TechRadar News

Looking for a different day?

A new NYT Connections puzzle appears at midnight each day for your time zone – which means that some people are always playing 'today's game' while others are playing 'yesterday's'. If you're looking for Friday's puzzle instead then click here: NYT Connections hints and answers for Friday, August 8 (game #789).

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

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

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

NYT Connections today (game #790) - today's words

(Image credit: New York Times)

Today's NYT Connections words are…

• INSIDE
• PRODUCT
• MAZE
• SYRUP
• FRUIT
• KNOCK-KNOCK
• LET ME IN
• MUFFIN
• PRACTICAL
• GHOST
• RETURN
• SILK
• YIELD
• TWILIGHT
• DAD
• CASPER

NYT Connections today (game #790) - hint #1 - group hints

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

• YELLOW: The final outcome
• GREEN: Funny ha-ha 
• BLUE: Spooky love stories 
• PURPLE: Add a word that goes with flakes

Need more clues?

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

NYT Connections today (game #790) - hint #2 - group answers

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

• YELLOW: RESULT 
• GREEN: KINDS OF JOKES 
• BLUE: MOVIES FEATURING SUPERNATURAL ROMANCE 
• PURPLE: CORN-Y THINGS 

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

NYT Connections today (game #790) - the answers

(Image credit: New York Times)

The answers to today's Connections, game #790, are…

• YELLOW: RESULT FRUIT, PRODUCT, RETURN, YIELD
• GREEN: KINDS OF JOKES DAD, INSIDE, KNOCK-KNOCK, PRACTICAL
• BLUE: MOVIES FEATURING SUPERNATURAL ROMANCE CASPER, GHOST, LET ME IN, TWILIGHT
• PURPLE: CORN-Y THINGS MAZE, MUFFIN, SILK, SYRUP

• My rating: Hard
• My score: 3 mistakes

I’ll be honest, I absolutely hated Connections yesterday. Not just because I failed, but because it didn’t make sense to me – even in hindsight.

I didn’t do much better today, but at least I understood the links once I finally limped across the line with four correct groups.

My mistake meltdown came in trying to connect what became KINDS OF JOKES. With just eight tiles left I eventually got lucky, but equally I was annoyed that I hadn't seen the easiest of today’s quartets. 

Yesterday's NYT Connections answers (Friday, August 8, game #789)

• YELLOW: MAKE EQUAL, AS A SCORE DRAW, EVEN, SQUARE, TIE
• GREEN: BESTOW CONFER, GRANT, PRESENT, VEST
• BLUE: THINGS YOU MIGHT DO DURING A BORING CLASS/ MEETING DOODLE, DOZE, PASS NOTES, SPACE
• PURPLE: WORDS AFTER THE LETTERS "MS" DEGREE, MARVEL, PAC-MAN, PAINT

What is NYT Connections?

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

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

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

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

It's time to unwind after a long week of work – and what better way to do so than streaming one or several new movies and shows that have dropped this week?

Unsurprisingly, Wednesday's second season has been the talk of the entertainment town over the last few days. However, many of the world's best streaming services have also got something interesting to offer, so one of the biggest Netflix TV Originals of all-time isn't the only thing worth watching this weekend. Hop to it, then, and learn more about what else has arrived. – Tom Power, senior entertainment reporter

Wednesday season 2 part 1 (Netflix)

The queen of anti-socialness, goths, and the macabre has returned. Wednesday Addams is back on Netflix for her sophomore semester at Nevermore Academy – and, considering it's been nearly three years since the hugely successful show's first season dropped, it's high time we were reunited with the psychic teen.

Wednesday season 2 won't be an easy ride for Jenna Ortega's eponymous character, though. Indeed, season 2's official trailer reveals she'll be in a race against time to save the life of her self-installed bestie Enid Sinclair. There'll be numerous other intriguing subplots to unravel, too, with the wider Addams Family having a larger role to play in proceedings than last time. Here's hoping one of the best Netflix shows' second outing is as kooky and spooky as its forebear. – TP

• Watch Wednesday season 2 on Netflix

King of the Hill season 14 (Hulu/Disney+)

Can you believe that little Bobby is all grown up? Hank and Peggy have returned to Middle America after a lengthy retirement, and it's incredible how little their world seems to have changed in the interim. In reality, King of the Hill has been off our screens for the last 16 years and it's the perfect dose of nostalgia for the week.

The Hulu and Disney+ show as charming as I remember, and has somehow managed to pick up exactly where prior seasons left off. If you sewed them together, you'd hardly be able to see the seams. Frankly, that's a massive two fingers up to every other legacy sequel in existence. I promise it'll exceed your expectations, even if you can't fully remember the full Hank-lore (you don't really need to). – Jasmine Valentine, entertainment writer

• Watch King of the Hill season 14 on Hulu (US) and Disney+ (internationally)

The Pickup (Prime Video)

I love watching an action movie on a Friday night but, after reading early reviews for one of August's new Prime Video movies, I'm not getting my hopes up for The Pickup. Billed as a heist comedy, it pairs Eddie Murphy and Pete Davidson as armored truck drivers caught in a criminal ambush.

On paper, this Prime Video flick sounds like the kind of plot that could set up thrilling action sequences – but I’m not convinced this one’s going to deliver. Considering Murphy's track record with buddy comedies, his performance might be the only bright spark in what could otherwise be a car crash of an Amazon film. – Amelia Schwanke, senior entertainment editor

• Watch The Pickup on Prime Video

Platonic season 2 (Apple TV+)

From the moment I saw the trailer for Platonic season 2, I knew the Apple TV+ series was onto another winner. The BFF (and things have strictly remained platonic) chemistry between Rose Byrne's Sylvia and Seth Rogen's Will is exactly the kind I hope I grow up to have, even if it comes with too many midlife hurdles for two people to handle. As Apple puts it: “The duo tries their best to be each other’s rock – but sometimes rocks break things.” That only means one thing... a huge helping of delicious drama.

As far as I'm concerned, tension and fallouts between two close friends hits a lot harder than any lovers tiff or evil rivalry, and there's a genuine honesty and vulnerability that comes alongside the laughs this summer. It's the perfect summer comedy we've not yet been treated to, if you discount the bloody accidents on the golf course. – JV

• Watch Platonic season 2 on Apple TV+

The Monkey (Hulu)

Osgood Perkins scared us all silly with the Nicolas Cage-led serial killer horror Longlegs, last year. Now, he's back with something very different. Based on a Stephen King story, The Monkey follows twin brothers who discover a cursed toy monkey that causes horrific deaths around them.

It leans more into comedic gory horror than straight-up frights and there are some seriously shocking moments throughout. One of seven new horror movies on Prime Video, Shudder, HBO Max, and more you can't afford to miss this month. – Lucy Buglass, senior entertainment writer

• Watch The Monkey on Hulu

Freaky Tales (HBO Max)

Pedro Pascal has been a busy boy lately. Whether he's navigating the apocalypse in The Last of Us season 2 or putting his spandex on for The Fantastic Four: First Steps, he's been on our screens a lot, including in this anthology flick that's out now on HBO Max.

In Freaky Tales, we follow an NBA star, a corrupt cop, a female rap duo, teenage punks, neo-Nazis and a debt collector set against an '80s backdrop. It's as varied as it sounds, with the movie showing us four interconnected stories that are sure to entertain. With a 74% Rotten Tomatoes critical rating, it's worth checking out this weekend – LB

• Watch Freaky Tales on HBO Max

Stolen: Heist of the Century (Netflix)

Remember those headlines about the world's biggest diamond heist back in 2003? The story of Italian thieves robbing an estimated $100–500 million from the vault of the Antwerp World Diamond Centre in the dead of the night has left many scratching their heads since, especially considering most of the gems have never been recovered.

Now, thanks to authors Scott Andrew Selby and Greg Campbell, we finally have the inside story. Their book Flawless detailed the daring operation and it’s now been adapted into a Netflix documentary. Think real-life Ocean’s Eleven, produced by one of the minds behind the brilliant Don’t F**k with Cats. I’m already sold. – AS

• Watch Stolen: Heist of the Century on Netflix

You might also like

• I watched the first six episodes of Alien: Earth, and it's the Peter Pan-inspired sci-fi horror show I didn't know I needed
• Foundation season 3's most explosive episode yet is just the start of 'the wheels coming off the wagon', star teases
• The Batman Part II will reportedly feature one of The Dark Knight's most famous allies – and their inclusion might kill off a big fan theory about its plot

• Foundation season 3 may have just killed off another major character
• This season's latest chapter sets the scene for five more incendiary episodes to come
• One of the sci-fi series' stars has teased what'll happen next

Foundation season 3 episode 5 is out now – and it's an explosive entry that might set up an incredibly tense and life-threatening second half of the Apple TV+ show's latest installment.

