Feed aggregator

UK organizations are investing heavily in AI tools. According to latest Office for National Statistics (ONS) figures, 29% of UK businesses were using at least one type of AI technology as of June 2026.

Yet despite growing adoption, many organizations are still focused on the wrong question.

Most AI discussions start with the technology. Which models should we use? Which platform should we buy? How quickly can we deploy it?

While these decisions matter, they are not what determines success. The biggest obstacle to AI transformation is not technology but the people.

Organizations have managed to navigate major technology shifts before, from cloud migration and agile development to Zero Trust security. AI, however, represents something different. Rather than simply introducing new tools, it is changing how work gets done.

The businesses that succeed will be those that give AI tools the context they need to be useful and the people who know how to put that context to work.

The first shift: changing the mindset

One of the most striking trends in AI adoption is the gap between individual use and organizational readiness. According to recent research, nearly one in five UK workers now uses generative AI daily, yet only one in ten UK organizations has successfully scaled AI or embedded it into core operations.

Employees are embracing AI, whilst organizations are still figuring out how to turn that enthusiasm into lasting business value.

The people seeing the biggest gains aren't treating AI like business software. They're treating it like another worker. They know how to provide context. They know how to set expectations. They know how to review outputs, challenge assumptions and refine results. In short, they're managing AI rather than simply using it.

That distinction matters. AI isn't behaving like traditional software. It behaves more like a worker. The people getting the most value from it are those who know how to brief it, direct it and assess its output.

For many employees, that way of working is still emerging. AI is often used as a faster route to information rather than a collaborative tool capable of helping shape and execute work. Increasingly, every employee will need to become an AI manager. The ability to brief, direct and challenge AI will become a core workplace skill, regardless of role or department.

The second shift: treating upskilling as a business priority

Organizations cannot hire their way to AI readiness. There simply aren't enough people with the necessary expertise, and external hiring alone cannot replace the institutional knowledge already sitting inside the business.

The businesses moving fastest are investing in the people they already have. They're helping employees develop the confidence, judgement and practical skills needed to work alongside AI every day.

Giving employees access to AI tools is only the starting point. Organizations also need structured training, clear governance and opportunities for practical experimentation. The real opportunity lies not in teaching people to use AI, but in helping them understand how to work alongside it, applying critical thinking and human expertise to deliver stronger outcomes than either could achieve alone.

As AI capabilities continue to evolve, the organizations that treat learning as an ongoing discipline rather than a one-off training program will be best positioned to adapt.

The third shift: redesigning the organization itself

The final shift is organizational.

For decades, businesses have been structured around functions, departments and clearly defined handoffs. These models were designed to create accountability and specialization, but they can also introduce friction.

As AI begins working across teams and systems, some of those traditional boundaries begin to blur. Processes that once relied on handoffs between departments can increasingly be coordinated through connected digital workers. Activities that once required multiple individuals, applications and approval chains can increasingly be coordinated through connected digital workers.

While organizational structures will remain important, leaders will need to reconsider how work is organized and how value is generated.

Businesses will increasingly organize around outcomes rather than functions. As AI takes on more coordination work, we may start to see entirely new operating models emerge, including businesses capable of generating enormous value with far fewer employees than would have previously been possible.

AI runs on context

Many organizations are focused on the latest AI capabilities, but AI can only work with the context it has access to. The more relevant context AI can access, the more valuable its outputs become.

This is why digital transformation remains so important in the age of AI. Not because every organization needs to reinvent how people work overnight, but because information needs to flow more freely across the business.

Too often, transformation is viewed as a large-scale overhaul. Some of the most impactful improvements come from making existing workflows smarter.

For years, organizations have viewed printing as a blocker to digital transformation. Although when connected to modern workflows, cloud platforms and business systems, the print action can become an enabler instead. It can help trigger processes, route information and connect systems in ways that support both employees and AI.

Sometimes transformation starts with a major initiative. Sometimes it starts by making familiar actions more intelligent. The goal is the same: making information easier for both people and AI to access and use.

AI transformation is a people transformation

As organizations race to adopt AI, it's easy to assume success will be determined by the technology itself.

The real differentiator will be how effectively organizations prepare their people and unlock the knowledge already sitting across the business.

AI can only work with the context it has access to, and much of that context still lives in documents, workflows and everyday processes.

The organizations that see the greatest value from AI won't necessarily be those with the most advanced models. They'll be the ones that make information easier to access, enable people to work differently and create an environment where humans and AI can contribute together.

Manage your data with the best business cloud storage.

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

• Slack Code will let workers code as a team
• Dedicated channels will let coding teams track changes in real time
• A wide range of AI agents will be available to use

Cracking those complex coding tasks could soon be a lot easier thanks to a new release from Salesforce and Slack.

The new Slack Code looks to brings AI-powered coding into the team workspace, changing how employees build and work, making coding "multiplayer".

The launch will allow teams to set up dedicated channels where teams can focus on a singular task or project, sharing it with their fellow workers to ensure nothing gets missed.

Slack Code

"We’re in the middle of a coding boom that’s changing how people build and work," a Salesforce blog post announcing the news declared, citing the popularity of coding agents such as Claude Code, Claude Tag, Devin from Cognition, GitHub Copilot, ChatGPT, and Vercel.

"Now the challenge is connecting what you’re building to the conversations your team is having, all in one interface."

Slack Code will look to tackle everyday tasks such as updating a web page or fixing a bug, but is also designed for the next big project or idea. Users simply tag in their coding agent of choice, which then creates a code channel to begin.

(Image credit: Salesforce)

Anyone in the channel can see the conversation, review code diffs, check live previews, give feedback, and approve the work before it ships, without the need to ever leave Slack.

