità Rossa, Passionale, and Qualità Oro), drinking it neat as an espresso, and using the Jolie Evo to add extra hot water to make an americano.
I set up and cleaned the machine following the directions in the instruction manual, and allowed the components to air-dry before reassembly.
- For more details, see how TechRadar tests, reviews, and rates products.
- First reviewed: August 2026
- ZeroBEC observes Greatness PhaaS evolving to bypass MFA and phish Microsoft 365, iCloud, Yahoo, and Google Workspace accounts
- Attackers spoofed RingCentral emails post‑ShinyHunters breach, luring victims to fake Microsoft 365 logins that capture authentication tokens
- Greatness is sold on Telegram for $289/month, enabling access to Outlook, Teams, SharePoint, OneDrive, and more across multiple regions
Microsoft 365 users have been getting phishing emails spoofing RingCentral, designed to steal their accounts even if they were protected by multi-factor authentication (MFA), experts have warned.
Security researchers ZeroBEC claim to have observed a phishing-as-a-service (PhaaS) platform called Greatness evolve to also target MFA accounts, as well.
Greatness used to be a simple credential phishing platform. However, in recent times, it evolved to target not just Microsoft 365 accounts, but also those of iCloud, Yahoo, and Google Workspace.
Grabbing MFA-approved authentication tokensZeroBEC notes that RingCentral recently suffered a data breach at the hands of the infamous ShinyHunters hackers, meaning there is a good chance (although not confirmed) that the threat actors exfiltrated a list of emails belonging to RingCentral customers from that attack, and used it in this attack.
Now, RingCentral customers have been getting emails that look as if they are coming from the company itself, despite being mailed from an unknown mail server, and despite failing SPF and DMARC checks. The emails are the standard fake voicemail and performance-review notifications which, if clicked, redirect the victim to attacker-owned infrastructure spoofing the Microsoft 365 login page.
Through this malicious landing page, Greatness operators are able to capture MFA-approved authentication tokens, bypassing the login process entirely and moving straight into victim accounts.
From there, they would enumerate Outlook mailboxes, Teams conversations, and SharePoint sites. They would also access OneDrive files, contacts, calendars, and registered applications through Microsoft Graph.
The number of victims is unknown at the time, but ZeroBEC says Greatness has been active for at least four years now, targeting users in the US, UK, Australia, Canada, and South Africa.
According to BleepingComputer, the platform is being advertised for sale on Telegram channels with “thousands of subscribers”, and is currently being offered for a monthly fee of $289.
Via BleepingComputer
- On August 4, users in Russia reported widespread disruptions to 20+ VPNs
- The attack targets the IPs and entire subnets of major hosting providers
- On August 5, connectivity appeared to have been restored for some
Citizens attempting to bypass Russia's stringent digital censorship are facing fresh obstacles, as a new wave of aggressive government blockages has knocked numerous popular VPN services offline.
On August 4, users across the country reported severe connectivity issues, as reported by Meduza, a Russian independent news outlet. The restrictions appear to have affected more than 20 popular services, making it increasingly difficult for citizens to connect to an open internet.
If you rely on circumvention tools to access independent news, finding the best VPN apps that work in Russia is more crucial than ever. However, this latest blackout suggests the Russian state communications regulator, Roskomnadzor, is deploying more sophisticated methods to sever the country's digital lifelines.
These ongoing blockages are becoming a daily reality for citizens, proving that Russians need a VPN to access Google and Apple as unexplained nationwide outages hit core web infrastructure.
On his Telegram channel, Leonid Volkov, an executive at the Anti-Corruption Foundation (FBK), described the current offensive against independent services as "the largest in history," though the exact scale of the disruptions remains difficult to quantify.
A new era of internet censorship in RussiaThe recent outages appear to involve broad, aggressive network restrictions.
According to Russian tech outlet SecurityLab, these new restrictions have targeted the IP addresses and entire subnets of major hosting providers, taking a blunt-force approach to network censorship.
Specifically, Volkov speculated on Telegram that Roskomnadzor’s enhanced blocking capabilities might stem from data harvesting. He suggested that domestic Russian applications, such as banking and marketplace apps, could be tracking users and feeding a massive database of IP addresses to state censors for targeted strikes.