Titled 'Where Tyrants Spend Eternity', Foundation season 3's latest chapter is full of shocks aplenty, one of which appears to be the demise of another major character. Full spoilers immediately follow for this season up to and including episode 5, so turn back now if you haven't watched it yet.

Someone's going for an unexpected space walk! (Image credit: Apple TV+)

Following Hari Seldon's supposed death in Foundation season 3 episode 2, the sci-fi epic appears to bump off another member of its primary cast – Brother Dawn – in 'Where Tyrants Spend Eternity'.

That might not come as a big surprise to some viewers. After all, a previous version of Dawn, i.e. one of three clones of Cleon I who tyrannically rule the Galactic Empire, was killed in Foundation's first season. The fate of season 3's iteration has been debated over the last couple of weeks, too, especially after Dawn formed a uneasy alliance with Gaal Dornick, who has ties to Empire's arch-rivals in the Foundation and the Second Foundation, in this season's third entry.

Unsurprisingly, Dawn was just a pawn in Gaal's all-consuming quest to try and defeat The Mule, aka this season's Big Bad. In episode 5, Gaal convinces Dawn to form an Imperial blockade around Kalgan, the independent world that The Mule took control of in Foundation season 3's premiere. Dawn does so by blackmailing Tarisk, one of the most influential members of Empire's Galactic Council, which is based on Clarion Station, and whom Kalgan's jurisdiction falls under, who helps to force the rest of his councillors to approve Dawn's request.

Dawn doesn't realize he's going to regret blackmailing Tarisk (Image credit: Apple TV+)

The aim of the blockade? Prevent food and medicine entering Kalgan due to the Imperial armada surrounding the planet, and effectively starve The Mule and his forces until they surrender. Okay, Kalgan's civilian population would also suffer, but Gaal insists it's a small price to pay to thwart this season's primary villain.

Predictably, things don't go according to plan. Once the armada is in position around Kalgan, The Mule reveals he foresaw what would happen and prepared to counter it. After taking control of the nearest jump gate to Kalgan in episode 3, his troops laced it with Blue Cobalt, an explosive substance that, when fired into a nearby star, causes a huge, powerful, and incredibly hot solar flare to erupt and incinerate anything in its path. Long story short: the armada is destroyed in the blast, and Kalgan and its populace are burned to a crisp.

Fleeing the scene and locating an airlock that'll allow him to reunite with Gaal with the aid of the latter's Foundation operative, Dawn soon learns Kalgan and the Imperial armada's destruction was also part of Gaal's own scheme. Indeed, Gaal tells him – via their telecom devices – that the Foundation needed Empire's fleet to be destroyed as that would accelerate its decline and allow the Foundation (plus the covert Second Foundation) to thrive.

Who knew that one of Empire's biggest archnemeses would betray you, eh, Dawn? (Image credit: Apple TV+)

Furious and understandably feeling betrayed, Dawn severs ties with Gaal. However, as he prepares to leave the aforementioned airlock, Dawn is confronted by a grief-stricken Tarisk, whose family was killed on Kalgan. Despite Dawn's pleas to put down his weapon, a vengeful Tarisk opens fire, causing the chamber to depressurize and sucking the pair of them out into space.

Now, Dawn managed to put on a full spacesuit before he was confronted by Tarisk, so it's highly likely he isn't dead. Nonetheless, Brother Dusk, the only one of Empire's three rulers who's still running the show – remember, Day has fled the coop to pursue his own quest in Mycogen – believes Dawn has perished. That much is made clear in the holographic message Dawn sends to Dusk before the council meeting, Kalgan's destruction, and everything that comes after.

Is Dusk about to go all Emperor Palpatine/Darth Vader on the galaxy? (Image credit: Apple TV+)

So, what next for the ailing Empire as it tries to hold onto the dwindling amount of power it wields in the galaxy? Cassian Bilton, who plays Dawn, wouldn't be drawn on what happens next when I asked him for details ahead of this season's debut.

However, Terrance Mann, who portrays Dusk in one of the best Apple TV+ shows, was more forthcoming. And, with a Black Hole Gun – think of it like the planet-killing Death Star in Star Wars – at his disposal, the emotionally devastated Dusk might be so consumed by rage and sadness that said weapon is taken for a ride or two very soon.

"Dusk, and Demerzel and the other Cleons for that matter, are desperately clinging on, "Mann said. "And now things have occurred that force him to make really hard choices about who's going to live, who's going to die, and what's ultimately going to save Empire.

"This whole season is all about the wheels coming off the wagon, and he's left having to do something he never thought he'd have to do," Mann continued, "A large part of that has to do with thinking Dawn has died. At that point, he's completely lost and thinks 'I have nothing else to do but exert this power [the Black Hole Gun] that I have and eradicate everything except for Empire."

If that doesn't set the stage for five more unmissable chapters of this Apple TV Original, I don't know what will. New episodes of Foundation air every Friday, so we've got a wait on our hands to see where things go from here.

You might also like

• Foundation season 3 is shedding new light on one of its most complex characters – and their actor suggests things might not end well for her: 'She's spiraling inwards'
• Who is Magnifico Giganticus in Foundation season 3? The Apple sci-fi series' most important new character explained
• 'We're the last to find out': Foundation stars don't know if it'll return for season 4 – but they haven't given up hope for the Apple TV+ show's renewal

• OpenAI’s GPT-5 model will soon be available alongside Apple Intelligence
• That means more powerful ChatGPT features on your Apple device
• But this move may make life harder for Apple’s rumored chatbot

OpenAI has just released its latest GPT-5 artificial intelligence (AI) model, and it’s set to bring a whole host of powerful new features to the ChatGPT chatbot. Thanks to Apple’s integration of ChatGPT into Apple Intelligence, that could bring benefits to a wider set of people than merely ChatGPT users – but it also brings some serious risks to Apple.

When might we see GPT-5 in Apple Intelligence? Well, news outlet 9to5Mac thinks it has the answer. The publication claims that Apple told it that GPT-5 will arrive on Apple devices with the release of iOS 26, iPadOS 26, and macOS Tahoe. That’s due to happen in early September, meaning there’s only about a month to wait before GPT-5 is directly incorporated into Apple’s products.

You'll apparently be able to access GPT-5 in a few different ways on iPhones and Macs. The most obvious being when you use Siri for more complex questions that it can't answer, but also with Apple's Writing Tools and Visual Intelligence, which uses your iPhone's camera to help answer questions about what you're looking at.

The new features in GPT-5 include different personalities (including Cynic, Listener, Nerd, and Robot), fresh theming options, and an improved Voice Mode. Google users will also soon be able to connect their Gmail, Google Calendar, and Google Contacts accounts directly to ChatGPT.

GPT-5 comes in a few different sizes, including the regular GPT-5 model, GPT-5 mini, and GPT-5 nano. These offer different levels of reasoning and are all available to free users. If you’re paying $200 a month for ChatGPT Pro, you’ll also get access to the deeper GPT-5 pro and GPT-5 thinking models, which take longer to respond but offer more insightful responses.

The wait goes on

(Image credit: Future/Jacob Krol)

While ChatGPT can be used in conjunction with Apple Intelligence, you have to specifically invoke it. Most queries to Apple’s AI will use Apple Intelligence, but if there’s something that Apple’s model can’t answer, it will ask if you want to hand off to ChatGPT.

Right now, relying on ChatGPT for more in-depth queries doesn’t reflect particularly well on Apple Intelligence, but there are signs that Apple is working to correct that. The company is reportedly developing its own AI chatbot to rival ChatGPT and others (despite previously telling TechRadar “that was never the goal, and it remains not our primary goal”).

Unfortunately, we don’t know when that chatbot will be ready, which could be why Apple is moving so swiftly to incorporate GPT-5 into Apple Intelligence: it might be an attempt to keep Apple’s AI relevant while the company works on its own solution. But the longer we have to wait – and the more features OpenAI adds to ChatGPT – the higher the expectations will be on Apple’s chatbot, and the harder it will be for Apple Intelligence to eventually usurp its rivals.