Every code channel maps to one project, meaning it (should) so it stays focused and tidy, with the ability to mute or archive when it's not needed.

All users will be able to follow the work as it happens, including live previews, in their own tabs, with the ability to see exactly what has changed, with old code being struck through. When it's ready to preview, the team can watch the HTML output run live, and when a task is completed, the channel archives itself, with the record remaining as an audit log.

Organizations can tailor agent behavior and connect with existing tools, including auto-provisioning of accounts, authenticating users, and automating parts of the workflow to fit how your company works.

"Coding agents changed how individuals build. Now Slack Code changes how teams build," Salesforce's blog post adds.

"Squeezing complex, multi-turn agent work into standard threads creates noise, and letting developers retreat into private browser tabs means losing the thing that makes teams effective: shared context. Slack Code gives your team a dedicated place to work with an agent on a specific coding session or project, together, right in Slack."

Why is the car race in this part of D.C.? Will it be streaming? Who's paying for it? Here's what to know before the Freedom 250 Grand Prix hits the National Mall this weekend.

(Image credit: Andrew Harnik)

The USS Abraham Lincoln is now on its way home after spending nine months at sea. Experts weigh in on the mental-health challenges of prolonged confinement and isolation aboard an aircraft carrier.

(Image credit: U.S. Navy via Getty Images)

U.S. officials say the ever-growing population of wild horses that roam federal lands in Colorado is depleting natural resources. But some activists say efforts to remove the horses are inhumane.

(Image credit: McKenzie Lange)

In dozens of states around the country, people are vandalizing, blocking and taking down Flock cameras. The surge in this type of activity underscores the growing resistance to this police surveillance technology.

(Image credit: Meg Anderson)

Experts warn Trump's transactional approach to alliance will erode political trust

(Image credit: JUNG YEON-JE)

Trump threatens economic pressure campaign against Iran, SCOTUS to weigh in on whether Trump's ballroom project, investors dump stocks and bonds despite government's efforts to calm market.

For StoryCorps, a mother remembers her daughter who was killed in a mass shooting by speaking with her daughter's best friend.

Critical infrastructure faces a common challenge. Across transport, utilities and communications, organizations are expected to deliver higher levels of performance, resilience and security while working with ageing assets and increasingly constrained budgets.

Large-scale replacement programs are rarely practical, affordable or necessary.

Instead, the organizations making the greatest progress are taking a different approach. They are building a much deeper understanding of the IT infrastructure they already own, using data and operational insight to improve performance, target investment and extend asset life.

This shift reflects a broader change in how infrastructure is managed. Rather than asking what needs replacing, operators are asking how existing IT assets can deliver more value.

Intelligent Infrastructure is helping answer that question by connecting assets, improving visibility and enabling better operational decisions.

For organizations responsible for critical infrastructure, success increasingly depends on understanding what is already in place, optimizing its performance and ensuring every investment delivers measurable value.

Different sectors, the same infrastructure challenge

While the infrastructure itself may differ between sectors, many of the challenges facing operators today are remarkably similar.

Whether managing a rail network, a water treatment facility, an electricity distribution network or a communications system, there is growing pressure to improve performance while operating within increasingly constrained budgets. At the same time, expectations around reliability, resilience and cybersecurity continue to rise, with customers and stakeholders expecting services to be available whenever they are needed.

Many organizations are also reliant on infrastructure that has been in operation for decades. In some cases, assets are performing well beyond their original design life. Replacing them is often costly, disruptive and difficult to justify when investment priorities are competing for limited resources.

Alongside this, operators are being asked to meet wider sustainability objectives, reduce unnecessary site visits and make better use of the resources available to them. The result is a complex balancing act between maintaining existing infrastructure, delivering operational improvements and planning for future requirements.

These challenges are not unique to one industry, and many organizations are asking the same questions. How can they improve resilience without significantly increasing costs? How can they make better use of existing assets? How can they prioritize investment with greater confidence? And how can they continue to meet rising expectations while working within finite resources?

Finding the answers starts with understanding infrastructure performance in greater detail. Not simply knowing whether an asset is operational, but understanding how it is performing, how its performance changes over time and what those changes might indicate. As infrastructure networks become larger, more connected and increasingly complex, that level of understanding is becoming essential.

Why understanding asset performance has become a strategic advantage

Infrastructure operators have never had access to more information about their assets than they do today. Connected devices, sensors and operational systems continuously generate information about the condition and performance of infrastructure. The opportunity now lies in making better use of that information to gain meaningful understanding to enable operational decision making, supporting desired business outcomes.

For many organizations, operational data exists across multiple systems, technologies and teams. Individual systems may provide valuable insight into specific assets, but they rarely provide a complete view of how the wider infrastructure network is performing. This can make it difficult to identify emerging issues, understand relationships between assets or recognize changes in performance before they begin to affect service delivery.

This becomes even more challenging where infrastructure has evolved over many years and comprises different technologies from multiple manufacturers. Legacy systems often operate alongside newer technologies, with each generating its own data and requiring its own management processes. Bringing these different sources of information together provides a far richer understanding of how infrastructure is performing.

Greater visibility allows operators to move beyond simply reacting to faults. It creates an opportunity to understand what normal performance looks like, identify trends over time and recognize when assets begin operating outside expected parameters. In many cases, these early indicators provide valuable time to investigate, prioritize activity and resolve issues before they affect customers or critical services.

This is where Intelligent Infrastructure begins to deliver real value. By securely connecting assets, bringing together operational data and providing a clearer understanding of infrastructure performance, organizations are better equipped to make informed decisions about how networks are managed, maintained and developed.

