TechRadar News

Marvel's Thunderbolts* movie has stormed into theaters worldwide, and, whether or not you plan on seeing it on a big screen near you, I suspect you'll want to know when it'll come to Disney+.

Well, the short answer is that the second Marvel film of 2025 doesn't have a streaming release date yet. That won't stop me speculating on when it might land on Disney's primary streaming platform, though, which I'll do shortly.

That's not all. Below, I'll also provide more information on what you need to know about Thunderbolts*, including its runtime and end-credits scenes. Before you proceed, though, make sure you read my Thunderbolts* review to see if it's worth watching in your local cinema and/or at home.

Does Thunderbolts* have a Disney+ release date yet?

Staring into the *ahem* void while you wait for Thunderbolts* to land on Disney+ like... (Image credit: Marvel Studios)

No. As I mentioned, Thunderbolts* doesn't have a confirmed Disney+ launch date yet, and it'll be a while before the Marvel Phase 5 movie makes its debut on one of the world's best streaming services.

On average, it takes three months for Marvel Cinematic Universe (MCU) films to be added to the Disney+ movie library. Indeed, at the time of publication, the most recent MCU flick – Captain America: Brave New World, which flew into theaters on February 14 – still isn't out on the service.

With Thunderbolts* only launching in multiplexes and local theaters globally in early May, it'll probably be August before the Florence Pugh-fronted team-up film lands on Disney+. For now, your only option is to watch it on the big screen.

As for my prediction of when it'll be available to enjoy in the comfort of your own home, I think Thunderbolts* will make its Disney+ bow on Wednesday, August 6.

That's a whole week after Ironheart, the final MCU Phase 5 project, will have ended. And, with new Disney+ movies usually releasing on the platform midweek, the aforementioned date seems like a solid guess to me.

What is the runtime of Thunderbolts*?

Thunderbolts* will take around two hours to watch (Image credit: Marvel Studios)

Thunderbolts* has a runtime of two hours and six minutes.

That includes its end credits crawl, which you should definitely sit through to fully appreciate the sheer amount of people who worked on one of the best Marvel movies of the Phase 5 era.

Does Thunderbolts* have a mid- and/or post-credits scene?

Metaphorically speaking, we're always tied to our seats until every Marvel movie's end-credits sequence has ended (Image credit: Marvel Studios)

Yes! Thunderbolts* has a mid-credits scene and a post-credits stinger. Where the MCU's future is concerned, the latter is far more important than the former but, in my view, they're both worth watching for different reasons.

Want more details on what is shown? My Thunderbolts* ending explained piece is packed with more information about those scenes, as well as plenty of speculation about what the film's end-credits scene means for the next two MCU movies – those being The Fantastic Four: First Steps and Avengers: Doomsday.

You might also like

• Here's how you can watch the Marvel movies in order
• Find out more about every Marvel Phase 6 project
• Get the lowdown on Spider-Man: Brand New Day

Marvel Unlimited monthly subscription: was $9.99 per month now $4.99 at Marvel
The ultimate digital comics subscription for Marvel fans is offering new and returning customers 50% off their first month, which is much cheaper than a streaming subscription to watch all the best superhero movies. To gain access to a library of more than 30,000 comics, use the code 'SPIDEY50' at checkout. You're able to use the Marvel Unlimited app on all iOS and Android devices, including the web, too. But, be quick – the deal expires on May 5!View Deal

• The highly anticipated Grand Theft Auto 6 has been delayed
• But we do finally have a release date
• It will launch on May 26, 2026

Developer Rockstar Games has finally given us an official release date for Grand Theft Auto 6, though simultaneously revealed that the game has been delayed.

As confirmed in a recent press release, the highly anticipated open-world crime game will now release on "May 26, 2026" - a year after its previously announced 2025 release window.

"Hi everyone," the press release begins. "Grand Theft Auto 6 is now set to release on May 26, 2026."

It goes on to apologize for the delay: "We are very sorry that this is later than you expected. The interest and excitement surrounding a new Grand Theft Auto has been truly humbling for our entire team. We want to thank you for your support and your patience as we work to finish the game."

It also provides a reason for the decision, stating that "With every game we have released, the goal has always been to try and exceed your expectations, and Grand Theft Auto 6 is no exception. We hope you understand that we need this extra time to deliver at the level of quality you expect and deserve."

It concludes by saying that more information will be shared "soon".

The first trailer for Grand Theft Auto 6 was released back in December 4, 2023, and has now been viewed more than 251 million times on YouTube.

The game will take place in the fictionalized US state of Leonida, home of the the iconic Vice City featured in previous games. It follows new protagonist Lucia and her partner as they go on the run from the law.

Industry analysts predict that the game's release will be "one of the biggest entertainment launches in history" and that it could potentially generate more than $1 billion in pre-order sales alone.

It will be available on PlayStation 5, Xbox Series X, and Xbox Series S.

You might also like...

• Gost of Yotei pre-orders live: buy now in the UK, retailers to watch in the US, and where to find the Collector's Edition
• Fortnite Chapter 6 Season 3 launch day: our live coverage now that Fortnite Galactic Battle is online
• Forget the Steam Deck, Cyberpunk 2077 on Nintendo Switch 2 is 'the more Cyberpunk way' to play, per CD Projekt Red

• The NotebookLM mobile app listings are now live
• The research app is coming very soon to Android and iOS
• Expect to hear more news at Google I/O 2025 on May 20

Having hinted at the imminent arrival of a NotebookLM app last month, Google has now put the AI research app live on the Google Play Store and Apple App Store for mobile devices. You can't download it yet, but you can pre-order it – and it gives us a good early glimpse of what the promising app looks like.

'Pre-ordering' the app means it will automatically download to your phone when it's launched in full, which according to the Apple App Store listing (via 9to5Google) is May 20 – the same day as Google I/O 2025 starts, where we'll no doubt hear more about this app.

(Image credit: Google)

If you've never come across NotebookLM before, it's an AI-powered resource tool that can collate information across documents, webpages, and more – you can get smart summaries, ask questions about the data, and even produce fake podcasts (which you can interrupt, if you want to).

It looks as though the app is going to include the podcast (or "audio overview") creator, based on the screenshots in these listings. We can also see screens for adding sources, and for the conversation panel where you can chat about the sources with the AI.

An alternative to Gemini

NotebookLM can produce audio podcasts for you (Image credit: Google/Apple)

Google will no doubt give us the full story about this app in a few weeks, but almost all of the web app features seem to have been carried over – including the ability to add notebooks and browse through the notebooks you've created on the web.

NotebookLM is a helpful alternative to Google Gemini for those times when you need AI's help with a specific set of material you've collected – perhaps a science research project or reports on a current industry trend.

Based on my experience with the tool, it's usually more straight-faced and serious than Gemini, and includes citations with all its responses – so you can check if it's made any mistakes. You get higher usage limits and more customization options in NotebookLM if you're subscribed to a Google One AI Premium plan.

It looks as though there's going to be a wealth of AI news coming out of Google I/O 2025 from May 20 – including, perhaps, an update on the AI Mode for search – and we will of course bring you all the announcements as they're made.

You might also like

• Google's AI Mode can explain what you're seeing even if you can't
• Testing out Mind Maps in Google's NotebookLM
• NotebookLM vs a 218-page string theory research paper

• A third British retailer has been hit with a cyberattack
• The M&S incident is likely a ransomware attack from ScatteredSpider
• Retailers are at risk due to high downtime costs

Luxury department store Harrods has become the third British retailer to be hit by a cyberattack in a matter of days, with the firm restricting internet access at its sites following an attempt to gain access to its systems.

The incident closely follows the confirmation that an earlier attack targeting Marks & Spencer, causing widespread outages in stores and the retailer’s online platform, is allegedly the work of ScatteredSpider hackers.

The third incident, a cyberattack on retail giant Co-Op, forced the supermarket to take down parts of its IT systems and take proactive measures to defend against the attackers.

There are certainly similarities between the reports of the incidents, and in such quick succession it seems likely there may be a link between them, but the extent of the attacks is yet to be seen - here’s what we know so far.