You might also like

• OpenAI GPT-5 launch live – all the latest news as Sam Altman unveils the new model
• GPT-5 is here – 5 things you need to know about OpenAI’s ‘most useful’ model yet
• Apple’s rumored ChatGPT rival could fix the thing I hate most about Siri – and it can’t come soon enough

• HBO Max confirmed “aggressive” plans to tackle password sharing, following in Netflix’s footsteps
• The HBO Max password sharing crackdown will begin to be enforced more rigorously in September 2025
• Subscribers are already complaining about the change to their accounts

Head of streaming and gaming at Warner Bros. Discovery JB Perrette confirmed at the company’s second quarter earnings call that HBO Max will begin an “aggressive” crackdown on password sharing as of September 2025.

It’s a move that follows in the footsteps of rival streaming services such as Netflix and Disney+, marking a 180-turn from previous marketing campaigns encouraging users to share their account details (we all remember those Netflix adverts). The idea is to close existing loopholes and drive profits, though the benefits of doing so aren’t expected to be seen until 2026.

As it stands, HBO Max is undoubtedly one of the best streaming services on the market. It hosts original content including The Last of Us, The White Lotus, Euphoria and Peacemaker, with a quality backlog of classic movies and TV shows added to the platform every month.

However, in a world where streamers are getting more expensive and inaccessible to groups of people (e.g., if you’re a family in the same home, it makes sense to share one account, however, HBO Max does not offer a family subscription tier), the decision to make password sharing harder is one that’s already not going down well with its subscribers.

Early X/Twitter comments from users show that subscribers are already complaining. One user said “that's one way to lose subscriptions quickly,” with another adding, “back in the early Netflix days we all shared passwords like Pokémon cards. Now they want a DNA test before I can watch Euphoria??”

Others have noted how close the announcement is to the pulling of all Cartoon Network content from the platform (don’t worry, you can still find a lot of this on Hulu in the US, while those in Australia can find the network’s shows on Stan). It feels like HBO Max is piling on bad news after bad news, but let’s dive deeper into the methods behind their madness.

Opinion: HBO Max’s ‘aggressive’ password crackdown isn’t exactly the news I wanted to hear

Perette explained during the call that months of testing have gone into “who’s a legitimate user who may not be a legitimate user,” adding that the next step is to “turn on the more aggressive language around what needs to happen” so they “are putting the net in the right place, so to speak”. Later in 2025, the crackdown will be happening “in a much more aggressive fashion,” adding “the message language right now has been a fairly soft, cancel-able message, [which will] start to get more fixed and such that people have to take action as opposed to right now, sort of having to be a voluntary process.”

After all of this is established “the real benefit will start probably in the fourth quarter and then kick in in 2026”. But as WBD claims to have added 3.4 million streaming subscribers to its platforms over the last quarter, can we really expect them to hit their new overall target of 150 million subscribers by 2026 if measures like this are being introduced? The fact I’m now having to jump through more hoops just to stream the content I want to is incredibly offputting, and I wouldn’t be surprised if more people turn to free streaming alternatives, or dare I say it, cable TV.

It’s not something that HBO Max itself is worried about, though, with its eyes on the prize of a hefty theatrical slate heading our way over the next few years. As WBD’s CEO David M. Zaslav explained on the call: “one of the assets that we have at this company is that we have such – so much compelling storytelling IP that people know everywhere in the world, whether it's Batman, Superman, Wonder Woman, Lord of the Rings. And then we'll call those the big tent poles, Harry Potter. And then smaller tent poles like The Fugitive, Goonies, Gremlins, Practical Magic that everybody knows.

“Our strategy is [to] light up strategically those big tent poles so that we have two or three of those a year, which provide real stability. We got a great script on Lord of the Rings with Peter Jackson that we're already – that we're moving forward on, and we'll be giving you more detail on that.”

You might also like

• HBO’s Harry Potter TV show is supposed to be a fresh start, but Nick Frost’s Hagrid says otherwise
• Superman's most surprising cameo has got me asking 3 big questions about James Gunn's next DCU project
• IT: Welcome to Derry – everything you need to know about the new Max horror series

We are in the midst of a Copernican shift in enterprise intelligence. We are no longer just automating or augmenting tasks - we are delegating intent. This is not just IT automation; it’s agency. And it’s changing everything…

In this new paradigm, Agentic AI doesn’t just support human workers - it collaborates with them, anticipates their needs and acts independently to drive outcomes. It marks a seismic shift in how we think about intelligence at work.

Agentic AI is about autonomous execution. It doesn’t wait for prompts. It plans, decides and acts—often without human intervention. This is not just a new toolset; it’s a new mindset.

Are we prepared to rethink how work is designed, how systems are integrated and how humans and machines co-evolve across the enterprise?

From prompt to purpose

Generative AI (GenAI) has captured the public imagination with its ability to generate text, images and code. But it is fundamentally reactive - dependent on human input to produce output. Agentic AI, by contrast, is proactive. It understands goals, decomposes them into tasks, orchestrates tools and adapts its strategy in real time. It’s the difference between a brilliant assistant and a self-directed colleague.

In 2025, forward-thinking enterprises are no longer choosing between GenAI and Agentic AI - they’re combining them. GenAI fuels ideation and content creation. Agentic AI delivers execution at scale.

The rise of the autonomous enterprise

Agentic AI is already transforming how businesses operate. In finance, agents autonomously monitor compliance, flag anomalies and initiate remediation workflows. In manufacturing, they optimize supply chains in real time. They can use the internet, make purchases and approve orders.

However, the real revolution is internal. Agentic AI will become the connective tissue of the enterprise - linking systems, surfacing insights and taking action across silos. It’s not just about doing more with less. It’s about doing what was previously impossible.

From process to possibility: Reinventing the enterprise core

Agentic AI isn’t just accelerating existing workflows - it’s reimagining them. In finance, autonomous agents now reconcile transactions in real time, detect anomalies before they escalate and dynamically adjust forecasting models based on live market signals. In HR, agents are transforming talent management by continuously scanning internal and external data to identify skill gaps, recommend personalized learning paths and even initiate retention interventions before attrition risks materialize.

Meanwhile, in Sales & Marketing, agents orchestrate hyper-personalized campaigns, adapt messaging based on behavioral signals and autonomously optimize pricing strategies across channels. These aren’t incremental improvements - they’re structural shifts that collapse cycle times, eliminate friction and unlock entirely new value pools. The result is a more fluid, responsive enterprise where strategy and execution are no longer separated by process drag.

Human + Agent: A new collaboration model

The fear that AI will replace humans misses the point. Agentic AI doesn’t eliminate human work—it elevates it. It frees professionals from the tyranny of the inbox and the spreadsheet, allowing them to focus on judgment, emotional intelligence and radical innovation.

Imagine an operations lead at a manufacturing firm working with an AI agent that monitors equipment health, predicts maintenance needs and autonomously adjusts production schedules to avoid downtime. Or a commercial strategist at an energy company whose agent tracks global commodity prices, models regulatory impacts and recommends contract renegotiations in real time.

These agents aren’t waiting for instructions - they’re anticipating change, acting on insight and reshaping how decisions are made. This isn’t science fiction - it’s already happening in the most forward-thinking enterprises.

Designing for Trust and Impact

To harness the full potential of Agentic AI, organisations must go beyond deployment. They must design for trust. That means:

• Transparent reasoning: Agents must explain their decisions in human terms.
• Ethical alignment: Agents must be trained on values, not just data.
• Human-in-the-loop governance: Autonomy must be bounded by accountability.

This is not just a technical challenge - it’s a leadership imperative.

The new literacy: Prompting, supervising, orchestrating

As Agentic AI becomes embedded in workflows, a new skillset is emerging. Prompt engineering is just the beginning. Professionals must learn to supervise agents, interpret their outputs and orchestrate multi-agent systems. This is the new digital fluency.

Business leaders, as well as HR and L&D, must act now to build these capabilities. The future belongs to those who can lead teams of humans and machines.

Conclusion: From intelligence to intent

Agentic AI is not just the next wave of automation. It’s a new form of intelligence - one that acts with intent. As we enter this new era, the question is not whether we will use Agentic AI, but how we will shape it to reflect our highest aspirations.