From operational insight to better decisions

Traditionally, many maintenance programs have been built around routine inspection schedules or simply responding to faults after they occur. While these approaches continue to play an important role, they do not always reflect the actual condition of assets or where engineering effort will have the greatest impact.

A clearer understanding of infrastructure performance allows operators to take a more informed and prognostic approach. Maintenance activities can be prioritized according to operational need, performance trends and potential risk rather than treating every asset in the same way. Engineering teams are able to focus their attention where it will deliver the greatest benefit, while emerging issues can be investigated before they develop into larger operational problems.

Over time, this supports a more informed approach to asset management. Maintenance ceases to be reactive and becomes increasingly proactive driven by the condition and performance of assets rather than fixed schedules alone. This can help reduce unplanned outages, improve network resilience and minimize disruption for customers and end users.

The benefits extend well beyond day-to-day operations. Better insight into infrastructure performance also supports longer-term planning by helping organizations understand where investment will deliver the greatest value. It provides greater confidence when deciding which assets continue to perform effectively, where upgrades are required and how infrastructure investment can be prioritized over time.

For organizations operating large and complex infrastructure estates, this is becoming increasingly important. Every investment decision carries an opportunity cost, making it essential to understand where resources will have the greatest impact. Better operational insight provides the evidence needed to support those decisions while helping organizations maximize the value of the infrastructure they already own.

It also contributes to wider organizational objectives. Targeted preventative maintenance can reduce unnecessary site visits, improve the efficiency of engineering teams and support sustainability ambitions by reducing travel and extending the useful life of assets where appropriate.

Understanding infrastructure will define its future

The debate about critical infrastructure often focuses on what needs replacing. In reality, the bigger opportunity lies in making better use of what already exists.

Organizations that combine connectivity, operational data and engineering expertise will be able to make better decisions about maintenance, investment and long-term asset management. They will improve resilience, extend asset life and deliver stronger outcomes without relying solely on large-scale capital programs.

Intelligent Infrastructure gives operators the insight to act earlier, invest with greater confidence and make better use of the assets they already own. Over time, that leads to more resilient networks, longer asset life and better outcomes for the customers and communities that rely on them.

We've reviewed and ranked the best internet security suites.

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

For organizations looking to gain an edge by adopting AI tools, the decision has been dominated between large language models.

Every new release promises better reasoning, greater accuracy and more capabilities than the last.

But as foundation models become increasingly powerful and widely available, the real differentiator for organizations won't be which model they choose - it will be how effectively they deploy agentic AI to solve real business problems.

At the recent 2026 Gartner Data and Analytics Summit, the message from analysts was clear; standalone AI models are outdated and autonomous, interconnected AI agents are the future. But to get real value from agentic AI, IT leaders should prioritize “high-frequency, low-complexity use cases, apply guardrails and upskill the workforce.”

To this statement, I would add that the next phase of enterprise AI isn't about having access to the biggest model. It's about building better agents.

That means moving beyond seeing AI as simply another chatbot and instead treating it as an operational capability that can augment employees, automate repetitive work and unlock efficiencies throughout the organization.

The biggest AI opportunity isn't conversational

For many people, AI still means asking ChatGPT a question and receiving an answer. That's certainly one application, and conversational interfaces have played an important role in making AI accessible. But that’s only scratching the surface of what's possible.

Some of the most valuable agentic AI deployments are focused on operational work rather than conversation. For example, in higher education, AI agents are helping students navigate support services by answering routine questions about enrolment, campus facilities or administrative processes, freeing staff to deal with more complex issues.

In transport, agents can combine timetable information, onward travel options and customer support into a single interaction, saving users from having to search across multiple services.

Elsewhere, organizations are using AI agents to qualify sales enquiries, summarize complex documentation, and extract information from contracts. Tasks that previously required hours of manual effort.

The common thread isn't simply that these agents can answer questions. It's that they remove administrative burden and enable people to focus on work that requires judgement, expertise and human interaction.

Better agents start with better problems

One of the biggest misconceptions surrounding agentic AI is that organizations should begin by deciding where they want an AI agent.

In reality, for organizations looking to develop their AI offering, the first question shouldn't be, "Where can we deploy AI?" It should be, "What business problem are we trying to solve?"

Starting with technology often leads organizations to build agents that deliver little value because they're automating the wrong process. The organizations seeing the strongest returns are those identifying operational pain points first, in areas where employees spend large amounts of time on repetitive administration, searching for information or manually moving data between systems.

When those challenges are clearly understood, agentic AI becomes an effective operational tool rather than technology looking for a use case.

The model is becoming less important than the implementation

Today's leading AI models are improving at an extraordinary pace. They're becoming more capable, faster and increasingly cost-effective which means that the competitive advantage is shifting

Organizations are unlikely to outperform competitors simply because they've selected one foundation model over another. Instead, success will increasingly depend on everything that surrounds the model.

Data quality is one of the biggest differentiators. An AI agent is only as good as the information it's grounded on. Contradictory documentation, inconsistent policies, or poor knowledge management inevitably reduce the quality of outputs.

Prompt engineering is another often overlooked capability. Designing an agent isn't simply a matter of giving it instructions and expecting consistent results. Well-designed prompts establish clear guardrails, define the agent's responsibilities and ensure it stays focused on its intended purpose.

The same is true of governance. A good agent shouldn't answer every question it's asked. It should understand its role, know when to decline requests that fall outside its remit and operate only with the information and permissions it genuinely needs. Applying the principle of least privilege isn't just good security practice, it also improves efficiency by reducing unnecessary processing and keeps costs under control.