No known link

Despite the timings and the similarities, there’s no official link between the incidents, although it’s likely that all three incidents are ongoing, so more information may be revealed as the attacks unfold.

In the meantime, security teams should be ultra vigilant, and should deploy the best endpoint protection software to keep their organisation safe.

SonicWall’s Executive VP of EMEA, Spencer Starkey explains ransomware, “holds victim organizations' business operations hostage, which uniquely impacts retailers and other organizations that provide daily, direct services to their customers”.

As we’ve seen in the continued disruption of the M&S attack, these incidents can have a devastating effect, with the supermarket’s share price falling 7%, causing millions in lost sales and downtime.

Starkey explains these attacks are particularly harmful to retailers as they directly affect revenue generation, which provides additional leverage for the attackers, and warns that companies should be operating on the presumption that they will inevitably be targeted, so should develop a comprehensive incident response plan.

Not if, but when

Concerns are rising for businesses across the country, and retail attacks have risen sharply since the development of AI tools designed to help cybercriminals send out more frequent and sophisticated attacks.

The head of the National Cyber Security Centre (NCSC), Dr Richard Horne, has warned that these attacks should serve as a “wake-up call” for all organizations, and urges security leaders to ensure they have “appropriate measures in place to help prevent attacks and respond and recover effectively.”

Ex-NSA cyber chief, Cody Barrow, also warns generative AI is “accelerating the threat landscape,” and that sophisticated attacks like social engineering and adaptive malware campaigns are now available to even low-skilled attackers.

In particular, criminals are able to send out an incredible number of social engineering attacks, that are designed specifically to trick victims into giving attackers access to their networks.

Security best practices recommend using the best antivirus software, best malware removal software, and training all staff to recognise phishing attacks.

These attacks outline just how vulnerable the sector is, and Barrow argues this is “due to the volume of customer data and the high cost of operational downtime”. Although, the reality is that almost all sectors are facing more threats than ever.

“Retailers must assume they are targets. Rehearsed incident response plans, clear communication protocols, and multi-factor authentication for administrative access are now essential. Endpoint detection and response tools should be standard, not optional. For consumers, vigilance is crucial: update passwords, monitor financial activity, and watch for scams exploiting recent breaches.”

You might also like

• Take a look at our picks for the best malware removal software around
• Check out our choice for best antivirus software
• M&S checkout chaos persists as cyberattack fallout continues

• Amazon’s Alexa+ AI has just hit 100,000 users in the US
• Yet the firm’s CEO has warned agentic technology is still “primitive”
• Alexa+ is rolling out in waves, but is still missing some key features

Like many other Big Tech firms, Amazon jumped on the artificial intelligence (AI) bandwagon by launching its own AI assistant named Alexa+ in February. Now, the company has announced that the tool has been rolled out to 100,000 users.

Right now, Alexa+ focuses on providing a conversational experience with an AI chatbot, although Amazon aims to imbue it with “agentic” abilities in the future. That would enable it to work as an agent on your behalf, interacting with third-party apps in order to carry out the tasks you ask of it.

That little tidbit came from Amazon CEO Andy Jassy during the company’s recent earnings call. There, Jassy explained that “Our intention is for agents to perform wide-ranging complex, multi-step tasks,” and that “there haven't been action-oriented agents like this until Alexa+.”

Yet the most recent update to Alexa+ is missing some of the key features that Amazon demonstrated in February, as pointed out by The Washington Post. It wouldn’t be the first AI to arrive without promised abilities – Apple’s AI-powered Siri has faced numerous delays and disruptions – and it shows how difficult it can be to pull ahead in the ongoing AI arms race.

Even Jassy admitted that Alexa+ was facing challenges. In general, agentic AI tools are still “primitive,” the Amazon CEO noted, with Amazon’s own Nova Act AI model aiming to improve their accuracy from a mere 30% to 60%. Eventually, the goal is to achieve over 90% accuracy, but clearly Alexa+ isn’t there yet.

Limited availability

(Image credit: Amazon)

Amazon has previously stated that it aims to launch Alexa+ in waves, so we should expect to see availability increase in the US over the coming months. There's no news about an international rollout yet.

Of those users who have tried it out so far, few have shared their thoughts or given insights into how well it works. That changed recently, though, when user Pretty2be took to Reddit to answer questions on the AI chatbot.

During the session, Pretty2be stated that Alexa+ is good at interpreting natural speech and is able to avoid interrupting you. Other positive feedback included its speech synthesis capabilities, although the user criticized its tendency to push you towards Amazon Music, even when you try to use Spotify.

We don’t know when Alexa+ will roll out to more users, but it’s clear that there’s still work to be done, judging by both Pretty2be’s feedback and Amazon CEO Jassy’s comments. Hopefully the missing features will arrive before the next 100,000 users sign up.

You might also like

• Alexa+ explained: 9 things you need to know about Amazon's new AI-powered assistant
• Alexa+ early adopter gives their thoughts: "Old Alexa was hot garbage but this new one was a pleasant surprise"
• I wasn’t a fan of the new Echo Show 15 and 21, but Alexa+ has changed my mind

• New Microsoft accounts will use passkeys by default, company reveals
• Existing users can also delete their passwords
• The login experience has an updated interface

Microsoft will be making all new accounts passwordless by default in a bid to push users towards more secure authentication methods that revolve around passkeys, push notifications and security keys.

"Microsoft and dozens of other organizations have taken the Passkey Pledge to work toward increasing the implementation and adoption of passkeys over the coming year," Microsoft explained in a blog post.

Naturally, existing users can also opt in to delete their traditional passwords via account settings in favor of more secure options.

Microsoft goes passwordless by default

Claiming that it "helped lay the foundation for an entirely new era of authentication," Microsoft's blog post takes us back ten years to the launch of Microsoft Hello, which allowed users to log into machines with secure biometrics.

"To keep your digital life safe, you need a way to sign into any account without a password," Joy Chik and Vasu Jakkal explained.

But why is it important to move away from less secure simple passwords?

Apparently, Microsoft observed more than 7,000 password attacks per second last year, marking a twofold year-over-year increase.

However, while passkeys make for a more secure account, the company noted that old-style accounts with basic passwords can expect to see even more attack attempts. All the more reason to go passwordless across all accounts – not just your Microsoft account.

They're also speedier, offering a log-in experience eight times quicker compared with the password and MFA combination.

Passkeys for consumer apps like Xbox and Copilot already became available in 2024, and now, more than a million are being registered each day.

Other recent announcements include an updated login user experience, with a simpler style introduced in March 2025 prioritizing passworldess methods, all of which just in time for yesterday's World Password Day.

From now, "new users will have several passwordless options for signing into their account and they’ll never need to enroll a password."

If Microsoft's decision to go passwordless has got you inspired to transition to passkeys across more services, check out the entire list of participators across the FIDO Alliance's Passkey Pledge, including Amazon, Apple, Dashlane, Google and tens more.

You might also like

• World Password Day top deals: Top offers from trusted password managers
• Check out the best password managers and best password generators
• We've listed the best authenticator apps

• A complete specs list for the Samsung Galaxy S25 Edge has leaked
• This largely lines up with what we've heard before, but includes new details like the presence of wireless charging
• Pricing has also leaked, and as with earlier leaks this puts the S25 Edge between the S25 Plus and Ultra

We’re likely less than two weeks away from the launch of the Samsung Galaxy S25 Edge, with May 13 rumored to be the big day. But ahead of that a major leak has tipped almost every spec and feature.

Reputable leaker Roland Quandt writing for WinFuture has claimed that the Samsung Galaxy S25 Edge will be just 5.85mm thick, weigh 163g, and have a titanium frame, with Gorilla Glass Victus 2 on the back and Gorilla Glass Ceramic 2 – which is said to be particularly durable – on the front.

Speaking of the front, that apparently holds a 6.7-inch 1440 x 3120 AMOLED screen with a 120Hz refresh rate, and there’s said to be an ultrasonic under-display fingerprint sensor.

Quandt also claims the Samsung Galaxy S25 Edge has a Snapdragon 8 Elite chipset, 12GB of RAM, and a choice of 256GB or 512GB of storage.