This wave coincides with reports that Russia's Ministry of Digital Development (Mintsifry) is planning to tighten control over corporate VPNs.
Sources suggest the government wants to establish a continuous monitoring mechanism with hosting providers to quickly exclude disguised circumvention tools from a whitelist of permitted corporate IP addresses. This aligns with ongoing fears that Russia's solution to its VPN crackdown is a state-owned VPN.
(Image credit: Getty Images)Paper VPN is among the services publicly acknowledging connectivity issues, with some users reporting completely broken connections while others simply experience slower speeds.
In a post published on Wednesday (August 5), the company confirms that "the connection has been restored for most users."
Amnezia VPN — which recently strengthened its apps to fight Russia's new approach to VPN blocking — was caught up in the government's wider summer crackdown. The provider previously restored its premium service in Russia following a devastating state-sponsored cyberattack that lasted a month and a half.
Regarding the August 4 wave, however, the provider stated on its channels that its service is "working reliably" in Russia.
"Friends, over the past two days we have seen many reports of massive blockages of other VPNs. Not long ago this happened to us, but the new wave did not affect us," wrote Amnezia VPN, adding that the team is also currently testing a new "Greenland" location for bypassing whitelists.
TechRadar has reached out to the provider for further comment on the current situation. We will update this story as we learn more.
- Audeze is "deeply involved" with the creation of future PlayStation audio tech
- Peter James, marketing director at Audeze, says there are products in "various stages of development"
- He adds that it's Audeze's goal to bring gamers "better sound"
Premier audio brand Audeze is growing its partnership with Sony and is currently "deeply involved" with the development of future PlayStation audio hardware.
Sony acquired Audeze, a leading audio technology brand and pioneer of high-end gaming headphones, in 2023 to continue its focus on innovation in PlayStation audio. Audeze still operates independently and continues to develop multi-platform products, while also benefiting from access to the PlayStation ecosystem.
Speaking to TechRadar Gaming Managing Editor, Rob Dwiar, during a visit to its California HQ, marketing director Peter James called the partnership "unique" considering Audeze still runs as its own brand.
However, Audeze's "primary function within PlayStation is to develop the next generation of PlayStation audio" and to always offer the "best performance" and "the best sound," he went on to say.
"We're more deeply involved with the creation of PlayStation audio products [than ever before], where you could kind of think of us as the audio Skunk Works for PlayStation."
In terms of specific ways of approaching PlayStation audio products' development, James said, "acoustically speaking, we are developing the sound of all future PlayStation audio products," while he also revealed that there are products that have yet to be released and are in "various stages of development."
The marketing director described how Audeze aims to bring the audio quality of planar magnetic drivers to a wider range of people, and being part of Sony and PlayStation can boost that enormously. "[We're] bringing people to better sound, who in many cases wouldn't know what planar magnetic technology is, [and] wouldn't know why it sounds better."
"They're going to buy [an official headset] because it's a PlayStation headset, but now they're getting an elevated audio experience, and potentially, once they become aware of it, chasing down even more elevated experiences," James said.
"The more people get those [drivers], the more that raises their expectation of the next headphone they buy", he went on to say.
He added that the path ahead is going to make Audeze hardware and the tech more "more ubiquitous" and more accessible: "Previously, the most accessible planar gaming headset was $299, and then we went with Sony and made it $149 [with the Pulse Elite headset], so it's slowly pushing its way down to be much more accessible."
And with Audeze adding another planar magnetic gaming headset to the market in its Audeze Maxwell 2 ANC, as well as being part of every PlayStation audio product henceforth, the future is looking bright and full of premium audio for gamers.
It's back-to-school sales season, and Amazon UK is getting in on the action with discounts across laptops, headphones, appliances, and all the useful tech you might need for the year ahead in school, college, or uni. I've had a look through the retailer's latest round of offers and picked out 15 of the best deals that I think are worth buying if you're shopping today.
• See all of today's best deals at Amazon UK
A new year often means a new laptop, and there's one offer in particular that stands out to me as a fantastic all-around choice: it's this Samsung Galaxy Book4 Edge for £549 (was £829). With a Snapdragon X processor, 16GB of RAM and a 1TB SSD, you'll get fast and efficient performance, as well as incredible battery life, and more than enough storage for all of your needs.