Ultimately, the quality of the implementation matters far more than the size of the underlying model.

Trust is built through iteration, not ambition

Another common mistake is assuming that agentic AI should be deployed through large-scale transformation programs. But, in practice, the opposite is usually true.

The organizations achieving meaningful outcomes tend to start with a tightly defined use case that delivers measurable value. This creates confidence among employees, demonstrates return on investment, and provides an opportunity to refine governance before expanding into more sophisticated workflows.

Agentic AI should be treated as a product that evolves over time rather than a project that's complete on launch day. However, that also changes the way organizations approach testing.

Traditional software is designed to deliver identical outputs every time. AI agents are probabilistic by nature, meaning consistency becomes just as important as functionality. Continuous testing, monitoring, and refinement therefore become integral to successful deployment.

Just as importantly, employees need to be part of the journey. Building trust isn't about unveiling an AI agent once it has been completed. It's about involving teams throughout development, helping them understand how decisions are made, where guardrails exist and how the technology is intended to augment and not replace their expertise.

Better agents will define the next phase of AI

It’s inevitable that the AI industry will continue producing faster, cheaper and more capable foundation models. But what isn't inevitable is whether organisations will be able to translate those capabilities into meaningful business value.

The next AI race won't be won by those with access to the biggest model. It will be won by those that build trusted, well-governed agents capable of solving real operational problems.

Understand your costs with the best accounting software for small 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

Metal Gear Solid 4: Guns of the Patriots has become something of a legend over the years, being the one big game in the franchise that never left its original platform – the PlayStation 3.

Review info

Platform reviewed: PS5
Available on: PS5, PC, Xbox Series X|S, Nintendo Switch, Nintendo Switch 2
Release date: August 27, 2026

Finally, the port once thought impossible has returned in the new Metal Gear Solid Master Collection Vol.2. The first collection packaged the classic stealth trilogy with the MSX and NES games that preceded it, and this time around the package is headlined by the return of Metal Gear Solid 4 and the duo of the PlayStation Portable’s excellent Peace Walker and underrated Game Boy Color spinoff Ghost Babel.

Off the bat, it may seem like a bit of a raw deal for the second to only feature one bonus title compared to the first collection, which offers three of the greatest games ever made along with four bonus titles and two graphic novels.

However, the work done on Metal Gear Solid 4 is far more than the bare-bones re-releases of the first game and Bluepoint's 2011 HD Collection versions of Metal Gear Solid 2 and 3 found in its predecessor, while Peace Walker is a phenomenal mainline entry and Ghost Babel ranks among the series' greatest spinoffs (and is better than the NES games at least).

It's a fairer trade than the numbers suggest.

War has changed

(Image credit: Konami)

Metal Gear Solid 4 is the crown jewel of this collection, and unlike the first – which had some major issues at launch – I'm happy to report that this is a phenomenal port. Running at 4K and a solid 60 frames per second (fps) throughout with massively improved loading times between areas, it's hard to see any flaws in this version of the stealth classic… well, aside from those that were already there.

Long thought unportable due to being made specifically for the esoteric PS3 system architecture and a hefty number of licensed products appearing in-game, Metal Gear Solid 4 remains mostly intact.

You still have the iPod; you still see the PS3 console in your plane. However, Konami has gone back and remixed some elements to make sense on modern consoles, with the famous "we're on PS3" scene changed to your platform of choice – even if the almost 20-year difference in Christopher Randolph's voice is pretty obvious.

In fact, the only thing I noticed that was missing was the now-redundant installation menu featuring Snake smoking between chapters.

(Image credit: Konami)

Despite its notoriety, Metal Gear Solid 4 is a strange game. I'd even argue it's the worst of the six main games in the series, even if it is still a great time.

It gets off to a really strong start with chapters set in the Middle East and South America, featuring some incredible stealth gameplay with a unique battlefield stealth mechanic which allows you to manipulate the ongoing war between rebel militia and PMC groups.

It's a genius bit of game design, and it’s a shame that it's abandoned by the end of the second chapter.

After an Act 3, which boils down to a not-so-great tailing mission, Metal Gear Solid 4 goes off the rails with its plot, lengthy cutscenes, and gameplay segments.

Some of these are highlights of the game, such as the entirety of Act 4 or the final boss fight that ranks among the best in the series. However, some plot decisions and retcons make for the messiest story in the franchise's history – I won't go into detail, but a particular retcon about Metal Gear Solid 3's Para-Medic is a terrible choice that I'm still bugged by years later.

(Image credit: Konami)

What's there is still a great game I'd recommend to anyone, though it's a shame some of the most interesting gameplay elements are abandoned almost entirely, even if the rollercoaster ride of set-piece and fan-service moments is a joy to play through.

The opening to Act 4 in particular is one of my favorite gaming moments ever and still gives me chills to this day.

We're nomads, and we always will be

(Image credit: Konami)

Peace Walker, on the other hand, is an unsung part of the saga. Originally released on the PSP, it was somewhat overlooked, but this game is Metal Gear Solid 5 in all but name.

Taking place between Snake Eater and The Phantom Pain, Peace Walker is Metal Gear crossed with Monster Hunter, with a mission-based design that allows for short bursts of play. It introduced base management, weapon development, and nicking soldiers via balloon aspects that The Phantom Pain would later expand upon.

Best Bit

(Image credit: Konami)

Returning to Shadow Moses in Act 4 of Metal Gear Solid 4 is one of the greatest bits of fan service ever deployed in a game, and the needle drop of The Best Is Yet To Come still sends chills down my spine almost 20 years later.