I tried 70+ best AI tools.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

• OpenAI's GPT-5 is now available, but many users are angry with the new release
• GPT-5 has replaced the previous AI model selection with some users claiming the upgrade "is horrible"
• ChatGPT Plus subscribers now have limits to how often they can access reasoning models, and have lost access to the older more reliable ones like o4-mini and o4-mini-high

OpenAI just released GPT-5, the next generation of the company's AI model that will power ChatGPT for the foreseeable future.

In an hour-long livestream broadcast yesterday, OpenAI CEO Sam Altman and his team showcased GPT-5's capabilities and improvements over its predecessor, GPT-4o.

However, not even 24 hours later and social media sites like Reddit are being flooded with criticisms of the new AI model, with many users left unimpressed with the next generation of ChatGPT.

One Reddit thread titled "GPT-5 is horrible" has nearly 3,000 upvotes and over 1,200 comments filled with users dissatisfied with the new release.

U/headwaterscarto said, "I like how the demo they were like – “if it gets something wrong, no worries, just ask again. I’m actually going to run 3 prompts at once and pick my favorite.” Like, how is that better?" Another says, "Sounds like an OpenAI version of 'Shrinkflation'"

Many users miss the previous 4o and 4.1 models, with plenty of comments saying things like "I miss 4.1. Bring it back," and "They should’ve let us keep the old models while they fix the new one."

There's also uproar from ChatGPT Plus subscribers who feel like the latest AI model release has actually limited the functionality of the paid subscription. The new GPT-5 Thinking model is limited to 200 messages a week, and Plus subscribers no longer have access to the wide variety of AI models that used to be available, as OpenAI now claims GPT-5 is able to reason when it needs to.

ChatGPT literally got worse for every single Plus user today.There's no way to reliably get thinking models anymore.Before we had o4-mini, o4-mini-high and o3.Now we have GPT-5 Thinking with 200 messages per week and a router that exclusively routes you to some small and…August 8, 2025

Are people against change? Or is GPT-5 as bad as the internet makes out

pic.twitter.com/1u0MOGvJWUAugust 7, 2025

A lot of the uproar surrounding GPT-5 is based on the overpromising from Sam Altman, who hyped up the latest announcement as if it were going to revolutionize the world and the way we interact with AI.

Hours before the official GPT-5 reveal, Altman tweeted an image of the Death Star from Star Wars looming over the horizon of a planet, hinting at a ground-breaking revolution from OpenAI's next AI model. Instead, while GPT-5 smashes benchmarks compared to its predecessors, it's an incremental upgrade compared to the initial AI revolution when ChatGPT first launched.

For many, having access to the reliable ChatGPT-4o models, which Altman claimed were like talking to a college student versus GPT-5's PhD-educated expert, was better than GPT-5's one-size-fits-all approach.

Lots of users are reporting GPT-5 performing worse than 4o, but we've yet to properly test out the new AI model to know for sure if that's truly the case. One thing is for sure: OpenAI's paid subscribers feel hard done by the new release, and the company better iron out the launch bugs, such as slow and poor responses, otherwise its loyal fanbase will look elsewhere.

You might also like

• GPT-5 free tier – 3 cool features you can try now for nothing
• GPT-5 is here – 5 things you need to know about OpenAI’s ‘most useful’ model yet
• Guess the movie from just emojis – try the TechRadar AI Challenge!

• A new AI search feature has been spotted in Apple Maps
• It's available in the latest iOS 26 developer beta
• A full rollout of the software is expected in September

As updated iOS 26 beta versions continue to roll out, we're spotting new features in the software that Apple hasn't officially announced: yesterday it was AirPod charging lights, and today it's AI search in Apple Maps.

The team at 9to5Google has spotted that the search function in Apple Maps now encourages users to "search the way you talk" – so you might type out or speak out a search like "find cafes with free Wi-Fi" for example.

This is now live in the fifth developer beta version of iOS 26, so you won't see it yet if you're running the public beta. It might also get tweaked or even pulled before the final version of iOS 26 launches in September, though it seems likely that it's here to stay.

It's certainly an easier and more intuitive way to search for places, if you need to go beyond the basics and specify some additional criteria for your results – such as Wi-Fi access, food options, electric vehicle charging, wheelchair access, or whatever it is.

Apple Maps vs Google Maps

The update gets Apple Maps closer to Google Maps, shown here (Image credit: Future)

It's possible that the new functionality has only gone live in the US for the time being, because after updating my iPhone to the iOS 26 developer beta 5, I was unable to get the natural language search feature to show up in Apple Maps here in the UK.

What I could do was run a similarly phrased search on Google Maps, and the results were somewhat patchy: it identified cafes without any problem, but it listed several that definitely do not offer free Wi-Fi (and I've been in them and asked).

It shows up a limitation in AI features such as these, which is that while the AI might be great at understanding what you're asking for and converting your spoken words into text, it's still relying on cold, hard maps data to get you accurate results.

In terms of raw data, Google Maps perhaps still has the edge over Apple Maps – but upgrades like this one in iOS 26 will mean Apple Maps becomes more appealing and useful, and may be enough to grab some users back from Google Maps.

You might also like

• Apple Maps gets a tasty Michelin star upgrade
• 5 great new features in the iOS 26 public beta
• The macOS 26 Tahoe update is also on the way

• Unidentified hackers broke into Bouygues Telecom systems
• The hackers stole names, contract data, and more
• Bouygues Telecom urges victims to be wary of incoming emails

French telco giatn Bouygues Telecom has confirmed suffering a cyberattack in which it lost sensitive customer data.

In a short announcement published on its website, the company said it detected the attack on August 4, and following an investigation, determined threat actors stole people’s contact details, contract data, civil status data (or company details), and IBAN numbers.

We don’t know exactly when the intrusion happened, who did it, if it was a ransomware attack or not, or how many people were affected - but acccording to official company data, Bouygues Telecom has 26.8 million subscribers.

Phishing potential

The teleco started notifying affected individuals via SMS, ousted the attackers, and allegedly implemented additional safeguards to prevent further incursions.

“We have blocked the malicious access, increased monitoring of our systems, and implemented additional necessary security measures,” the company said.

Bank card numbers and Bouygues Telecom account passwords were not affected.

While there is still no evidence the information stolen in the attack is being abused in the wild, the very nature of the archive gives it plenty of potential.

Cybercriminals can use it to craft convincing phishing emails, impersonating the company and forcing the victims into action by threatening to terminate their account.

Knowing information such as contract data and IBAN numbers could convince the victims that they are talking to legitimate representatives of the organization.

In a short FAQ, Bouygues said it “strongly recommends” users stay vigilant, and never share their login credentials, or passwords.

“Be especially cautious of calls from fake banking advisors who may try to gain your trust by mentioning your name or account number. If in doubt, hang up and call your bank or advisor back on their usual number.”

Finally, the company stressed that the IBAN number alone is not enough to initiate financial transactions, but users should still monitor their accounts closely.

Via TechCrunch

You might also like

• Top Canadian telecom firms may have been hit by Chinese Salt Typhoon hackers
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

• A new report suggests The Batman Part II will feature Robin
• The rumor emerged less than 24 hours after the movie's filming start date was revealed
• If true, it raises questions about the DCU's own Batman film

The Batman Part II is still over two year away from grapple gunning into theaters, but a new rumor could not only have huge implications for the movie itself, but also James Gunn's DC Universe (DCU).

Let's start with the latest piece of gossip. Yesterday (August 7), industry insider Jeff Sneider suggested that Robin, the Dark Knight's most famous sidekick, is part of the superhero movie's cast. Sneider made the claim in the latest addition of his InSneider newsletter and says he was made aware of Robin's inclusion after sitting down with a "well-placed source" earlier in the week.

Sneider's assertion comes less than 24 hours after Warner Bros. Discovery (WBD) announced filming would begin on The Batman Part II in early 2026. Considering the timing of Sneider's claim, some observers would argue its arrival is all a bit too coincidental. After all, if Sneider spoke to his source before WBD confirmed when the cameras would start rolling on the DC comic book movie sequel, why not reveal it sooner?

Robin's apparent inclusion in Matt Reeves' next Batman film raises even more questions and potentially kills off a big fan theory about The Batman Part II's story. So, let's break down the biggest ones that are on my mind right now.