If you're just after a few handy extras, then you can get this pair of Anker USB-C Charger Cables for £5.99 (was £7.99), a Warmco Portable Handheld Fan for £19.88 (was £24.99), and an Inui 45W 10,000mAH Portable Charger for £18.66 (was £21.99).
Have a browse through all of my top picks from Amazon's latest sale below, and be sure to stick with TechRadar over the next few weeks for even more back-to-school deals and buying advice from our expert team.
15 best back-to-school deals at Amazon UKDisplay - 15.6 inches
Processor - Snapdragon X
RAM - 16GB
Storage - 1TB
It was slightly cheaper over Prime Day, but this is a great price for a relatively modern laptop that comes with a well-performing and efficient Snapdragon X CPU. On top of that, you get 16GB of RAM, a surprisingly large 1TB SSD for storage, and incredible battery life, which represents good value for money overall. I like it as a light, portable and capable all-around everyday machine for school, college, and at home.View Deal
This is the biggest saving I've seen on a Microsoft 365 Family subscription since Black Friday, so it's excellent value for money overall if multiple people prefer to use Word, Excel, PowerPoint, Outlook, and other Microsoft Office programs. This version grants up to six users access across multiple devices for 12 months, so it's better value than the single subscription. If you just want that, though, it's also currently on offer at Amazon for £69.99 (was £84.99).View Deal
It's always handy to have a portable charger to take with you when you're on the go, and this 10000mAh one from popular brand Inui is a smart buy in my eyes. For starters, it's small and super-slim, making it easy to carry around in a pocket or slip into your bag, while still offering enough juice to charge some of the latest phones and tablets to above 50%. It has two USB-C ports and one USB-A, meaning you can charge multiple devices at once. There's even a handy detachable USB-C cable included for free.View Deal
The Nespresso Vertuo Next is a small and convenient pod coffee machine — and it's now back to its lowest price this year. With multiple drinks in multiple sizes available at the touch of a button, it could be the ideal no-fuss pick for your room or shared kitchen.View Deal
If you want a cheap and basic Bluetooth speaker, then you can't go wrong with this Anker Soundcore for almost £25. Sure, it's not going to give you the biggest and boldest audio quality, but it's still a solid option at an affordable price. Battery life is an impressive 24 hours, too, so it's great to use around the home or take it with you when on the go.View Deal
The Google Pixel 10a is our pick for the best bargain phone for its balance between affordability and functionality. It features a Tensor G4 processor and a display with a 120Hz refresh rate, which offers solid performance at this price point. You also get a Gorilla Glass 7i screen for better drop protection and scratch resistance. For £100 off, you can get this smooth and affordable smartphone for its lowest price yet, or save another £50 on top if you're OK with 128GB of storage.View Deal
Can you ever have too many USB cables? The box full of them under my desk would suggest you can. However, if you want a couple more to leave around the house or take with you when out and about, this pair of USB-C to USB-A cables from top brand Anker is now super-cheap.View Deal
Display - 15.6 inches
Processor - Intel Core 3
RAM - 8GB
Storage - 256GB
This version of the Asus Chromebook Plus offers a solid amount of power for the price. It includes 8GB of RAM and an Intel Core 3 processor, boosting performance, boot times, and load times. There's also a standard 256GB of storage for the price, and battery life is solid at up to 12 hours on a single charge. That's more than enough to last a full working day. At a little under £250, it's a smart buy for school or college, or just as a home laptop for light use and everyday tasks.View Deal
The Ninja Blast Portable Blender is a powerful all-in-one portable blender — and it's now got a small price cut at Amazon. It has a 530ml container with a leakproof lid and sip spout, while the carry handle means you can take it with you to the gym or on your walks. It’s capable of blending ice and frozen fruit, as well as any liquids and powders, while one charge gives you up to two hours of battery life.View Deal
This Warmco Portable handheld fan is a slightly more high-tech and serious-looking device. We haven't reviewed it, but it does have good Amazon reviews and the price is right thanks to this latest deal. According to the user reviews, this looks to be a surprisingly powerful little fan that also offers excellent battery life if you need to take it with you when on the go.View Deal
If you picked up those USB cables for cheap, you might also want a charger plug to use with them. There are several on sale right now at Amazon, but I like this Ugreen option the best. It's a 100W fast charger with three USB-C ports and one USB-A port, offering support for multiple phones, laptops, tablets, earbuds, and more, with the option to charge more than one at a single time with the correct arrangement. The small and compact design, is the ideal finishing touch.View Deal