While not quite on the level of the original trilogy, Peace Walker has a great story, which helps show how Big Boss eventually became the person we saw as a villain in the original MSX games – arguably better than The Phantom Pain did. It also features one of the best supporting casts in the series, with Kazuhira Miller returning from Metal Gear Solid and being fleshed out into one of the series' best, while the likes of Amanda, Paz, and Professor Galvez are memorable additions to the series canon.

I say Metal Gear crossed with Monster Hunter, but nowhere is that more felt than in the bossfights, which have you take on massive mecha over a prolonged period. While I like this in theory, these bosses are unfortunately the worst the series has to offer – and aside from the titular Peace Walker and true final boss – are not memorable in the slightest.

(Image credit: Konami)

Peace Walker also features co-op and versus multiplayer modes, which are present in this new collection, though I wasn’t able to test them over the review period. From my experiences with the HD Collection version, Peace Walker multiplayer is a blast.

This port is a bit rougher than Metal Gear Solid 4, mostly because there isn't too much you can do to make PSP assets shine in 4K. However, the addition of new higher-quality versions of the comic-book-like cutscenes by Ashley Wood is a welcome bonus

It's not over yet

(Image credit: Konami)

If you played Master Collection Vol 1, you will have a good idea of what to expect this time around. Both main games come with a Screenplay Book and Master Book, which features rundowns of the story, background info on characters, and strategy-guide-esque details to let you find secrets in-game.

There are also soundtrack selections and the ability to download other language versions of the games if you please. The Metal Gear Solid 4 edition also features the Metal Gear Solid 4 Database app, which was released on PS3 around the launch of the original game, although sadly its version of Metal Gear Online is not present.

And while the bonus titles included the original MSX2 version of Metal Gear and Metal Gear 2 – which I wrote about how good they are a few years back – and the less-good NES versions of them, this time around the only additional title is Metal Gear: Ghost Babel for the Game Boy Color.

(Image credit: Konami)

Ghost Babel takes gameplay akin to Metal Gear 2, but fuses it with some elements of the PS1 classic while adapting it to a stage-based formula for handheld. While it doesn't quite match the main series, Ghost Babel should not be slept on, as it's one of the best spinoff titles the franchise has to offer.

Sadly, that's it though. It's a shame, as the likes of Metal Gear Acid or Metal Gear Rising would've been much appreciated additions.

Metal Gear Solid Master Collection Vol.2 is another package of two fantastic games, with some great bonuses included. While it's mostly a vehicle to get Metal Gear Solid 4 out of PS3 jail, Peace Walker and Ghost Babel are two brilliant additions that you might have missed the first time around.

Should you play Metal Gear Solid Master Collection Vol.2?Play it if...

You didn't have a PS3
Metal Gear Solid 4 is out of PS3 jail after almost 20 years. For those who have been waiting, this is one of the biggest dream ports to finally get, and being able to play it in a version that sticks to a solid 4K resolution and 60fps is worth a return trip even if you did play it on PS3.

You never played Metal Gear Solid: Peace Walker
Yes, Metal Gear Solid 4 is the main event of this collection, but Peace Walker is one of the best games the series has to offer. Not to mention an essential part of the storyline heading into Metal Gear Solid 5.

Don't play it if...

You haven't played Metal Gear Solid Master Collection Vol 1.
Metal Gear Solid 4 was made with the idea that you'll have the knowledge of the original trilogy of Metal Gear Solid games at a bare minimum, leading to a plot that will be extremely hard to follow for anyone not caught up. Not to mention that all of the emotional impact will be nullified without prior knowledge

Accessibility features

Both Metal Gear Solid 4 and Peace Walker have subtitles and controller remapping features, but otherwise the accessibility features included in this collection are the same as they were back in the original releases.

This means no color blind support, audio description, or gameplay tweaks are available – which could cause issues during moments in both games when tons of button mashing is required.

How I reviewed Metal Gear Solid Master Collection Vol.2

I played 12 hours of Metal Gear Solid Master Collection Vol.2 on PS5 Pro using a DualSense Wireless Controller, on an LG C5 OLED TV with an LG-US70TY soundbar.

During this time, I played a full run of Metal Gear Solid 4 while skipping most cutscenes (as I had replayed the game very recently for comparison) through to the end of Act 4. I played through to the end of Chapter 1 of Metal Gear Solid Peace Walker, and roughly half of Metal Gear: Ghost Babel. I have played all of these games multiple times before this review, and was able to compare to other versions where applicable.

First reviewed August 2026

Metal Gear Solid 4: Guns of the Patriots has become something of a legend over the years, being the one big game in the franchise that never left its original platform – the PlayStation 3.

Review info

Platform reviewed: PS5
Available on: PS5, PC, Xbox Series X|S, Nintendo Switch, Nintendo Switch 2
Release date: August 27, 2026

Finally, the port once thought impossible has returned in the new Metal Gear Solid Master Collection Vol.2. The first collection packaged the classic stealth trilogy with the MSX and NES games that preceded it, and this time around the package is headlined by the return of Metal Gear Solid 4 and the duo of the PlayStation Portable’s excellent Peace Walker and underrated Game Boy Color spinoff Ghost Babel.

Off the bat, it may seem like a bit of a raw deal for the second to only feature one bonus title compared to the first collection, which offers three of the greatest games ever made along with four bonus titles and two graphic novels.

However, the work done on Metal Gear Solid 4 is far more than the bare-bones re-releases of the first game and Bluepoint's 2011 HD Collection versions of Metal Gear Solid 2 and 3 found in its predecessor, while Peace Walker is a phenomenal mainline entry and Ghost Babel ranks among the series' greatest spinoffs (and is better than the NES games at least).

It's a fairer trade than the numbers suggest.