For the cameras, there’s apparently a 200MP f/1.7 main one with optical image stabilization (OIS) and a 12MP f/2.2 ultra-wide with a 120-degree field of view and no OIS. It also apparently has a 12MP f/2.2 front-facing camera.

Rounding out the specs, Quandt claims the Samsung Galaxy S25 Edge has a 3,900mAh battery, wireless charging, and an IP68 rating for water and dust resistance.

Image 1 of 3

(Image credit: WinFuture / Roland Quandt)Image 2 of 3

(Image credit: WinFuture / Roland Quandt)Image 3 of 3

(Image credit: WinFuture / Roland Quandt) Similar to what we've heard before

This is mostly stuff we’ve heard before, but there’s a little more detail here than in previous leaks, such as the mention of wireless charging, the exact build materials, and the details beyond megapixels in the camera specs.

Since we have largely heard this before and since it comes from a reputable source it may well be accurate, in which case – as expected – the main compromises Samsung has seemingly made to get the phone so thin include the absence of a telephoto camera, and the low battery capacity.

Despite these compromises, the Samsung Galaxy S25 Edge likely won’t be cheap, with Quandt claiming it will start at €1,249 (roughly $1,415 / £1,065 / AU$2,205) in Germany, for which you’ll get 256GB of storage, and saying it will rise to €1,369 (around $1,550 / £1,165 / AU$2,415) for 512GB.

The price conversions we’ve included are unlikely to be accurate, but notably that would put it between the price of the Samsung Galaxy S25 Plus and the Samsung Galaxy S25 Ultra in Germany, which is where we expect it to land globally based on other rumors.

You’ll supposedly be able to choose between Titanium Jet Black, Titanium Icy Blue, and Titanium Silver shades, as you can see in the renders above, and apparently it will ship in Germany around the end of May, which makes sense if it’s announced on May 13.

You might also like

• We’re now almost certain the Samsung Galaxy S25 Edge will land on May 13 – and we have a clearer idea of its cameras
• Samsung Galaxy S25 Edge: latest news, rumors, and everything we know so far
• Samsung Galaxy Z Flip 7: latest news, rumors, and everything we’ve heard so far

• LinkedIn's Wire Program becomes BrandLink
• It focuses on delivering advertisement videos from creators and publishers
• AI, leadership and business advice will be the first topics to be covered

LinkedIn has announced some changes to its ad revenues with the rebrand of Wire Program to LinkedIn BrandLink, which it says is designed to "[help] brands reach their audiences in new ways with video from top creators and publishers."

Product Management VP Lindsey Edwards explained, "BrandLink delivers more relevant video content to members while also allowing advertisers to align with editorial content from trusted publishers and, now, some of the world’s top creator voices."

From launch, content themes will include artificial intelligence, leadership, small businesses, innovation and entrepreneurship.

LinkedIn launches updated BrandLink program

The Microsoft-owned social networking platform says advertisers see 130% higher video completion rate and 23% higher views compared with standard video ads when using BrandLink, adding that members who view those ads are 18% more likely to become a lead for their company.

The news comes after a series of LinkedIn video-related announcements over the past year or so, including full-screen vertical video displays, a refreshed carousel and an updated UI to make it easier for audiences to engage with brands.

Edwards referred to video as "one of the fastest growing formats on LinkedIn," and in February 2025, we learned the medium was growing twice as fast as any other post format.

Some of the early creators to join the program include B2B figures such as Steven Bartlett, Bernard Marr, Allie K. Miller, Rebecca Minkoff, Candace Nelson, Guy Raz, Gary Vaynerchuk, and Shelley Zalis.

The latest announcement highlights how 91% of B2B marketers believe that video will help them accelerate advertising ROI more than any other format.

With BrandLink, advertisers can target specific locations and languages, with LinkedIn boasting it "supports a growing list of global publishers who produce region-specific content."

The tool is available to use now on desktop and mobile applications.

You might also like

• LinkedIn now wants to be more like...TikTok?
• We've listed the best social media management tools
• Check out our roundup of the best free video editing software

• AMD is reportedly prioritizing the Radeon RX 9060 XT 16GB GPU for retailers
• It corroborates rumors that it may cancel or discontinue the 8GB GPU
• If it's true, it could be a sign that Team Red won't make 8GB GPUs going forward

AMD's Radeon RX 9060 XT GPU is on the horizon (it's projected to launch in May), with 16GB and 8GB variants. However, a new report suggests one of them may not be readily available for consumers at launch.

As reported by Wccftech, AMD could be set to prioritize the Radeon RX 9060 XT 16GB variant over the 8GB counterpart in terms of availability for retailers at launch, according to Benchlife. It comes after the rumor that hinted at the 8GB model being cancelled or discontinued, and this new rumor certainly suggests that something is going on with the 8GB model.

Essentially, this means consumers may not see the 9060 XT 8GB available on shelves or online retailers apart from rare occasions. If this is true, it could be a huge indication of AMD's future stance on low VRAM capacity GPUs.

Nvidia's RTX 5060 Ti recently launched with 16GB and 8GB variants, with the latter receiving a lot of flak; consumers have been expressing displeasure with 8GB graphics cards for a while, and Benchlife claims that this is AMD's reasoning behind the potential case of prioritizing the 16GB GPU.

The Radeon RX 9060 XT series is expected to be unveiled at Computex, with a potential May 21 launch - but apparently, it won't actually go on sale until June (considering AMD's Q2 2025 estimation). Regardless, this puts the GPUs in direct competition with the RTX 5060, expected to launch on May 19. So, we could be in for another fierce battle between these two tech giants.

(Image credit: Future / John Loeffler) Please, let this be the end of 8GB GPUs...

If this rumor is legitimate, it suggests to me that Team Red is well aware that PC gamers are increasingly of the opinion that 8GB of video memory is simply not enough for modern games at high resolutions. If there's anything good to come from this, it's that we could soon see the end of 8GB graphics cards.

Triple-A games aren't easy going regarding VRAM usage, and while 8GB of VRAM isn't completely useless, I'd say it's a limitation for those looking for better performance. Gaming at 1080p is still the most popular resolution (according to Steam's hardware survey), but more demanding games with higher graphics settings enabled, like Cyberpunk 2077 or Black Myth Wukong, will end up choking 8GB GPUs.

It's also worth mentioning the cost of GPUs in this current market with inflated pricing; a prime example is the XFX Speedster Radeon RX 7900 XTX, sold at $1,129.99 on Best Buy (it's a three-year-old GPU), despite its $999 launch price. Inflation like this makes other low-end graphics cards seem much less appealing, as they no longer have low prices that justify the compromises made to performance. If you're going to be spending that kind of money on a GPU, you'd be better off looking at more powerful cards.

At the very least, I feel that modern graphics cards should come with 12GB VRAM at a minimum, because anything less just won't suffice in 2025 and onwards...

You may also like...

• Tariffs could further disrupt the stability of the PC hardware market - and this new study suggests it's best to buy right now
• Two new laptops with Nvidia's RTX 5060 are coming, and leaked performance makes it seem like 8GB GDDR7 might actually be enough
• Trump's tariffs may be paused, but that won't stop GPUs being overpriced and expensive

• Veeam study find budget increases for cybersecurity aren't enough; businesses need smarter resilience strategies
• Ransomware payments are dropping, but attackers are shifting tactics faster than businesses can adapt
• Declining attack rates don't mean ransomware is beaten

New research has revealed the scale of recent ransomware evolvution, warning it remains a dominant threat to organizations worldwide.

A Veeam study, which gathered insights from 1,300 CISOs, IT leaders, and security professionals across the Americas, Europe, and Australia, found nearly three-quarters of businesses were impacted by ransomware over the past year.

Cybersecurity measures seem to be having some effect, with businesses facing ransomware incidents dropping slightly from 75% to 69% - and ransomware payments are also decreasing, as in 2024, 36% of affected businesses chose not to pay, and 60% of those who did paid less than half of the demanded ransom.

Declining attack rates are no reason for complacency

"Organizations are improving their defenses against cyber-attacks, yet 7 out of 10 still experienced an attack in the past year. And of those attacked, only 10% recovered more than 90% of their data, while 57% recovered less than 50%," said Anand Eswaran, CEO of Veeam.