Let's tackle the aforementioned fan theory first. Some fans have speculated that The Batman 2 will be heavily inspired by 'The Long Halloween', one of the Caped Crusader's most famous comic book series. Their evidence? The Batman's follow-up is slated to arrive in October 2027, so a tale based on Jeph Loeb and Tim Sale's iconic literary works would befit the film's forthcoming release date.

Robin's possible appearance in The Batman Part II pours cold water on that theory because he isn't in 'The Long Halloween'. Okay, The Batman Epic Crime Saga's next big-screen entry could take creative liberties with said literature and include Robin. It wouldn't be the first time a superhero film has changed elements of a comic series it's based on. Nevertheless, I'd be amazed if Reeves and fellow co-writer Mattson Tomlin shoehorned Robin into their story if it is based on 'The Long Halloween', particularly as it would heavily impact the timeline of events from said story.

What does this rumor mean for the DCU's Batman movie, which is also set to include Robin? (Image credit: DC Studios)

Now it's time to try and address some of those big questions I previously mentioned – and, surprise surprise, they're based around a long-running discussion about whether The Batman Epic Crime Saga is secretly being folded into the DCU.

Here are the facts: When James Gunn announced the initial DCU Chapter One line-up in January 2023, one of the projects he discussed was a Batman movie. Titled The Brave and the Bold, it's expected to draw heavily from Grant Morrison and Andy Kubert's fan-favorite 'Batman and Son' comic series. In it, the Dark Knight is stunned to learn he has a son named Damien, whom he takes in and tries to raise. However, without getting too deep into spoiler territory, there's more than meets the eye about the seemingly genetically perfect, yet rage-filled Damien.

Now, even if Robin is confirmed to be part of The Batman Part II's cast, we don't know which one will appear. It could be any one of Dick Grayson, Jason Todd, Tim Drake, Damien Wayne, or one of the less well-known female Robins in Stephanie Brown or Carrie Kelley. There are other Robins, but those listed are the most well-known.

If – and it's a big if – The Batman 2 features Damien, though, fans will rightly ask the question: is The Batman Epic Crime Saga eventually going to be part of the DCU? After all, it would be silly, not to mention incredibly confusing, if we have two Batman films that exist in separate cinematic universes and feature the same super-duo.

For what it's worth, Gunn and Reeves have largely shut down the idea of the Reeves-Verse being absorbed into the DCU. Crucially, though, Gunn has always left the door slightly ajar – the DC Studios co-chief most recently telling Entertainment Tonight: "I’ve said it before. We [Gunn and Reeves] have talked about it before, but it’s not what’s happening right now." (NB: Gunn didn't emphasise the words in italics, but I've done so to hammer home my point that the door isn't closed on these two universes potentially merging).

A post shared by Matt Reeves (@mattreevesla)

A photo posted by on

There is evidence that Gunn's and Reeves' Batman films will continue to exists in separate shared universes. In June, Gunn said "I think I have a way in" when asked for an update on The Brave and the Bold's script. He also revealed an unnamed writer was busy penning it.

Now, you could argue that those things could be related to Reeves' own Batman flick. He and Tomlin were still thrashing out The Batman 2's screenplay when Gunn made those comments. The fact that Gunn didn't mention the name of his movie's writer, plus his "I think I have a way in" quote, could be interpreted as him and Reeves secretly agreeing to fold The Batman's epic crime saga into the DCU. To me, though, that's akin to fans clutching at straws and trying to manifest the integration of the former into the latter.

Frankly, considering Robert Pattison's Bruce Wayne had only been operating as Batman for a couple of years in The Batman, the only way he'd be old enough to have a kid in The Batman 2 is if it featured a big time jump. Sure, that's a possibility, but I don't see it happening.

The best thing Gunn, who's active on social media, can do right now is confirm or deny Sneider's claim. If Robin is set to appear in The Batman Part II, Gunn – or even Reeves himself – could give us some indication as to which one will feature in Reeves' next big-screen project. That's the only way to officially put another pin in this never-ending debate about the state of The Batman Epic Crime Saga's ongoing independence to the DCU.

In the meantime, let me know what you think. Should these two universes remain separate, or is it time for Pattinson's Caped Crusader to make the leap to the DCU? Sound off in the comments.

You might also like

• Here's everything we know so far about The Batman Part II, including its release date
• Find out how to watch the Batman movies in order
• Or see if you agree with my raking of the best Batman movies

• Avalanche Studios' Contraband is seemingly no more
• The game only ever received one public-facing CG trailer
• There has been no news on the game in the four years since its announcement

Xbox Game Studios has canceled Contraband, a game I wouldn't blame you for not remembering, as we only ever saw it once in the form of a CG trailer.

Contraband, being developed by Just Cause and Mad Max's Avalanche Studios, was initially revealed in 2021, accompanied by a trailer that didn't give much away beyond its gritty 1970s setting.

Initially reported by Bloomberg, it seems that Contraband has now been canceled, or has at least had its development halted for the time being. Reporter Jason Schreier states that the project has been canceled "after four years of radio silence."

Avalanche Studios followed up with confirmation on its own website, saying: "Over the past several years, Avalanche Studios Group and Xbox Game Studios Publishing have collaborated on Contraband. Active development has now stopped while we evaluate the project's future. We're thankful for the excitement we've seen from the community since we announced and will give an update on what's next as soon as we can."

A cancellation at this point may not be all that surprising for Contraband, but regardless, it's not a good look for Xbox. The company recently canceled two high-profile games in Rare's Everwild and The Initiative's Perfect Dark reboot. That's in conjunction with another recent round of mass layoffs, where thousands of employees across several studios were let go.

Back in July, Xbox boss Phil Spencer described the actions taken as "tough decisions," claiming Xbox's "platform, hardware, and game roadmap have never looked stronger" despite the seeming instability of the Xbox Game Studios umbrella.

Also check out...

• After Concord's flop, Sony admits its live-service plan 'is not entirely going smoothly' but says it will continue and learn from its mistakes
• Hollow Knight: Silksong was a no-show once again at the Nintendo Indie World Showcase, but here's everything else that was announced
• Mafia: The Old Country will have quality and performance modes on PS5 and Xbox and is officially confirmed to be PS5 Pro-enhanced

• Google pledges $1 billion in funding for AI education and training across the US
• Eligible students also get 12 months' Google AI Pro, with Gemini 2.5 Pro and more
• The news comes just weeks after Microsoft announced a similar scheme, but at 4x the value

Google has confirmed a commitment to spend $1 billion over the next three years to support AI education and training across the US, targeting colleges, universities and nonprofits.

Alphabet CEO Sundar Pichai explained in a blog post more than 100 colleges and universities have already signed up to its AI for Education Accelerator program, which hopes to make AI and career training free for every college student in America.

Google's training efforts add to its already extensive involvement within education – the company already runs schemes to get Chromebooks into the hands of students in multiple countries and regions.

Google wants to train America's students in AI

In the post, Pichai explained his own personal motivation for the project: "Having regular access to computers in grad school changed my life, and led me on the path to Google. It’s my hope that bringing the best AI tools to college students will open up new worlds for them, too."

The latest announcement sees eligible students across the US, as well as Japan, Indonesia, Korea and Brazil, able to sign up for a year's free Google AI Pro plan with Gemini 2.5 Pro access, Deep Research via Gemini 2.5 Pro, Veo 3, higher limits for Jules, NotebookLM and 2TB of storage.

Boasting about its presence across the education sector, and offering a small dig at Microsoft which previously dominated the area, Pichai noted that more than four in five of the top 100 US universities now use Google Workspace for Education.

However, Google isn't the only company investing in upskilling the current generation of students and our future workforce.

Microsoft President Brad Smith recently shared details of a $4 billion investment in AI and cloud technology for K-12 schools, community and technical colleges, and nonprofits. Other companies in this space, including Meta, OpenAI, Anthropic and Amazon have all pledged their own support, too.

You might also like

• These are the best AI tools and best AI writers
• Check out our roundup of the best online learning platforms
• Microsoft is spending $4 billion to push AI in schools, universities and more

• Available for Galaxy Buds 3 and Galaxy Buds 3 Pro
• Requires Android 16 and One UI 8
• One UI is still a beta for some Galaxy phones

As promised, Samsung has begun the rollout of Gemini support to its Galaxy Buds 3 earbuds. It's a free update, and it's easy enough to get and install. But there's one caveat: you might need to install beta software on your phone.