War has changed

(Image credit: Konami)

Metal Gear Solid 4 is the crown jewel of this collection, and unlike the first – which had some major issues at launch – I'm happy to report that this is a phenomenal port. Running at 4K and a solid 60 frames per second (fps) throughout with massively improved loading times between areas, it's hard to see any flaws in this version of the stealth classic… well, aside from those that were already there.

Long thought unportable due to being made specifically for the esoteric PS3 system architecture and a hefty number of licensed products appearing in-game, Metal Gear Solid 4 remains mostly intact.

You still have the iPod; you still see the PS3 console in your plane. However, Konami has gone back and remixed some elements to make sense on modern consoles, with the famous "we're on PS3" scene changed to your platform of choice – even if the almost 20-year difference in Christopher Randolph's voice is pretty obvious.

In fact, the only thing I noticed that was missing was the now-redundant installation menu featuring Snake smoking between chapters.

(Image credit: Konami)

Despite its notoriety, Metal Gear Solid 4 is a strange game. I'd even argue it's the worst of the six main games in the series, even if it is still a great time.

It gets off to a really strong start with chapters set in the Middle East and South America, featuring some incredible stealth gameplay with a unique battlefield stealth mechanic which allows you to manipulate the ongoing war between rebel militia and PMC groups.

It's a genius bit of game design, and it’s a shame that it's abandoned by the end of the second chapter.

After an Act 3, which boils down to a not-so-great tailing mission, Metal Gear Solid 4 goes off the rails with its plot, lengthy cutscenes, and gameplay segments.

Some of these are highlights of the game, such as the entirety of Act 4 or the final boss fight that ranks among the best in the series. However, some plot decisions and retcons make for the messiest story in the franchise's history – I won't go into detail, but a particular retcon about Metal Gear Solid 3's Para-Medic is a terrible choice that I'm still bugged by years later.

(Image credit: Konami)

What's there is still a great game I'd recommend to anyone, though it's a shame some of the most interesting gameplay elements are abandoned almost entirely, even if the rollercoaster ride of set-piece and fan-service moments is a joy to play through.

The opening to Act 4 in particular is one of my favorite gaming moments ever and still gives me chills to this day.

We're nomads, and we always will be

(Image credit: Konami)

Peace Walker, on the other hand, is an unsung part of the saga. Originally released on the PSP, it was somewhat overlooked, but this game is Metal Gear Solid 5 in all but name.

Taking place between Snake Eater and The Phantom Pain, Peace Walker is Metal Gear crossed with Monster Hunter, with a mission-based design that allows for short bursts of play. It introduced base management, weapon development, and nicking soldiers via balloon aspects that The Phantom Pain would later expand upon.

Best Bit

(Image credit: Konami)

Returning to Shadow Moses in Act 4 of Metal Gear Solid 4 is one of the greatest bits of fan service ever deployed in a game, and the needle drop of The Best Is Yet To Come still sends chills down my spine almost 20 years later.

While not quite on the level of the original trilogy, Peace Walker has a great story, which helps show how Big Boss eventually became the person we saw as a villain in the original MSX games – arguably better than The Phantom Pain did. It also features one of the best supporting casts in the series, with Kazuhira Miller returning from Metal Gear Solid and being fleshed out into one of the series' best, while the likes of Amanda, Paz, and Professor Galvez are memorable additions to the series canon.

I say Metal Gear crossed with Monster Hunter, but nowhere is that more felt than in the bossfights, which have you take on massive mecha over a prolonged period. While I like this in theory, these bosses are unfortunately the worst the series has to offer – and aside from the titular Peace Walker and true final boss – are not memorable in the slightest.

(Image credit: Konami)

Peace Walker also features co-op and versus multiplayer modes, which are present in this new collection, though I wasn’t able to test them over the review period. From my experiences with the HD Collection version, Peace Walker multiplayer is a blast.

This port is a bit rougher than Metal Gear Solid 4, mostly because there isn't too much you can do to make PSP assets shine in 4K. However, the addition of new higher-quality versions of the comic-book-like cutscenes by Ashley Wood is a welcome bonus

It's not over yet

(Image credit: Konami)

If you played Master Collection Vol 1, you will have a good idea of what to expect this time around. Both main games come with a Screenplay Book and Master Book, which features rundowns of the story, background info on characters, and strategy-guide-esque details to let you find secrets in-game.

There are also soundtrack selections and the ability to download other language versions of the games if you please. The Metal Gear Solid 4 edition also features the Metal Gear Solid 4 Database app, which was released on PS3 around the launch of the original game, although sadly its version of Metal Gear Online is not present.

And while the bonus titles included the original MSX2 version of Metal Gear and Metal Gear 2 – which I wrote about how good they are a few years back – and the less-good NES versions of them, this time around the only additional title is Metal Gear: Ghost Babel for the Game Boy Color.

(Image credit: Konami)

Ghost Babel takes gameplay akin to Metal Gear 2, but fuses it with some elements of the PS1 classic while adapting it to a stage-based formula for handheld. While it doesn't quite match the main series, Ghost Babel should not be slept on, as it's one of the best spinoff titles the franchise has to offer.

Sadly, that's it though. It's a shame, as the likes of Metal Gear Acid or Metal Gear Rising would've been much appreciated additions.

Metal Gear Solid Master Collection Vol.2 is another package of two fantastic games, with some great bonuses included. While it's mostly a vehicle to get Metal Gear Solid 4 out of PS3 jail, Peace Walker and Ghost Babel are two brilliant additions that you might have missed the first time around.

Should you play Metal Gear Solid Master Collection Vol.2?Play it if...