Still, businesses cannot afford to relax. Major law enforcement actions against groups like LockBit and BlackCat have disrupted large-scale operations, but this has inadvertently led to a surge in smaller, independent attackers.

Companies still need to adopt proactive defenses and use anti-ransomware tools alongside the best endpoint protection solutions.

"As the nature and timing of attacks evolve, it is essential for every organization to transition from reactive security measures to proactive data resilience strategies. By adopting a proactive security approach, investing in strong recovery solutions, and fostering collaboration across departments, organizations can significantly reduce the impact of ransomware attacks," Eswaran added.

The rise in data exfiltration-only attacks, where hackers bypass encryption to steal private data directly, is another alarming trend. In this environment, relying solely on antivirus software may not be enough.

Although 69% of companies say they are prepared for an attack, this confidence drops sharply after an incident. Only 44% have tested backup plans, and just 30% have a formal chain of command in place.

Regular training, cross-team cooperation, and investment in strong recovery frameworks are necessary. Although budgets for cybersecurity and recovery have increased, they remain insufficient.

As a result, Veeam advises companies to implement thorough procedures such as the 3-2-1-1-0 data rule, which ensures multiple immutable backups free of malware before restoration.

You might also like

• Downloaded something dodgy? Follow our best malware removal advice
• Nail the basics with the best firewalls available now
• Most popular chip designer in the world turns 40, and it started in a wooden barn

Now that we've reached May, it's time for Disney+ to bring another wave of fresh movies and shows to the service, and if there's one thing we can rely on Disney+ for, it's for a real good bingeable series.

TV shows reign supreme this month on one of the best streaming services, one of them being the new series of Doctor Who whose episodes will come out weekly and will end with its finale on May 31. And let's not forget the abundance of new Star Wars titles landing on May 4 (appropriate), as well as more episodes of one of the best Disney+ shows, Andor.

And it doesn't end there, because at long last, Disney+ subscribers in the US will also finally be able to catch Spiderman: Across the Spider-Verse this month, being one of the movies leaving Netflix in May 2025 and finding its new place in the library of best Disney+ movies.

Everything new on Disney+ in May 2025

Arriving on May 1

Rise Up, Sing Out season 2 (TV show)
Spider-Man: Across the Spider-Verse (movie)

Arriving on May 2

Genghis Khan: The Secret History of the Mongols season 1 (TV show)

Arriving on May 3

Doctor Who season 2, episode 4 (TV show)

Arriving on May 4

Star Wars: Tales of the Underworld (TV show)
Star Wars: Galaxy’s Edge (TV show)
Star Wars: Rise of the Resistance (TV show)

Arriving on May 6

Andor season 2, 3 more episodes (TV show)

Arriving on May 7

Broken Karaoke season 3 (TV show)
Firebuds season 2 (TV show)
Hamster & Gretel season 2 (TV show)
Big City Greens season 4 (TV show)

Arriving on May 9

History’s Greatest of All Time with Peyton Manning season 1 (TV show)
The Toys That Built America season 3 (TV show)
The UnXplained season 7 (TV show)

Arriving on May 10

Doctor Who season 2, episode 5 (TV show)

Arriving on May 13

Andor season 2 finale (TV show)

Arriving on May 17

Doctor Who season 2, episode 6 (TV show)

Arriving on May 19

Tucci in Italy (TV show)

Arriving on May 20

Minnie’s Bow-Toons: Pet Hotel season 1 (TV show)

Arriving on May 24

Doctor Who season 2, episode 7 (TV show)

Arriving on May 28

Me & Winnie the Pooh season 2 (TV show)
Playdate with Winnie the Pooh season 2 (TV show)

Arriving on May 31

How Not to Draw season 3 (TV show)
Doctor Who season 2 season finale (TV show)

You might also like

• 'Nothing else to live for': Andor season 2 star Faye Marsay breaks down episode 6's soul-crushing moment for Vel in the hit Star Wars show
• How to watch the Star Wars movies in order: chronological and release date
• This new Disney+ documentary about Freddie Flintoff is the most inspiring movie I've seen all year – even though I've never watched a cricket match

When DeepSeek-R1 launched recently, it immediately captured the attention of the global artificial intelligence community, prompting major players such as OpenAI, Microsoft, and Meta to investigate its seemingly novel approach to model distillation. Yet, beneath the excitement around distillation lies a more nuanced and impactful innovation: DeepSeek's strategic reliance on reinforcement learning (RL).

Traditionally, large language models (LLMs) have been refined through supervised fine-tuning (SFT), an expensive and resource-intensive method. DeepSeek, however, shifted towards reinforcement learning, optimizing its model through iterative feedback loops. This method dramatically reduced costs, up to 90% compared to traditional methods such as those used by ChatGPT, while delivering comparable or even superior performance in various benchmarks.

The Real Revolution: Democratizing AI Knowledge

While model distillation, the method of teaching smaller, efficient models (students) from larger, more complex ones (teachers), isn't new, DeepSeek’s implementation of it is groundbreaking. Its true innovation is transparency. By openly sharing comprehensive details of their methodology, DeepSeek turned a theoretically solid yet practically elusive technique into a widely accessible, practical tool.

This openness accelerated adoption exponentially. Within weeks, the initial 60 distilled models released by DeepSeek multiplied into around 6,000 models hosted by the Hugging Face community. Developers around the globe now have practical blueprints for creating powerful, specialized AI models at significantly reduced scales.

By reducing the barrier to entry, DeepSeek’s open source strategy enables organizations of various sizes and sectors to explore sophisticated AI solutions that previously seemed out of reach. The widespread availability of distilled models means more specialized applications can emerge rapidly, opening doors to innovation in fields such as healthcare, finance, manufacturing, and education.

Implications for Businesses

For businesses, this marks a major turning point. The costly IT infrastructure required for traditional LLMs often barred smaller enterprises from adopting cutting-edge AI. DeepSeek’s distilled models promise powerful, tailored AI capabilities at a fraction of previous costs. Organizations can now easily leverage AI optimized specifically for their unique datasets, fostering deeper insights, operational efficiency, and enhanced competitiveness.

Moreover, these distilled models significantly lower the environmental impact associated with AI deployment. With sustainability becoming a central business imperative, companies can now align their AI strategies with broader corporate responsibility goals, reducing their carbon footprint without sacrificing technological capabilities.

Europe's Moment to Lead

Historically trailing behind AI powerhouses like the US and China, Europe is uniquely positioned to capitalize on DeepSeek’s approach. Europe's strength in open source collaboration, exemplified by initiatives like OpenEuroLLM and entities such as Mistral AI, aligns perfectly with DeepSeek’s ethos of openness.

Instead of competing in a costly arms race of extensive GPU infrastructure, European companies can lead by deploying energy-efficient, smaller-scale models. Given Europe’s significantly higher energy costs, this method of distillation presents a strategic advantage: sustainable and efficient AI solutions that are attractive to enterprises, consumers, and regulators alike.

Moreover, Europe’s regulatory landscape, which emphasizes data privacy and consumer protection, is particularly well-suited to smaller, more transparent models. By embracing DeepSeek’s distillation practices, European organizations can not only comply with stringent regulations more easily but also differentiate themselves globally through responsible AI practices.

Challenges and the Road Ahead

Despite its promise, model distillation isn’t without pitfalls. Poor implementation can inadvertently amplify biases or errors present in teacher models. These biases, if unchecked, could lead to unfair outcomes, regulatory scrutiny, or loss of consumer trust. However, with careful attention, rigorous testing, and responsible governance, these risks can be mitigated effectively.

Another challenge lies in ensuring the ongoing quality and consistency of distilled models. As the model pool grows exponentially, maintaining standards becomes more complex. The AI community will need robust verification processes and continual improvements to distillation techniques to sustain quality across thousands of models.

Training expertise is also critical. Despite the democratization of access, skilled personnel are necessary to effectively apply these distilled models to specific use cases. Investment in workforce development, continuous education, and community knowledge-sharing will be essential components in realizing the full potential of DeepSeek’s innovations.