Gemini doesn't run on the earbuds themselves – it's far too demanding to run natively – so it needs your phone or tablet to act as a messenger between your mouth and Gemini's ears.

As 9to5Google reports, at the moment the updates appear to be exclusively for phones running the One UI 8 atop Android 16. That's factory-fitted on the Galaxy Z Fold 7 and on the Galaxy Z Flip 7, but for the Samsung Galaxy S25 series One UI 8 is still a beta – ie. a pre-release version of the operating system, that while available (to a select group of users) is meant for testing.

(Image credit: Samsung)How to get Gemini on the Galaxy Buds 3 and Buds 3 Pro

If you have a Galaxy S25 Series or older, the stable version of the One UI 8 software isn't due until late September 2025. However, there is a beta version and from next week it will be available for more devices including the Galaxy S24 Series, the Galaxy Z Fold 6 and the Galaxy Z Flip 6. The beta will expand further next month to include the S23, Z Fold 5, Z Flip 5 and several A-series models.

Once you've got One UI 8 up and running, the next stage is to check for software updates for your earbuds… and then pray. I'm being dramatic there, but only slightly: the Gemini upgrade is reportedly "wonky" and results may differ from device to device.

If you've got the update, you should now see a new "Set up Google digital assistant" option at the top of the Voice Controls section in Samsung's app. This enables you to set up the Gemini integration with your smartphone or tablet.

You might also like

• Our in-depth Samsung Galaxy Buds 3 Pro review
• Oops, they did it again: Samsung Galaxy Buds 3 FE leak on Samsung's own site
• The best earbuds for every budget, all TechRadar tested

• Unidentified hackers access Air France and KLM through a third-party service provider
• The attackers stole names, contact details, and more
• Passport data was not compromised

Air France and KLM Royal Dutch Airlines have confirmed recently suffering cyberattacks in which both airlines lost sensitive customer data.

The companies, both owned by the same airline holding firm, sent out data breach notification letters to affected customers, and in a statement shared with Tweakers, KLM said the incident happened when threat actors broke into a third-party service provider.

“Unusual activity was detected on a third-party platform used by our contact centres, which led our IT security team, together with the third-party system involved, to swiftly implement corrective measures to put an end to the incident,” the company also told Cybernews.

Was it Scattered Spider?

We don’t know exactly how many people were affected by the breach, but the airlines transport more than 80 million people every year.

The information stolen in this attack include people’s full names, contact details, Flying Blue numbers and tier levels, and subject lines of service request emails.

Luckily, passport numbers, payment card details, passwords, or Flying Blue Miles (the airline’s loyalty program) balances were not stolen.

There was no word on the attackers, and no one claimed responsibility for the attack.

However, in late June 2025, the FBI warned Scattered Spider hackers were now increasing targeting airlines.

Scattered Spider works by impersonating company staff, and convincing support employees from the IT department that they lost access to their corporate accounts.

After gaining initial access, they map out the company, identify high-level individuals, and then repeat the process until they gain access to accounts through which they can steal data.

The hacking group struck Qantas in early July 2025, and Hawaiian Airlines in late June, and Russian Aeroflot, American GlobalX, and Canadian WestJet have all suffered similar incidents in recent months.

You might also like

• WestJet investigating possible cyberattack - make sure your data is safe
• Take a look at our guide to the best authenticator app
• We've rounded up the best password managers

• A video of an iPhone 17 Pro dummy unit in orange has leaked
• This shade is far more eye-catching than most Pro iPhone colors
• Viewers are very mixed on whether it's a good color or not

Leaks suggest that – along with understated black, white, and dark blue shades – the iPhone 17 Pro and iPhone 17 Pro Max might be sold in orange, and now a video has given us our best look yet at a dummy unit in that color.

Shared by leaker Majin Bu, you can see that this is quite a striking shade, and far more colorful than Apple’s Pro models tend to be. Whether it’s a good color is more debatable – Bu says it “looks so good”, but many of the replies to their post are a lot more negative.

@AnxiousHolly, for example, said “this looks hideous”, while @black0nder said “the worst color in iPhone history”. But @YSL_Laurentttt said “Let that be official and I’m getting the 17 Pro”, and @vernons said “if it really looks like that, it could be the winner in this year's upgrade.” So, this really does seem to be a love-it-or-hate-it shade.

iPhone 17 Pro orange looks so good pic.twitter.com/N9ehzP6ldnAugust 7, 2025

A good move

Whatever you think of this specific color choice, though, we’d say it’s a positive sign that Apple might offer such a shade, as it could mean the company is getting more comfortable with the idea that its Pro phones don’t always have to be sold in smart, understated shades.

We’ve long been calling for more colorful premium phones, so perhaps Apple is finally listening. But whether we’ll see more might depend on how well this orange shade sells.

And that’s assuming Apple really does launch an iPhone 17 Pro in this color. For now, we’d take this rumor with a pinch of salt, but we should find out soon, as the iPhone 17 series is likely to land in early September, with one leak pointing to September 9.

You might also like

• New iPhone 17 rumors tease pricing and colors – and there's good news and bad news
• The foldable iPhone could have much smaller screens than the Samsung Galaxy Z Fold 7
• The iPhone 17 Air might come in these 4 colors – and there's an interesting claim about its chipset too

Enterprise cloud demands have evolved beyond the early drivers of adoption. Now, more than ever, businesses must be prepared to scale at a moment’s notice. Meeting unexpected user demand while maintaining service levels is an ongoing challenge. At the same time, cloud computing workloads are becoming increasingly specialized, requiring optimized solutions across CPUs, DPUs, and AI accelerators to fully harness cloud-native architectures and frameworks.

As a result, cloud spending has become a critical operational priority. According to Flexera, nearly half of all workloads and data are now in the public cloud. In fact, 72% of IT decision-makers (ITDMs) are prioritizing cloud optimization as a key organizational initiative to drive cost savings.

These trends highlight the importance of right-sizing cloud solutions. Enterprises must ensure that their infrastructure is optimized for their specific business needs and workload requirements. The right cloud strategy delivers flexibility, security, performance, and cost efficiency, all of which are fundamental to maintaining a competitive edge.

Why the Time is Now

Cloud computing has long been the backbone of modern digital infrastructure, primarily built around general-purpose computing. However, the era of one-size-fits-all cloud solutions is rapidly fading in a business environment increasingly dominated by AI and high-performance computing (HPC) workloads. Legacy cloud solutions struggle to meet the computational intensity of deep learning models, preventing organizations from fully realizing the benefits of their investments.

At the same time, cloud-native architectures have become the standard, as businesses face mounting pressure to innovate, reduce time-to-market, and optimize costs. Without a cloud-optimized IT infrastructure, organizations risk losing key operational advantages—such as maximizing performance efficiency and minimizing security risks in a multi-cloud environment—ultimately negating the benefits of cloud-native adoption.

Moreover, running AI workloads at scale without an optimized cloud infrastructure leads to unnecessary energy consumption, increasing both operational costs and environmental impact. This inefficiency strains financial resources and undermines corporate sustainability goals, which are now under greater scrutiny from stakeholders who prioritize green initiatives.

Beyond performance gains, security is yet another critical consideration when selecting cloud-optimized hardware that often goes underappreciated. Cloud-optimized hardware often provide a strong suite of advanced security features, such as confidential computing. These technologies ensure that sensitive data remains encrypted while in use, reducing the risk of physical DIMM (Dual In-line Memory Module) attacks or virtual threats in hyperconverged infrastructure environments.

As the risks of data breaches continue to escalate – both in financial and reputational terms – organizations must recognize that leaving cloud environments unprotected is no longer an option. The rise of sophisticated cyber threats, from rogue hackers to nation-state-sponsored actors, makes enhanced cloud security a non-negotiable priority.

Your 2025 cloud optimization playbook

Furthermore, as industries push forward with the adoption and deployment of AI tools, IT leaders must ensure that their cloud infrastructure can support compute-intensive workloads while balancing cost, security, and efficiency considerations. While each organization's computing needs are unique, IT teams embarking on hardware modernization should consider the following:

Performance: Are your cloud instances equipped for the level of compute performance your business requires? Cloud infrastructure must support a range of workloads, from web front-end applications to in-memory analytics and heavy transactional processing.

Cost and efficiency: Can you reduce your cloud footprint by running the same workloads on fewer servers? Prioritizing instances with high compute density allows businesses to run more VMs or containers per server, achieving significant cost and energy efficiency benefits.