You didn't have a PS3
Metal Gear Solid 4 is out of PS3 jail after almost 20 years. For those who have been waiting, this is one of the biggest dream ports to finally get, and being able to play it in a version that sticks to a solid 4K resolution and 60fps is worth a return trip even if you did play it on PS3.

You never played Metal Gear Solid: Peace Walker
Yes, Metal Gear Solid 4 is the main event of this collection, but Peace Walker is one of the best games the series has to offer. Not to mention an essential part of the storyline heading into Metal Gear Solid 5.

Don't play it if...

You haven't played Metal Gear Solid Master Collection Vol 1.
Metal Gear Solid 4 was made with the idea that you'll have the knowledge of the original trilogy of Metal Gear Solid games at a bare minimum, leading to a plot that will be extremely hard to follow for anyone not caught up. Not to mention that all of the emotional impact will be nullified without prior knowledge

Accessibility features

Both Metal Gear Solid 4 and Peace Walker have subtitles and controller remapping features, but otherwise the accessibility features included in this collection are the same as they were back in the original releases.

This means no color blind support, audio description, or gameplay tweaks are available – which could cause issues during moments in both games when tons of button mashing is required.

How I reviewed Metal Gear Solid Master Collection Vol.2

I played 12 hours of Metal Gear Solid Master Collection Vol.2 on PS5 Pro using a DualSense Wireless Controller, on an LG C5 OLED TV with an LG-US70TY soundbar.

During this time, I played a full run of Metal Gear Solid 4 while skipping most cutscenes (as I had replayed the game very recently for comparison) through to the end of Act 4. I played through to the end of Chapter 1 of Metal Gear Solid Peace Walker, and roughly half of Metal Gear: Ghost Babel. I have played all of these games multiple times before this review, and was able to compare to other versions where applicable.

First reviewed August 2026

A charter aircraft carrying eight people crashed at a remote radar site in western Alaska on Thursday, killing all on board, the U.S. military said.

(Image credit: Andrew Harnik)

Tens of thousands of people in Indiana are entering a second week without power after a deadly storm toppled trees and power lines. Schools remain closed and businesses have been unable to reopen.

(Image credit: Nam Y. Huh/AP Photo/Nam Y. Huh)

Federal agents seized electronic devices from former Rep. Eric Swalwell and searched his Washington home as part of an investigation into allegations of sexual misconduct by the ex-Democratic congressman from California.

(Image credit: Rich Pedroncelli)

Jurors on Thursday heard Duane "Keffe D" Davis tell detectives in 2008 that his nephew Orlando "Baby Lane" Anderson fired the shots that killed Tupac Shakur.

(Image credit: Steve Marcus)

• More than 2,500 organizations, including Cisco, Samsung, AWS, Airbus U.S. Space & Defense, Thales, and the London Stock Exchange Group, have credentials harvested during a supply-chain attack on LiteLLM
• LiteLLM was not directly hacked by the hacking group TeamPCP, which found their way in thanks to a compromised build of an open-source security scanner
• Some of the credentials still work, nearly five months after the original breach, indicating that there is still a persistent security risk until they are changed

Security firms CloudSEK and Hudson Rock have claimed more than 2,500 organizations have had credentials harvested in a supply-chain attack on LiteLLM.

LiteLLM, an open source gateway which translates API calls for over 100 large language models into a single OpenAI-compatible format, was not directly compromised in the attack, as hackers targeted a known vulnerability in Aqua Security's Trivy.

The list included many large and critical service providers, including but not limited to Cisco, Samsung, Salesforce, and Amazon Web Services, as well as Airbus U.S. Space & Defense, Thales Group, Deutsche Bahn, Munich Re, and the London Stock Exchange Group.

An attack that is still a concern nearly five months later

The original attack occurred on March 24 2026 and was spearheaded by a financially motivated hacking group called TeamPCP, which compromised Trivy, an open source security tool that scans for vulnerabilities.

The modified package, which was subsequently downloaded and 'invited' in by LiteLLM without checking its ID- an automated process that essentially allowed a poisoned version of the trusted tool in- gained server administrator privileges and then installed a stealer.

The stealer compromised credentials and secrets far more valuable than corporate data, including Cloud keys, SSH keys, Kubernetes tokens, environment variables, repository and package-publishing tokens, and AI provider keys.

These are arguably worse from a security standpoint than a singular breach because of both the scale of the attack and the fact that hackers now had a 'key' to many security doors rather than having to run exploits to get there.

The victim-scale research done by CloudSEK was further corroborated the following day by Hudson Rock, and it painted a grim picture of what was still an outstanding issue nearly 5 months after the original attack.

The irony is that some of the credentials still work: Independent researcher Kevin Beaumont said some of the compromised keys were still valid after he tested them, even as the impacted organization insisted it had 'rotated' those keys to new ones.

CloudSEK's figures indicate 2,500-plus companies and 434,000 CI/CD pipelines were compromised, while Hudson Rock has released a 153 GB archive of the exfiltrated material after examining a 195 TB file it had obtained. Both firms are running domain-lookup tools so organizations can check their own exposure online.

Whether these revelations lead organizations to double-check their use of AI tools in multiple mission-critical instances that could compromise not only customer data but their own trade secrets down the line remains to be seen.

I've said it before and I'll say it again: new HBO Max spinoff Stuart Fails to Save the Universe is even better than original series The Big Bang Theory.

Sure, it's come under fire for its potentially AI-ridden title sequence and we've still yet to see Sheldon (Jim Parsons) and Leonard (Johnny Galecki) make a return, but I'm having so much fun watching that I've barely given these minor issues a second thought.