The overarching benefits of DeepSeek’s open-source distillation methodology—a combination of economic efficiency, sustainability, and transparency—far outweigh the potential drawbacks. As businesses and nations recognize the opportunity, this innovative approach could very well redefine the future trajectory of AI development worldwide.

DeepSeek's blend of reinforcement learning, model distillation, and open source accessibility is reshaping how artificial intelligence is developed and deployed. This revolutionary approach holds significant promise not only for technological advancement but also for democratizing AI, driving sustainable innovation, and positioning regions like Europe as leaders in the global AI landscape.

Check out our comprehensive list of the 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

Marvel's Thunderbolts* movie is out now, and it's a film you won't want to miss.

I'm not just talking about how good it is, either (be sure to read my Thunderbolts* review to find out why). Indeed, it's far more important than many people realize, especially when it comes to setting up future Marvel Cinematic Universe (MCU) projects.

Below, I'll explain how Thunderbolts* lays the foundation for events still to come. I'll also answer your biggest questions about the Marvel Phase 5 film, including the foremost one on everyone's minds prior to its release. That being, what's with the asterisk in its title?

Full spoilers immediately follow for Thunderbolts*. Turn back now if you haven't seen it yet.

What does the asterisk mean in Marvel's Thunderbolts* movie?

Valentina (third left) uses her street smarts and marketing skills to turn the Thunderbolts* into The Avengers 2.0 (Image credit: Marvel Studios)

In short, the group's unofficial name, The Thunderbolts, is replaced with an official moniker, The New Avengers.

So the reason for the asterisk in the title is due to the titular team being renamed in the film's final scene, which is reconfirmed by The New Avengers' logo appearing during the movie's end credits crawl. From that point on, Yelena Belova and company will go by their new team name.

That won't be a huge shock to some MCU devotees. Ever since Marvel added the asterisk to the movie's title, fans have theorized that the ragtag group would be renamed before Thunderbolts* ended.

However, considering that the eponymous team comprises anti-heroes, reformed villains, and superpowered misfits, some observers suggested they'd collectively be called The Dark Avengers. After all, this is the name given to a group of superpowered individuals in Marvel Comics that includes Bob Reynolds/Sentry and John Walker/US Agent, two of the characters who appear in this film.

Yep, you're looking at the latest version of Earth's Mightiest Heroes (Image credit: Marvel Studios)

The New Avengers' moniker is a better fit for the group in the MCU, though.

Indeed, The Avengers haven't existed since they defeated Thanos in 2019's Avengers: Endgame. Sure, Sam Wilson, the second person to adopt the mantle of Captain America, plans to reform Earth's Mightiest Heroes in Captain America: Brave New World, but he's yet to do so before the events of Thunderbolts*.

That delay provides Contessa Valentina Allegra de Fontaine with the perfect opportunity to introduce Belova and company to the world as The New Avengers.

After defeating Sentry's darker half (aka The Void) in his Shadow Realm, The Thunderbolts – a name derived from Belova's childhood soccer team that lost every match it played – find themselves back in New York. Moments after returning to their reality, they spot the power-hungry de Fontaine and quickly pursue her down Main Street.

"Wait... us?" (Image credit: Marvel Studios)

Before they can arrest her for her crimes, though, the group stumbles into a street-based live press conference, which de Fontaine has hastily arranged. There, the CIA's duplicitous overlord tells the assembled press (and those watching the live broadcast) that Belova and company have not only saved the day, but are also a secret project she's been working on.

It's here that de Fontaine rechristens the group as The New Avengers. Confused but seemingly won over, the group accepts the new title, but not before Belova tells de Fontaine, "We own you now".

Essentially, the slippery CIA director has to do as she's told now. Otherwise, Belova and co. will come clean about de Fontaine's actual secret project, which involved Sentry/The Void's creation, her attempts to avoid impeachment, and so many other shady business dealings she made.

How does the titular team defeat The Void?

The Void is the physical representation of the dark part of Bob's split personality disorder (Image credit: Marvel Studios)

Through the power of friendship and hugging!

I joke, of course – I don't mean to belittle the weighty thematic material explored in Thunderbolts*. As someone who's suffered with mental health issues in the past, I applaud its cast and crew for tackling such difficult subject matter, and for telling a story that'll strike an emotional chord with many cinephiles. It certainly did with me.

Anyway, the titular group beat The Void via unconventional means. Bob Reynolds' two superpowered alter-egos are arguably the strongest beings – human race-wise, anyway – in the entire MCU (sorry, Thor and The Hulk). That much is made clear by how easily Sentry deals with the rest of The Thunderbolts during their fight in the Watchtower.

The Void is even more powerful than Sentry. Again, that's evident by the ease with which his Shadow Realm spreads across New York City. (As an aside, I do wonder if Daredevil: Born Again season 2 will reference this event. After all, that Disney+ show is also set in New York, so, unless the series' sophomore season takes place prior to Thunderbolts*, it would be odd if it didn't mention the fact that a giant shadow literally overtook the city.)

Bob was experimented on in Malaysia by de Fontaine's OXE group (Image credit: Marvel Studios)

But I digress! In a bid to stop The Void, Belova – and, some time later, the rest of the team – enter the Shadow Realm, fight their way through their worst memories and traumas, and eventually find Bob, who's hiding in his 'best' worst memory.

Working together, the group eventually finds a memory of Bob's that shows he's not only been experimented on by de Fontaine's OXE research group, but also marked the first time that The Void manifested.

Long story short, The Void overpowers everyone except for Bob, who eventually fights The Void for control of his mind. As Bob pummels his darker self, though, The Void starts to silently overtake his consciousness with its shadow powers – a move that'll wipe Bob out of existence because The Void will have absorbed him into itself.

In a Hail Mary move, Belova – with Red Guardian's help – breaks free of her shackles, parkours her way to Bob, hugs him from behind, and tells him he's not alone. The rest of the team soon follow and silently embrace Bob, too. Overwhelmed by their humanity and comradeship, Bob breaks down and, in doing so, stops physically and mentally beating himself up.

Free of the darkness that consumed them, the group is suddenly transported back to New York. Belova asks Bob if he's okay, to which he responds that he is but has no idea what just happened. From here, we lead into the above section's stuff about de Fontaine, the press conference, and The New Avengers reveal.

Who dies in Thunderbolts*?

Taskmaster's time in the MCU was short and not very sweet (Image credit: Marvel Studios)

There's only one major character death in Thunderbolts* and it's one that was foreseen by fans many months before the film's release.

While appearing in Thunderbolts* first trailer and initial slate of press images last August, Olga Kurylenko's Antonia Dreykov/Taskmaster has been conspicuous by her absence in promotional materials ever since. In fact, in the weeks leading up to one of May 2025's new movies arriving in theaters, Taskmaster has been completely MIA.

Inevitably, that led MCU devotees to suspect she wasn't long for this world – and that proves to be the case. During the four-way fight between Belova, Walker, Ava Starr, and Dreykov in the film's first act, Taskmaster is shot in the head and killed by Starr. It's an unceremonious way for her to bow out of the MCU and, based on some early fan reactions following the launch of Thunderbolts*, it's a death that's not gone down well with some.

Are there any Marvel cameos in Thunderbolts*?

Anthony Mackie's Captain America is referenced, but he doesn't physically appear (Image credit: Marvel Studios)

No. Unlike Bucky Barnes' surprise cameo in Captain America 4, there are no unexpected appearances from new or returning MCU characters in Thunderbolts*.

There's the odd reference or two to those who populate Marvel's cinematic juggernaut. Wilson is namechecked a couple of times. Meanwhile, former US President Thaddeus 'Thunderbolt' Ross is briefly mentioned by de Fontaine during her Senate hearing; the CIA director reminding the assembled elected individuals of Ross Hulk-ing out in Brave New World.

Those rare occurrences aside, you won't see other Marvel characters at any point. Well, unless you count the appearance of a certain group by way of a space vessel, but I'm getting ahead of myself.

Is Geraldine Viswanathan's Thunderbolts* character Mel going to become Songbird in the MCU?