Security: Does your cloud instance provide the level of data protection you require? Confidential computing helps mitigate security risks by protecting data in use, reducing vulnerabilities in virtualized environments.

Ecosystem: Opting for processors powered by industry-standard x86 architecture simplifies cloud environments, making it easier to develop, maintain, and migrate applications with minimal disruption.

Unseen optimization, tangible impact

For IT decision-makers, understanding the cost implications of each 'unit of work' is crucial when selecting cloud instances. Traditional infrastructure forces enterprises to choose between overprovisioning resources – leading to unnecessary expenses – or under provisioning, which can cause performance bottlenecks. Cloud-optimized hardware changes this equation by enabling businesses to achieve more with fewer resources while maintaining high levels of performance, security, and efficiency.

As cloud technologies continue to evolve, enterprises that prioritize modernization will reap the benefits of seamless scalability, improved sustainability, and a resilient digital foundation for future innovation. In a world where agility is critical, cloud optimization is no longer a luxury—it is a business imperative for staying competitive in 2025 and beyond.

We list the best IT management tool.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

‘Your scientists were so preoccupied with whether they could, they didn’t stop to think if they should’. While this famous line from Jurassic Park is a poignant reminder of the dangers of unchecked ambition, it can also be applied to today’s rapidly evolving and fragmented AI landscape.

The mainstream availability of AI has compounded issues with shadow IT, as employees increasingly sidestep governance to deploy powerful, self-service AI tools. In this environment, many businesses are faced with how to manage the element of control when unmanaged AI systems start making critical business decisions based on fragmented, unverified data.

Like John Hammond’s ambitious yet doomed theme park, some organizations are now creating something powerful without fully understanding the risks or having proper containment measures in place.

It’s become a business imperative to find ways to ensure AI-ready data is trusted, compliant, and seamlessly connected. Here we explore the unintended consequences of AI-driven shadow IT and why businesses need a structured approach to data management to avoid costly mistakes.

The rise of AI-powered shadow IT

Shadow IT is not a new challenge, but AI takes it to a new level. With so many generative tools now readily available, employees can solve problems, generate content, or make recommendations at speed. This happens often without needing any technical expertise or approval.

This speed is both a blessing and a risk. In their enthusiasm to experiment and move fast, teams often pull data from disparate sources, bypassing enterprise-grade controls in favor of quick, isolated fixes. Over time, these short-term solutions accumulate, and organizations are left with a patchwork of systems, models and insights that don’t speak the same language.

The risk isn’t just that teams are duplicating efforts or misinterpreting data. Business-critical decisions affecting customers, supply chains, product development and strategic direction are increasingly being made based on unverified siloed information. When AI systems operating on flawed data foundations make recommendations that influence growth strategies, the potential for bias or error multiplies exponentially.

Unify and trust your data

The antidote to this growing risk is not to clamp down on experimentation. It’s to build the right data foundation, one that supports innovation while maintaining context and integrity.

This means giving employees access to high-quality, AI-ready data from across the business. It’s essential to build one harmonized layer that connects all business AI applications and ensures that everyone from developers to decision-makers can rely on a single source of truth.

This foundation keeps context intact, so the entire business can see where, how, when and why data was produced, building trust and accurately informing decisions. When data is unified, it also supports regulatory demands and keeps the business agile to future compliance requirements.

The cost of siloed data and duplicated spend

There’s a significant cost benefit to this too. When growth is the unanimous business goal, organizations cannot afford to hemorrhage spend on an inefficient IT landscape.

It’s estimated that organizations today spend up to 50% of their IT budgets on data and analytics, with a significant portion of that going on attempts to harmonize disconnected data sources. Yet, despite these efforts, many businesses still lack a continuous, unified data layer that brings these sources together in a coherent, usable way.

That’s not just inefficient, it’s a missed opportunity. In the age of AI, the power of data lies not just in how much you have, but in how well it’s connected. Without a shared foundation, AI models risk drawing the wrong conclusions or being trained on outdated information.

This in turn leads to additional budgetary pressures. Businesses need to confidently scale AI across functions, knowing insights are accurate, secure and compliant.

From raw data to business outcomes

To move from raw data to real business outcomes, organizations need more than just infrastructure. They need a strategic approach to data and analytics that supports decision-making at every level.

This means combining new technologies with existing business processes to create enriched, curated data products that deliver meaningful value. It means equipping users with advanced analytics, benchmarking tools and AI-powered insights applications that can both interpret the data and recommend actions.

This strategic approach helps limit the spread of shadow IT by reducing the need for employees to seek out unapproved tools or shortcuts. By aligning data initiatives with established governance frameworks and cultural values, organizations can ensure consistency, compliance and trust in the data being used. At the same time, it creates space for innovation and agility, enabling teams to move quickly and confidently within a well-defined structure.

When done right, the benefits are clear: smarter decisions, faster responses and better outcomes across the board.

Creating a culture of AI confidence

Ultimately, the question businesses need to ask is not whether they’re prepared to use AI, but whether they’re ready to do it responsibly and reliably.

Readiness starts with a strong data foundation, ensuring that information is accurate, accessible and well-governed. It means empowering teams with tools and guidance to innovate responsibly, creating a culture where experimentation with the right tools is encouraged.

The lesson from Jurassic Park was not that innovation is dangerous. It’s that innovation without structure, without guardrails and without consideration of the bigger picture can quickly spiral out of control.

We list the best data visualization tool and the best business intelligence platform.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

• OpenAI's GPT-5 debut included the reveal of a new integration that allows ChatGPT Pro users to link their Gmail, Google Calendar, and Google Contacts directly to the chatbot
• ChatGPT can summarize your day, create a schedule, and flag important emails
• Though an opt-in feature requiring user confirmation, privacy issues are likely to be raised

The spectacle of GPT-5 may have overshadowed some of OpenAI's other news during its livestream on Thursday, but the demonstration of ChatGPT's new ability to directly peruse and analyze a user's Google data caught my eye as a major moment in ChatGPT's development and OpenAI's battle for supremacy among AI chatbots.

The new feature enables ChatGPT users to connect the AI chatbot to their Gmail, Google Contacts, and Google Calendar data. The demo showed ChatGPT responding to a request to see a schedule of the following day by going through the user's calendar and email inbox, then rapidly compiling a complete and complex schedule, including important unread emails to respond to.

This may not sound like a breakthrough to anyone who’s already overwhelmed with their digital list of things to do, but ChatGPT sorting that information and putting it in front of you might actually lower your stress, at least judging by the demo.

It's easy to imagine an AI outline for your day, or a nudge to handle that still-unread message, reducing the mental workload by cutting out the tedious sorting and linking of scheduled events to relevant emails. You might say, “What’s on my plate today?” and see your calendar paired with that flagged email. That would mean no more toggling between Gmail and your calendar, squinting at what's urgent.

Further, the data could help ChatGPT learn more about you and your needs by reading the meeting invites you've sent, deadlines you've barely hit, and RSVPs you sent the second you got the invite. For now, this option is only available to ChatGPT Pro users, though OpenAI promised it would become more widely available soon.

ChatGPT won't sneak a peek at your messages

That said, the idea of handing over Gmail and Calendar data to ChatGPT might raise an eyebrow or two for good reason. Gmail could hold confirmation of doctor appointments and secret romantic rendezvous plans.

Don't worry about inadvertently sharing those details with ChatGPT, though. You'll need to opt in to link your accounts to ChatGPT and confirm actions before they occur, which will prevent any emails from being sent automatically.

Then again, there are plenty of smart scheduling bots and email add-ons that automatically pull event details or remind you about missed invites. But ChatGPT’s integration adds actual conversation to the mix. You don’t forward an email or set up complex rules of how the automated system should respond to certain family members. You just type in regular language, and it will act more like a human secretary.

Assuming you're okay with the concept, you can see how those who use Google and ChatGPT might value linking the two. Especially if you're not a fan of Google Gemini for one reason or another. You might long to have an AI chatbot connected to your Google account, but simply prefer ChatGPT to Gemini. OpenAI wants to give you that option.

If Google doesn't have an exclusive claim to linking your email and calendar to an AI chatbot, then OpenAI can hope to win out in other areas where it may feel it has the advantage, like the power of GPT-5. You just have to be okay with letting ChatGPT see which weddings you'll be attending in the next few months.