We've been hopping from universe to universe at lightning speed, with each variant of Stuart (Kevin Sussman), Denise (Lauren Lapkus), Kripke (John Ross Bowie) and Bert (Brian Posehn) being richer and more hilarious than the last. The TBBT cameos we have had have been ridiculously left-field, and frankly, I love the fact that I cannot guess what on Earth will happen in the next episode.

Given how the concept of the show itself can be taken literally anywhere in space and time, it makes logical sense for Stuart Fails to Save the Universe to run for multiple seasons. But according to its creators, there's "no big picture" for any kind of future... even though season 2 has already been picked up.

'This is chasing what seems to be fun and exciting to do'

(Image credit: HBO Max)

"It would be great if I had the big picture," Chuck Lorre tells me. "Wouldn't that be swell? I don't. There's no big picture. This is just chasing what seems to be fun to do... the big picture would require a bigger person."

Zak Penn adds, "We purposefully wrote ourselves into corners. We like doing that and then saying 'How are we going to get out of this?' That was definitely part of the design of how the team writes and how they wanted to approach it. It makes it more fun not really knowing where we were going to be and how we were going to get there.

"Eventually, we figured some stuff out and came up with an ending. But [the show] was never designed to be 'Oh, we know where all of this is going.'"

Given how easy it would be to create Stuart Fails to Save the Universe season 2 — both from a narrative and production point of view — this news is somewhat alarming. But don't panic... I've decided to take matters into my own hands.

For now, I'm continuing on with the perhaps-delusional belief that I could find myself in the season 2 writers room within a matter of months. So, I've pitched the team two ideas for future storylines.

(Image credit: HBO Max/CBS)

The first is addressing the really big picture: use Stuart Fails to Save the Universe to address the existing problem of George Cooper being dead, and bring him back to life through an alternate universe. That way, Young Sheldon fans get the ending they really wanted.

"We'll definitely keep your ideas in mind," Penn says. "There are an infinite number of universes within this universe, so the universe that you're looking for is going to exist by definition. Whether it makes it to TV is another question."

So far, so good. But I hammer this home with another idea: explore the episode 2 universe that reveals film and TV's Christine Baranski is an AI overlord.

Bill Prady responds, "We'll have to see how the rebellion goes and what happens afterwards. What do they do to her afterwards?"

"She'd be tried," Lorre agrees. "Well there we go — there's something we can do if we get a second season."

Bingo. Over to you, HBO.

• UK movie theatres considering smart glasses ban
• The ban would be over privacy and piracy concerns
• It's yet another instance of private bans restricting the tech

Privacy concerns have rocked Meta’s smart glasses business. They’re being ridiculed online as ‘perv glasses,’ and their use is banned in a growing number of spaces. In the UK, this could also soon include cinemas, though it’s for a very different p-word: piracy.

The main concern with camera glasses, like the ones Meta produces with its partners Ray-Ban and Oakley, is that people can record in a fairly secretive way. This is an issue for the general public as they don’t know when or if they’re being recorded, and for venues like cinemas and theatres as it allows movies and live shows to be subtly recorded by the audience.

With a stable enough internet connection, the viewing could even be live-streamed.

The ban hasn’t taken effect yet, but its consideration is another blow for smart glasses makers and users, who are finding it increasingly difficult to wear their specs because of their spy-camera capabilities.

(Image credit: Meta)

Meta would argue that recording isn’t covert. There is a light that comes on and stays on while you record. Plus, Meta has taken numerous steps to disable recording capabilities if this light is tampered with — either via a simple cover or by fiddling with the hardware.

However, critics have argued the light is easy to miss if you don’t know to look for it, especially if you're outdoors during the day or in a bright place.

In a dark theatre or cinema, the recording light would arguably be a lot more noticeable, though the proposed ban would reduce the risk that bad actors sneak through — while also addressing privacy concerns the glasses raise for other patrons, and banning glasses that allow more covert filming.

Because while Meta is the most well-known smart glasses maker, and so is taking the heaviest share of the backlash, it isn’t the only player in town. Some glasses manufacturers aren't as strict about tackling modders who disable their recording indicator, while others make the indicator even less noticeable or don’t ship with one installed.

(Image credit: Future / Hamish Hector)Going, going, gone?

Governments have also been considering action against smart glasses; however, we have yet to see much in the way of widespread bans. At this rate, we might not need to.

The court of public opinion has issued an overwhelmingly negative verdict on smart glasses. The privacy concerns for people wearing the glasses, and for passersby caught in their gaze who never consented to having their life captured by smart specs,

Even previous champions of the technology I’ve spoken with — journalists like myself who have used and tested, and loved these glasses — have stopped wearing them outside of our reviews. I’ve written before how I’m genuinely fearful I’ll get accosted in the street if I’m caught wearing these glasses.

Besides the threat of retaliation, the steady banning of these glasses across various venues makes them increasingly useless. That’s doubly true for folks who need prescription glasses — smart glasses bans often say that the glasses having prescription lenses is no excuse, meaning you could spend $500 / £500 / AU$750 plus on a pair of glasses you can’t wear anywhere.

(Image credit: Meta)

Is this the end of smart glasses then? A serious perception overhaul could help, though it’s unclear how that would be achieved without such major software changes or hardware alterations, such as removing the camera altogether — though, from experience, cameraless smart specs don’t impress me. It’s too useful a feature to lose out on.

To that end, it’s a lose-lose, and that’s a shame for a category I think is quite fun and cool when people aren’t being creeps with them. They’re also proving a handy tool for people with visual impairments as the glasses can see and describe the world for them.

We’ll have to wait and see what happens, but I expect it’ll continue to be a rocky road ahead for smart glasses.