There may be more to Mel (left) than meets the eye (Image credit: Marvel Studios)

Geraldine Viswanathan's Mel, aka de Fontaine's assistant, might not seem like the superhero type. In fact, to the average viewer, one of the primary supporting characters in Thunderbolts* will be seen as nothing more than de Fontaine's diligent albeit emotionally torn lackey.

Diehard Marvel fans, though, may pick up on the fact that Mel isn't just a simple background character. If you were paying attention, you'll have noticed Mel wears a gold necklace – one that has a gold bird sitting in a ring – throughout the film. Hmm, I wonder where we've seen that before...

Alright, enough teasing. Some fans believe Viswanathan is another Marvel hero hiding in plain sight. That's because, in Marvel Comics, there's a superhero called Songbird whose civilian identity is – wait for it – Melissa Gold.

Songbird's powerset includes the ability to create supersonic soundwaves, fly, and use echolocation (Image credit: Marvel Comics)

Now, this could be nothing more than a curious case of misdirection on Marvel Studios' part. That said, the fact that there's a character called Mel who wears a gold necklace with a songbird hanging at the end of it is too much of a coincidence in my and many other viewers' eyes.

It might not be a coincidence, either. In a January 2025 chat with ScreenRant, Viswanathan danced around a question about whether she's playing Songbird in the MCU.

Asked if she considers herself to be a songbird, she replied: "I have no idea what you're talking about. Never in my life have I heard something like that...Oh, look what you've done. Now, I'm all nervous."

Don't be surprised, then, if Viswanathan isn't just confirmed to be playing Songbird in the MCU at some point, but that she'll also return in the next Avengers movie as part of a yet-to-be-revealed second round of Avengers: Doomsday cast reveals. Sure, Doomsday may already have 27 confirmed characters, but there's always room for more!

Does Thunderbolts* have a mid-credits scene?

MCU fans won't want to leave their seats early for this Marvel movie (Image credit: Marvel Studios)

There is – and it's a call-back to a humorous line of dialogue Alexei Shoshtakov/Red Guardian utters around the film's midpoint.

The scene opens with Shoshtakov loitering in the cereal aisle of a grocery store. Striking up a conversation with a fellow shopper, he tries to convince her to buy a box of Wheaties, which is adorned with images of The New Avengers. He even poses next to the box, hoping that she'll recognize him (FYI, it doesn't work).

Earlier in the film, Shoshtakov jokes that, if The Thunderbolts save the day by putting a stop to de Fontaine's nefarious plans, they could not only end up on the front of Wheaties cereal boxes, but also be the "little kiddie toy" that children find among the cereal itself.

Despite appearing to convince his fellow cereal enthusiast to buy a box of Wheaties, though, Red Guardian doesn't realize that the person puts it back on a different shelf when he's not looking. Clearly, he put them off with his over-the-top persuasive tone!

What is the Thunderbolts* post-credit scene? And how does it set up The Fantastic Four: First Steps?

The Fantastic Four are seen – well, their starship is – in Thunderbolts* end credits stinger (Image credit: Marvel Studios)

Thunderbolts' end credits stinger takes place 14 months after the film's primary narrative.

Initially, it reunites us with Bucky Barnes, Belova, Walker, and Starr, who have returned to The New Avengers' headquarters – they might be stationed in The Watchtower, aka the building formerly known as Avengers Tower, but it's hard to tell. The quartet is sporting new costumes that look a bit corny (I'm hoping this is down to the corporate or commercial nature of The New Avengers, rather than the suits they actually decided to wear!).

Anyway, as they enter the main room of their HQ, they discuss the fact that Wilson/Cap 2.0 is suing them for trademark infringement. Wilson is upset they've snatched the Avengers moniker from under his nose. Despite Bucky meeting Wilson to smooth things over, Bucky reveals the conversation "went poorly".

As the conversation continues, Belova mentions some form of "space crisis," but, before we learn more, Shoshtakov interrupts to say he's fashioned a new, garish-looking tracksuit they can all wear. It's littered with sponsors, such as Tide, and has a giant 'New AvengerZ' logo on the front, which Shoshtakov suggests can circumvent Wilson's attempts to sue them for the team's new name.

Will we see The New Avengers greet The Fantastic Four in one of First Steps' likely end credits scenes? (Image credit: Marvel Studios)

Bob is also present, but he hasn't – and doesn't want to – use his powers to fight anyone (if it comes to that) because he's worried The Void will return. He says he "did the dishes", though, so at least he's of some use.

Suddenly, a data pad held by Belova lights up. Moments earlier, Bucky told Belova to use Earth's satellite system to learn more about these "space anomalies" they've been made aware of. Lo and behold, Bucky's idea works – and it's here that Thunderbolts* reveals its ties to The Fantastic Four: First Steps, which arrives in theaters on July 25.

The data pad alerts the group that a mysterious space vessel has appeared above Earth-616 via a multidimensional portal. Belova pulls up the footage on her tablet and, as a short motif of The Fantastic Four's MCU theme plays, we see the group's futuristic-looking rocket ship, complete with a huge 'F4' logo on it, barreling towards Earth.

Downey Jr's Doctor Doom will be the Multiverse Saga's primary antagonist (Image credit: Walt Disney Studios/Marvel Studios)

So, how does this film's post-credits scene set up what'll happen in The Fantastic Four's latest big-screen reboot?

The short answer is: I'm not sure, but we know First Steps is set in an alternate reality. That much was made clear last April when Marvel teased some key story details about the film and then released a poster that confirmed to fans that it wouldn't be set in the MCU. So, Marvel's First Family will exist in another universe for much of First Steps' runtime.

"For much of" are the key words here. Again, we've no idea about how First Steps, aka the first Marvel Phase 6 movie will end, but fans have speculated that the titular team will be forced to flee their universe – because First Steps' main villain Galactus destroys it, Robert Downey Jr's Victor von Doom (who'll be the Big Bad of Avengers: Doomsday and its sequel Secret Wars) casts them out, or for some other reason.

Regardless, Thunderbolts* confirms The Fantastic Four will find themselves in the MCU by the time First Steps' story ends. The *ahem* steps taken to get there, though, are still up for debate.

You might also like

• Find out how to watch the Marvel movies in order
• Get the lowdown on Spider-Man: Brand New Day
• See if you agree with our ranking of the best Marvel movies

Marvel Unlimited monthly subscription: was $9.99 per month now $4.99 at Marvel
The ultimate digital comics subscription for Marvel fans is offering new and returning customers 50% off their first month, which is much cheaper than a streaming subscription to watch all the best superhero movies. To gain access to a library of more than 30,000 comics, use the code 'SPIDEY50' at checkout. You're able to use the Marvel Unlimited app on all iOS and Android devices, including the web, too. But, be quick – the deal expires on May 5!View Deal

The adoption of technology can spark all kinds of emotions – not just in terms of the tech itself but also the brands behind them. As consumers, it’s widely accepted we often allow our emotions to influence our tech purchasing decisions. But to what extent do businesses or other large organizations use an emotional response to help make similar purchasing decisions?

The reason I ask that is because there’s a suspicion that as IT professionals – whether we like it or not – we’re increasingly relying on an emotional response to influence our technology choices. And increasingly, we’re seeing this in relation to the adoption of generative AI (Gen AI) solutions.

This may seem like an odd observation to make, especially when making objective investments in technologies such as SaaS platforms or security solutions where everything should be benchmarked against KPIs, ROI and other measures.

And yet, a clear pattern of behavior is emerging that links the adoption of Gen AI with a more ‘emotional’ purchasing response. Here at Tanium, we call it the ‘Emotional Hype Cycle’, and it’s already being used to influence our IT strategy. And while it leans on Gartner’s Hype Cycle, the big difference is the emotional factor and the important part this plays.

The ‘Emotional Hype Cycle’ explained

The easiest way to explain it is to look briefly at all eight stages of the curve to see just how much of an emotional rollercoaster it can be for enterprises and employees when making decisions about technology:

Triggered: A new innovation breaks onto the scene and becomes the ‘latest thing’. It’s not just new – the tech can be challenging and disruptive, triggering a range of emotions. Generative AI isn’t just another breakthrough -- it’s a seismic shift that has the potential disrupt industries and reshape the way we think about technology itself. As a game-changing technology, it has the ability to ignite excitement and deep unease and mistrust in equal measure.