The iO range is usually a more expensive option in Oral-B's range of electric toothbrushes, but this is an excellent deal on an entry-level option that's also bundled with a few handy extras. It doesn't boast many high-end features, but it has the basic ones you really need to brush your teeth effectively and efficiently, including a two-minute timer and a visible pressure sensor. Plus, you get a refill holder and a travel case included for free.View Deal
The Philips OneBlade 360 is a quick, simple and affordable electric razor that's suitable for a wide variety of face and body grooming needs. This jam-packed bundle was a little cheaper over Black Friday, but today's offer is still a solid one. This bundle includes the handle, as well as two spare blades, four different-sized combs, a body kit, and a USB-A charging cable.View Deal
The Wolfbox MF50 is a convenient and extremely easy-to-use air duster, with a single button to cycle through fan speeds, a trigger to fire out the air, and several different attachments to choose between depending on the cleaning job you need it for. It's completely wireless with USB-C charging and lasts up to four hours on a single charge on the lowest setting, too.View Deal
Perhaps you're moving into halls or other university accommodation and want an option to get some exercise in at home, well, this cheap pack of resistance bands could be useful. For under £10, you get five different-strength bands, ranging from light to XX-heavy, that can help with all sorts of different home training, yoga, pilates, and other exercises.View Deal
When Anthropic's CFO revealed that over 90 per cent of the company's code is now written by its own AI, it landed as a milestone.
Tasks that once consumed hours now take 30 minutes. The productivity gains are significant.
But Anthropic is an AI-native company with some of the world's best engineering talent.
For most enterprises, the question isn't whether AI can generate code at that speed. It's whether they can govern what it generates.
Shadow AI: The new shadow IT"Vibe coding", the term for generating code via AI, has moved into the mainstream. By some estimates, almost half of all new global code is now AI-generated. Developer productivity is up. So is debt that nobody fully understands.
Part of what's driving this is necessity. AI is helping close the engineering talent gap. Teams that lack experienced developers are using it to build at the pace the business demands. The problem is that the same shortage that makes AI indispensable also means there aren't enough senior engineers to review AI-generated code.
Research across Fortune 50 enterprises found that AI-assisted developers introduce security vulnerabilities at ten times the rate of their peers. Forty-five per cent of AI-generated code contains OWASP Top 10 vulnerabilities. Independent analyses indicate that technical debt increases by 30–41 per cent following the adoption of AI tools.
Traditional technical debt is at least visible. Engineers who cut corners know they did it. Vibe coding debt is different: developers often don't realize they have incurred it, because the code looks correct – right up until it doesn't.
We used to worry about Shadow IT. The new threat is Shadow AI: code generated at pace without architectural review, security auditing or institutional understanding of what's been built.
Unlike Shadow IT, which was typically contained within a department, Shadow AI compounds across organizational boundaries. In enterprises where systems are still deeply siloed, complex problems span multiple departments and platforms; no single team has a complete picture of what's been generated or what it depends on.
Enterprises have spent decades paying for yesterday's shortcuts. AI risks creating the next generation of legacy systems, only much faster.
Repeating the COBOL mistakeThe market is flooded with tools promising to read millions of lines of COBOL or Java and convert the functionality into a modern language. This is technically impressive, but strategically flawed.
Legacy systems are full of inefficiencies, redundancies and “swivel chair” workarounds baked in years ago to compensate for other systems' limitations. Translating code line-by-line replicates that bad logic in a newer language, now running on cloud infrastructure with a modern interface on top of old code and dated processes.
To get modernization right, organizations should avoid treating it as a technical exercise. Instead, they should step back and ask whether a process is still valid, not just how to replicate it.
True modernization is about reinventing how work gets done, designed around the employee and customer experience rather than the constraints of systems built decades ago. It is a state of consistent change, and it must be driven by the North Star of clear business objectives.
The agility layerProbabilistic AI needs to operate within deterministic boundaries to be safe and useful at enterprise scale. AI that can generate anything is not the same as AI that generates the right thing reliably with full traceability, in a context where every decision might be scrutinized by a regulator.