You might also like

• OpenAI's first AI Agent is here, and Operator can make a dinner reservation and complete other tasks on the web for you
• There’s a new AI agent ready to browse the web and fill in forms without the need to touch your mouse
• I tried this new AI agent that takes control of your mouse and keyboard to help get tasks done, and I can’t believe how good it is

• Hard drives sold as new contained decade-old internals from previously used Seagate and Western Digital units
• SMART data was manipulated to hide prior usage, masking serious mechanical and read error issues
• Attingo analysis revealed leftover user data, proving zeroing only touched the start of storage sectors

A recent case involving UnionSine-branded external hard drives has raised serious questions about the integrity of some products sold through Amazon’s marketplace.

An investigation by data recovery company Attingo uncovered supposedly "new" hard drives contained 2.5" HDDs from major brands like Seagate and Western Digital manufactured over a decade ago.

These findings contradict the product labelling, which lists a manufacturing date from spring 2025.

Used drives masquerading as new

The drives were sold under the identifier HD2510 during Amazon’s Tech Week promotional period, and were assumed by buyers to be new portable HDDs suitable for regular backup and storage use.

Attingo’s teardown of the drives revealed not just their age, but also signs of previous use, with the evidence extending beyond the hardware identifiers.

The company said some of the HDDs still contained fragments of user data, with only the beginning of the data fields zeroed out.

This practice, commonly used to simulate full data erasure, failed to eliminate all traces of prior use.

One recovered XML file even contained a timestamp from May 2024, pointing to very recent activity inconsistent with a supposedly unused device.

Some of the data traces were linked to TV recording systems, suggesting that the drives had once been in operational environments.

Attingo also noted inconsistencies in SMART (Self-Monitoring, Analysis and Reporting Technology) data, which is used by operating systems to assess hard drive health.

Although the displayed operating hours suggested zero usage, other key metrics, such as the read error rate, told a different story.

This kind of manipulation has been seen before in fraud cases involving repurposed HDDs, especially those sold through gray market channels.

The presence of numerous read errors and mechanical anomalies conflicted with the SMART status, indicating a fresh drive.

The fact that this issue emerged from a well-known retail platform, rather than a niche marketplace or obscure online store, is of particular concern.

Attingo CEO Markus Häfele expressed alarm at the findings, saying the situation was not only misleading to consumers but also potentially dangerous from a data protection standpoint.

"It's unfortunately well known that used hardware is occasionally declared as new on platforms like eBay or in direct imports from Asian retailers – but the fact that this is happening systematically as part of a widely advertised Amazon campaign is truly scandalous," Häfele said.

Though the incident has sparked renewed concern about the reliability of NAS HDDs and external storage devices offered through third-party sellers online, Amazon has yet to respond to the allegations.

Via Computerbase

You might also like

• These are the fastest SSDs you buy right now
• Take a look at some of the best external hard drives
• Watch how Chinese are converting thousands of Geforce RTX 5090 cards into AI accelerators and there's nothing Nvidia can do to stop it

OpenAI’s much‑anticipated livestream reveal of GPT‑5 crammed a lot into about an hour and a half of announcements and demonstrations. CEO Sam Altman didn't show any of the fear he claimed to feel about the new AI model, just a lot of pride at what he and his team had accomplished.

Much of the presentation showed off GPT-5's technical milestones and how they translate into powerful new and upgraded AI features for users. Altman also had some lighter fare to unveil for ChatGPT, offering more customization options and ways for users to link their existing online footprint to ChatGPT.

Here are the five most notable pieces to emerge from GPT-5's debut.

GPT-5 arrives

(Image credit: OpenAI)

GPT-5 is the next iteration of OpenAI's models, bigger and more powerful, but not dissimilar in its basic form. Even so, GPT-5 is big and complex enough to reach a new level in how it seems to reason.

Essentially, you no longer have to spoon-feed it context or restate complex prompts three times, or at least not nearly as often. Multifaceted questions like how changing interest rates might affect Gen Z homeownership trends in mixed markets might take several prompts refined multiple times to provide the answers you seek with earlier ChatGPT models, but GPT‑5 can unpack the whole thing.

Based on the demonstrations, GPT‑5 seems to parse each part separately and stitch it together. And it flags when there's a gap in its knowledge, which is far better than confidently hallucinating. It applies that way of thinking to how it interacts with users, too. While obviously not 'thinking', it does appear to read between the lines well enough to reflect a user's mood and even adjust its response to an expressed emotion.

The model is supposed to be particularly good with math and coding software. Good enough to handle the increasingly popular pursuit of 'vibe coding,' where you simply describe a feeling or a mood of a piece of software, and the AI produces the HTML, CSS, and JavaScript to match your design vision.

GPT-5 sizes

(Image credit: OpenAI)

GPT-5 comes in many sizes in addition to the standard version. There's also the smaller gpt-5-mini, and an even leaner gpt-5-nano, which lives solely in the API. The big news is that free ChatGPT users now get access to both GPT-5 and mini, while ChatGPT Plus subscribers enjoy higher usage limits across the board.

If you’re a Pro user paying $200 a month, you’ll now get unlimited GPT-5 access, along with access to the more powerful gpt-5-pro model and gpt-5-thinking. These both take longer to provide answers, but come back with deeper, more thoughtful responses.

There is no need to pick and choose yourself, either, though. ChatGPT now picks the right model automatically based on what you're asking and what plan you’re on.

ChatGPT custom personalities and colors

ChatGPT has a default, pleasantly bland personality, but GPT‑5 is advanced enough to offer more variety in the tone and style of the AI chatbot. If you don't want the usual neutral mode, you can choose “Cynic” for sarcasm with your answer, “Listener” if you’re venting and just need it to echo back understanding, "Nerd" for a side of geeky trivia, and "Robot" for the purely mechanical response.

These personalities don’t undercut the answers you get, but they do flavor the response. Therefore, you might get dry wit with productivity tips from the “Cynic” tone or gentle encouragement in your goals from “Listener.”

Additionally, the chatbot's appearance can now be altered with the new color themes. If you're a paid subscriber, you'll soon be able to adjust the look of ChatGPT instead of switching between the usual black or white.

Voice Mode

(Image credit: Future)

ChatGPT's Voice Mode is getting an audio glow-up of its own. OpenAI is rolling out a much-improved version that not only works with custom GPTs but also adapts its tone and speech style based on your instructions and overall vibe.

You can ask it to be snappier, slower, warmer, or whatever else you want. For ChatGPT Plus users, voice responses are now nearly unlimited. Free users still get access, too, with a few hours a day to chat hands-free.

To streamline things, the old Standard Voice Mode is being phased out entirely within 30 days. After that, everyone will be on the same upgraded experience.

Google connections for ChatGPT

(Image credit: Future)

Next week, ChatGPT Pro users will be able to hook up their Gmail, Google Calendar, and Google Contacts directly to ChatGPT. That means no more switching tabs to check if you're free next Tuesday or digging through threads to find that one email you definitely forgot to reply to.

Once connected, ChatGPT will pull in what it needs to help respond to your queries. OpenAI assured users that it will only pull in the minimum needed and only when it’s helpful.

You don’t need to say “check my calendar” or “pull that contact.” The AI will do so based on whether you request something that requires it, like scheduling a meeting. It will pick a time that works for you and write the email on your behalf. Other subscription tiers are scheduled to get access to the connections in the near future, so this won't be limited to Pro forever.

All of these upgrades leveraging GPT-5 point to OpenAI's bigger plans to make its AI models an intimate part of your life, not just a tool you occasionally turn to and feel annoyed about having to carefully parse answers. Smarter reasoning means less cleanup for the user. Vibe coding shifts AI from merely aping code to interpreting your intended use with software. The personalities and colors make the AI feel like it's more unique to you, not just a one-size-fits-all tool, especially with the more realistic voice and access to your email account and calendar

GPT‑5 mimics awareness better than any of its predecessors. That means it could blend into our routines and become as second nature to use as our smartphones. Or at least, that's what OpenAI and its investors likely hope to see happen.

You might also like

• OpenAI's first AI Agent is here, and Operator can make a dinner reservation and complete other tasks on the web for you
• There’s a new AI agent ready to browse the web and fill in forms without the need to touch your mouse
• I tried this new AI agent that takes control of your mouse and keyboard to help get tasks done, and I can’t believe how good it is