Hyped: People, businesses, sectors and governments get caught in the excitement of it all. While there may be some notable high-profile successes, there are plenty of failures along the way.

Rejection: After the excitement, it’s time for the post-hype headache as organizations face setbacks such as the technology failing to live up to expectations. As technology is rejected, some organizations will prohibit adoption. As a result, staff are left in limbo and shadow IT emerges.

Doubt: Some may view this rejection as a failure. Instead, the period – which is often shaped by caution, doubt and skepticism – allows enterprises to assess the technology with greater objectivity. It provides breathing room for organizations to consider ethical factors such as bias, as well as other issues like privacy and protection. Hesitancy is the default emotion here.

Comfort: Over time, organizations tend to develop a sense of gradual acceptance as they build their own test beds and case studies in controlled environments. To start with, access is limited, but these pilots are useful since they help to manage excitement and expectation.

Habitual: Now things start to pick up. Businesses get into the rhythm of using and applying the technology thanks to the establishment of common standards and governance. It’s at this point that we see the emergence of fair and effective user policies – along with more established working patterns – that not only set the tone but also the foundation for enterprise-wide rollouts.

Social: The technology becomes part of the social fabric of the organization, with adoption now widespread. Critically, when it comes to the technology’s strategic role, there is a connection and alignment between IT and the wider business.

Creative: This is a transformative stage for users and the enterprise as people use the technology to innovate and operate in productive new ways. Not only is this rewarding, but it’s also an early indicator of technology maturity.

What’s clear is that the Emotional Hype Cycle provides a clear roadmap to help organizations move from uncertainty to maturity as they adopt Gen AI. At its heart, it emphasizes a thoughtful, structured approach to integrating new technologies, ensuring decisions are guided by governance, ethical considerations and strategic alignment.

Practical strategies that deliver the smooth adoption of Gen AI

Crucially, it acknowledges human tendencies to react to ‘feelings’ or ‘hunches’ while offering practical strategies to help smooth the adoption of Gen AI. 

For instance, the ‘Doubt’ phase calls for a rigorous, ethical assessment of AI, addressing concerns like bias, privacy and security. This is followed by the ‘Comfort’ stage, which encourages organizations to run controlled pilots and test beds, allowing enterprises to experiment with Gen AI’s potential in a safe and manageable way.

Similarly, as organizations progress, the framework calls for the incremental development of governance and standards, something that is central to the ‘Habitual’ stage. In terms of a strategic approach, it acts as a brake on runaway excitement or hysteria while helping to establish processes and protocols for responsible use.

Wherever you look, the Emotional Hype Cycle offers a way to navigate the highs and lows of technology adoption. It empowers organizations to balance excitement with caution while embracing Gen AI in a way that is measured. Crucially, it allows businesses to embrace Gen AI as a transformative force for innovation and productivity.

We provide an extensive list of the 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

The Samsung Galaxy Z Flip 7 might not be a comprehensive upgrade on the Samsung Galaxy Z Flip 6 from what we’ve heard so far, with various aspects possibly staying the same. But there are some parts of the phone that likely will be substantially improved.

While nothing is certain yet, we can look at claimed leaks and rumors to get a good idea of what specs we might see.

So that’s exactly what we’ve done below, and where there aren’t any tips we’ve used educated guesses to fill in some of the gaps.

Samsung Galaxy Z Flip 7 predicted specs

If you’re familiar with the Samsung Galaxy Z Flip 6’s specs then you’ll note that the predicted specs in the chart above are a mix of old and new. But let’s start with some of the areas where the Samsung Galaxy Z Flip 7 will reportedly be an improvement on its predecessor.

One of the main upgrades could be the screen sizes, with multiple sources suggesting the Galaxy Z Flip 7 could have a roughly 4-inch cover screen, which is up from 3.4 inches on the Samsung Galaxy Z Flip 6. So that would make the cover screen significantly larger this year.

One of those sources also points to a slightly larger 6.85-inch foldable screen – up from 6.7 inches on the Samsung Galaxy Z Flip 7. So, in both cases you should have more space to interact with the phone, though this should have the greatest impact on the cover display, which is quite cramped on the Z Flip 6.

So far there’s no word on the resolution, but presumably the Samsung Galaxy Z Flip 7 will either match or better the Z Flip 6 here, meaning a foldable screen of at least 1080 x 2640, and a cover screen of at least 720 x 748.

We also fully expect the foldable display will have a 120Hz refresh rate, since the Galaxy Z Flip 6’s does, while the cover screen will probably be 60Hz.

The Samsung Galaxy Z Flip 6 (Image credit: Philip Berne / Future)

Another upgrade we’ll almost certainly see is the chipset, with the Samsung Galaxy Z Flip 7 rumored to use either a Snapdragon 8 Elite or an Exynos 2500.

So far there’s no agreement on which – and it could be that some regions will get one and some the other, but while the Snapdragon 8 Elite will likely outperform the Exynos 2500, either of these would be an improvement on the Snapdragon 8 Gen 3 in the current model.

RAM and storage however may not be improved, with the one leak on that front pointing to the Samsung Galaxy Z Flip 7 having 12GB of RAM and a choice of 256GB or 512GB of storage, just like the Z Flip 6. However, its performance may get a boost from a larger vapor cooling chamber.

One of the more disappointing things we’re hearing is that there might be no upgrades at all to the Samsung Galaxy Z Flip 7’s cameras, with leaks pointing to the same 50MP main, 12MP ultra-wide, and 10MP front facing cameras as last year.

Finally, there’s the battery, and multiple sources point to this being 4,300mAh, which would be a noticeable upgrade on the 4,000mAh of its predecessor. That’s great news, as 4,000mAh really isn’t much for a phone of this size.

What we don’t know is how fast it might charge, but the Samsung Galaxy Z Flip 6 supports 25W wired charging and 15W wireless, so expect at least that much.

You might also like

• Samsung Galaxy Z Flip 7 colors: every rumored and predicted shade
• Samsung Galaxy Z Flip 7 release date rumors: when might the new phone launch?
• Samsung Galaxy Z Fold 7: latest news, rumors, and everything we’ve heard so far

• Report warns sophisticated bots mimic human behavior so well outdated defenses don’t stand a chance
• Mobile apps are under siege, with a 160% rise in targeted bot traffic year over year
• CAPTCHA farms and rotating proxies help bots bypass basic defenses

The internet has entered a new era where automated traffic now accounts for more web activity than human users, new research says.

Radware's 2025 ecommerce Bot Threat Report claims the majority of traffic to online stores during the 2024 holiday season didn’t come from people. It came from bots.

For the first time, automated programs - ranging from simple scripts to AI-enhanced digital agents - accounted for 57% of all traffic, surpassing human visitors on e-commerce websites.

A smarter generation of bad bots

The report highlights the ongoing evolution of malicious bots, as nearly 60% now use behavioral strategies designed to evade detection, such as rotating IP addresses and identities, using CAPTCHA farms, and mimicking human browsing patterns, making them difficult to identify without advanced tools.

The only effective counter is equally intelligent detection - AI-powered defenses that can learn and adapt. Businesses must reassess their security stack and look beyond basic filters to solutions offering advanced DDoS protection and intelligent traffic monitoring.

“Bad bots are no longer just based on simple scripts - they’re sophisticated, AI-enhanced agents capable of outsmarting traditional defenses,” said Ron Meyran, Vice President of Cyber Threat Intelligence at Radware.

“E-commerce providers and online retailers that rely on conventional security measures will find themselves increasingly exposed, not just during the holidays but year-round.”

Mobile platforms have become a critical battleground, with a staggering 160% rise in mobile-targeted bot activity between the 2023 and 2024 holiday seasons. Attackers are deploying mobile emulators and headless browsers that imitate legitimate app behavior.

The report also warns of bots blending into everyday internet traffic. A 32% increase in attack traffic from residential proxy networks is making it much harder for ecommerce sites to apply traditional rate-limiting or geo-fencing techniques.