This is why high-stakes organizations are looking for an agility layer: a governed process platform that imposes structure, ensures auditability and keeps AI outputs within maintainable boundaries. The US Army followed this approach for security with agility to operate with certainty at speed.
Ordering ammunition in disconnected field conditions is mission-critical, with zero tolerance for ungoverned outputs. Similarly, pharmaceutical giant Merck's clinical supply chain, where regulatory scrutiny is intense and errors have patient safety consequences, required a platform that could accelerate delivery without sacrificing auditability.
These are examples of organizations trying to address the hardest part of modernization: discovery. They are using AI to extract specifications from even the most poorly documented legacy applications, converting them into visual plans covering UI, data models and process flows.
They're generating software components rather than custom code to reuse in other applications, accelerating development time and reducing technical debt. AI agents then build against those specs under human supervision, with developers assigning tasks and iterating throughout – at roughly 25 per cent of the time traditional approaches require.
The future is bespoke, not general-purposeThere is a temptation to conclude that general-purpose AI will soon render enterprise platforms obsolete. This is premature and in regulated environments, dangerous. General-purpose models are extraordinary at generating plausible outputs. They are not equipped to ensure those outputs are auditable, compliant or maintainable by teams that did not generate them.
The vibe coding wave is real, and so is the debt it is creating. The organizations that navigate the next decade successfully and avoid falling into the same old legacy traps will not be the ones that generated code the fastest in 2025 and 2026.
They'll be the ones to build governance and process boundaries into how AI was used from the start – so what was built quickly can still be understood, maintained and trusted years later.
We've reviewed, rated, and ranked the best laptops for programming.
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
- Forescout’s Vedere Labs found 15 flaws in TP‑Link Omada business networking gear, exploitable for RCE when chained with prior CVEs
- Weak trust shortcuts in zero‑touch provisioning exposed devices to client‑side code execution, hijacking, spoofing, and encrypted comms compromise
- TP‑Link released firmware updates; admins should patch immediately, with 1,800+ Omada controllers exposed online
TP-Link has patched more than a dozen vulnerabilities across multiple business networking products which could have been chained to achieve remote code execution (RCE).
Security researchers at Vedere Labs from Forescout found the flaws and published an in-depth report on the issues, which particularly affect TP-Link Omada, the company’s business networking platform for centrally managing enterprise and small-business network infrastructure.
It includes cloud-managed Wi-Fi access points, routers, switches, gateways, and controllers, all of which can be monitored and configured from a single interface.
Enabling "concrete attacks"These support zero-touch provisioning (ZTP), a mechanism that allows IT managers to deploy and maintain devices without needing to configure each one manually and on site.
However, ZTP has to establish trust between a factory-fresh device, and a controller with no human involved, so TP-Link used different shortcuts: from hard-coded keys and certificates shared across multiple devices, to default credentials, and from guessable serial numbers as “identity”, to weak session-key randomness.
Now, Forescout says 15 vulnerabilities its researchers discovered all allow for different ways of exploiting these shortcuts, meaning a flaw anywhere in the onboarding chain can compromise every device that goes through it. These bugs would need to be combined with two previously disclosed command-injection flaws, though.
“The vulnerabilities fall into four impact categories: client-side code execution, information disclosure, device hijacking and spoofing, and compromise of encrypted communications,” Forescout said. “Combined with two previously disclosed CVEs (CVE-2025-7850 and CVE-2025-7851), these flaws enable concrete attacks that let attackers infiltrate networks through controllers and client devices.”
Out of the 15 discovered flaws, 11 received CVE identifiers, and the rest did not receive a tracking number.
Forescout said there are more than 1,800 Omada controllers accessible from the wider internet. If you are using any of the devices from the platform, you should head over to TP-Link’s download portal and grab the latest firmware for your device model.
- Gigabyte graphics cards are rumored to be set for a big price rise
- It could be a hike of 20% to 40%, and other rumors are also pointing to major price hikes for GPUs more broadly
- This means now is likely a good time to make a purchase, particularly with certain cards which have avoided the worst of the hikes so far
Desktop GPU prices are set to rise if the rumor mill is right – with yet another prediction to that end being aired – and while graphics cards are already costly in some cases, now might be the time to buy certain models, based on a fresh roundup of average pricing.
First off, let's deal with the latest gloom nugget from the grapevine regarding GPU pricing, courtesy of @momomo_us on X (via Wccftech) who picked up a notice from CFD (a Japanese components supplier) warning of substantial price hikes on Gigabyte graphics cards.
The claimed rise (add seasoning, of course) is in the order of 20% to 40%, so at the latter end of that scale, we're looking at prices shooting up by almost a half in some cases. Existing orders from retailers buying boards from CFD may even be cancelled, the notice states.
I perhaps wouldn't put as much stock in this, except that it echoes a rash of recent reports contending that large price hikes are about to happen in the GPU world, which makes it all the more believable.
Only yesterday I reported on the situation in South Korea where RTX 5000 models are rumored to be getting price increases of up to 30%. So again, that's a hefty hike, and it follows other speculation about Nvidia charging its partners more for video RAM (which will push up costs, of course).
This apparent Gigabyte move will affect both Nvidia and AMD graphics cards in Japan, and there have been rumors about Team Red having Radeon price hikes in the pipeline due to more expensive VRAM, too. That only makes sense, as if VRAM is getting more costly for Nvidia, it will be for AMD too. And while some of these reports are based on regional activity, the cost issues will surely apply globally – none of these markets exist in a bubble.
Analysis: I think it's time to buy – especially with certain GPUs(Image credit: Gigabyte)The RAM crisis has a lot to answer for, and it very much looks like graphics card prices are going to be jacked up considerably during the remainder of 2026, with video memory being the primary driver here. Given that, it seems that if you are thinking about a GPU upgrade, now is likely the time to go for it, rather than waiting and being hit by price hikes.
That's especially true if you're considering a 16GB graphics card (as more gamers are these days), as the products with more VRAM are going to get hit the hardest for price rises as memory modules become even more expensive.
If we want a snapshot of the current GPU market as price tags stand, fortunately enough, TechSpot has just furnished us with an extensive (multi-region) pricing roundup as of the end of July which highlights some key points.
Firstly, low-end GPUs are still relatively unscathed, with offerings like the RTX 5050 and RTX 5060 8GB, and the 8GB spin on the 5060 Ti, not having been bumped up by all that much in the RAM crisis thus far. However, as for the 16GB incarnation of the RTX 5060 Ti, that has seen major price bumps, reinforcing that observation on 16GB pain (this card is up over 40% since November 2025 in the US, for example – and shows a 14% rise since April 2026).
In the tier above that, the RTX 5070 is holding at a reasonable price, and even if the sub-MSRP levels of late last year are well and truly gone, it remains a decent buy. (As long as you can stomach 12GB of VRAM at this price bracket, of course – and that's one of the reasons why this GPU hasn't been hit so hard, no doubt).
Above this level, Nvidia GPUs have witnessed some of the nastier price increases, as you've probably noticed.
On the AMD side, matters look more promising, particularly in the case of the Radeon RX 9070 and RX 9070 XT, with TechSpot finding that the 16GB models in the RDNA 4 family haven't suffered as much as Nvidia's GeForce offerings. The price of both these 9070 models has remained steady over the last few months, and in many regions, they're only up 10% or so compared to late 2025. (The 9070 XT is an exception in the US, mind, where it has risen by closing on 20% since last year).
If you're inclined towards one of these Radeon mid-rangers, or Nvidia's RTX 5070, I'd strongly suggest you buy now rather than wait based on what we're hearing from the rumor mill. Lower-priced graphics cards are likely worth grabbing, too, because they may be heading upwards in the pricing stakes before long – although notably the RX 9060 XT with 16GB has already become a good deal more costly, much like its RTX 5060 Ti 16GB rival.
As noted, Nvidia's higher-end graphics cards (RTX 5070 Ti and upwards) have already been hit with alarming price hikes, but they're likely to get even worse. So, this could be a case of keeping your wallet shut and holding out for possibly a lengthy time, as I can't really recommend them even at the current pricing.
The trouble is, we don't know how long the RAM crisis will continue, and some of the guesswork on that front is very bleak indeed – suggesting it might be the rest of this decade effectively. At this point, I'm inclined to believe we're in it for the long haul with PC component pricing woes, which is why I think it's best to buy a GPU now, at least with the mentioned models, assuming you're in a position to do so.