Perhaps the most alarming development is the rise of multi-vector campaigns combining bots with traditional exploits and API-targeted attacks. These campaigns go beyond scraping prices or testing stolen credentials - they aim to take sites offline entirely.

For businesses relying on the best ecommerce website builders or user-friendly platforms, the threat is clear. Security must evolve in step with the attackers. Platforms must also adopt dedicated mobile protections to defend against these increasingly sophisticated threats.

You might also like

• Take a look at the best VPN with antivirus available right now
• PC makers are planning plants in Saudi Arabia to try and avoid US tariffs
• These are the best password managers you can use

Apple does not hide its deeds under a bushel; When it "lost" the Epic (Fortnite) case last year, it quickly rolled out a series of changes to satisfy the injunction.

If no one else realized Apple had somehow won by losing, Epic certainly knew it, with Epic CEO Tim Sweeney stating on X (formerly Twitter), "The court battle to open iOS to competing stores and payments is lost in the United States. A sad outcome for all developers."

In answer to that ruling, Apple started charging a 27% commission on App Store apps that chose to manage their own billing and forced them to apply for a special tool that let app developers include a button to outside billing systems. Additionally, Apple required that developers still support Apple's in-app billing system as an alternative to third-party ones.

It was all there in black and white, but those decisions triggered what may be one of the most scathing rulings I have ever read. I mean, US District Court Judge Yvonne Gonzales Rogers who has been presiding over this case for years is pissed.

Was Apple trying to follow the letter of the law while skirting around the bits that might have any impact on its commission revenue business? As Judge Gonzalez sees it, yes.

Don't believe me? Read this unedited bit from the court ruling:

"In stark contrast to Apple’s initial in-court testimony, contemporaneous business documents reveal that Apple knew exactly what it was doing and at every turn chose the most anticompetitive option.

To hide the truth, Vice-President of Finance, Alex Roman, outright lied under oath. Internally, Phillip Schiller had advocated that Apple comply with the Injunction, but Tim Cook ignored Schiller and instead allowed Chief Financial Officer Luca Maestri and his finance team to convince him otherwise.

Cook chose poorly. The real evidence, detailed herein, (Case 4:20-cv-05640-YGR Document 1508 Filed 04/30/25 Page 2 of 80 3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 United States District Court Northern District of California) more than meets the clear and convincing standard to find a violation. The Court refers the matter to the United States Attorney for the Northern District of California to investigate whether criminal contempt proceedings are appropriate."

As the judge sees it, Apple may have used the ruling to further enrich its commission business:

"Prior to the Injunction, Apple did not allow 'linked-out purchases' and, thus, did not charge any commission for purchases made outside of, or off, its platform. Now, it does."

The filing shows how Apple worked to fulfill the injunction in such a way that, while maybe opening new third-party billing and revenue avenues to developers, it also solidified Apple's hold on the process and commission collection.

Judge Gonzalez paints this as a pretty nefarious operation with Apple willfully hiding and lying about its intentions.

The thing is, Apple has never lied about its intentions regarding the App Store.

The Apple way

Apple likes full-stack control for, well, control, but also so it can fulfill its security and privacy promises. An App Store where you can trust the majority of apps is not something that occurs naturally. You must set the parameters and ensure that every app that passes through those narrow gates does what it promised and is free from malware and spyware.

It's no accident that your best iPhone is not riddled with malware or that you rarely download a fake app (does anyone remember the early days of Android apps before Google Play?!). The more open Apple is with its platforms, the more risk it incurs for itself and its customers.

It's also clear, though, that Apple doesn't think much about methods or perception. If it was being forced to give third-party fulfillment systems access to App Store customers, then why not charge third-party developers for the pleasure? After all, the links to these platforms would come through Apple. Was that a deterrent to the development of such systems? Probably. Was that Apple's intention? I don't know.

If you believe the injunction, you know some within Apple knew that Apple's methods of compliance were too close to flouting the law. Those voices, though, did not win. Former Apple CFO Luca Maestri, once a voice of reason on Apple earnings calls, is named a key champion of potentially bad ideas.

As for what Apple does next, well, I'm sure it will fight, but there's also no way it can ignore Judge Gonzales' closing statement:

"This is an injunction, not a negotiation. There are no do-overs once a party willfully disregards a court order. Time is of the essence. The Court will not tolerate further delays. As previously ordered, Apple will not impede competition. The Court enjoins Apple from implementing its new anticompetitive acts to avoid compliance with the Injunction. Effective immediately Apple will no longer impede developers’ ability to communicate with users nor will they levy or impose a new commission on off-app purchases."

What happens next

If this ruling holds, which is not a certainty as Apple plans to appeal, it may trigger sweeping changes that could forever alter the way we use the App Store.

The loss of a fully closed system may not be the best thing for consumers. Still, it gives developers the flexibility many have been craving and effectively takes the Apple commission (a..k.a. the "Apple tax") off the table for transactions happening outside the platform, even if users found that link from inside Apple's App Store.

Epic is already planning for a best-case scenario and quickly announced that Fortnite will return to the App Store next week with, I'm sure, in-app purchases (those skins!), all taking you outside Apple's platform.

We will return Fortnite to the US iOS App Store next week.Epic puts forth a peace proposal: If Apple extends the court's friction-free, Apple-tax-free framework worldwide, we'll return Fortnite to the App Store worldwide and drop current and future litigation on the topic. https://t.co/bIRTePm0TvApril 30, 2025

This might be the first effective crack in Apple's "reality distortion field," one that refashions reality to suit its needs. We've seen it at work this year with Apple not delivering the full Apple Intelligence experience but telling consumers that the iPhone 16 comes with it. It has since quietly removed some of that language from product pages.

Apple has rarely admitted mistakes or shortcomings, and while this ruling isn't a sign that Apple will start now, it appears to have no choice but to deconstruct the distortion field it built around app developer choice.

As for the injunction, it reads like the final word on the subject:

"Apple willfully chose not to comply with this Court’s Injunction. It did so with the express intent to create new anticompetitive barriers which would, by design and in effect, maintain a valued revenue stream; a revenue stream previously found to be anticompetitive. That it thought this Court would tolerate such insubordination was a gross miscalculation. As always, the coverup made it worse. For this Court, there is no second bite at the apple. "

Ouch.

You might also like

• Apple ‘strongly disagrees’ with ruling that could change the App Store – as Fortnite says it could now return to iPhones
• Apple officially delays the AI-infused Siri and admits, ‘It’s going to take us longer than we thought’
• The Bella Ramsey Apple Intelligence ad that disappeared, and why Apple is now facing a false advertising lawsuit
• AI took a huge leap in IQ, and now a quarter of Gen Z thinks AI is conscious

• Triple Boost 14 Pro adds three screens to your laptop from a single USB cable
• It is designed for remote workers needing more display space on the go
• Matte 14-inch screens offer with 100% sRGB, ideal for creative professionals

Aura Displays has launched the Triple Boost 14 Pro, a new portable three-monitor setup for laptops that uses only one USB cable for both power and data.

Designed for remote workers, digital professionals, and laptop users in need of more screen space, the Triple Boost 14 Pro offers three 14-inch Full HD screens.

The company says this is the first time a full triple-display solution has been offered in such a compact, cable-free form factor.

Tri-Link technology

The Triple Boost 14 Pro is constructed with a full-aluminum housing and weighs 4.85lb (2.2kg). When folded, it measures 13.9 inches wide and 1.38 inches thick.

Each screen is a matte 1080p LED display, with a brightness of 300cd/m², a 1500:1 contrast ratio, and a 178-degree viewing angle. The displays cover 100% of the sRGB color space and refresh at 60Hz. Aura says this makes the unit especially suitable for design, coding, analysis, and multitasking.

The key to its single-cable setup is Aura’s proprietary Tri-Link technology, based on a custom-optimized SiliconMotion chipset. This chip handles power and video for all three screens through a single USB connection, removing the need for additional cables or adapters.

The product folds down into a compact form and is aimed squarely at users who travel often or work in temporary setups. The company sees the device as an ideal tool for those who need more workspace on the go but want to avoid carrying separate monitors or docking setups.

Pre-orders for the Triple Boost 14 Pro are live now, with shipments set to begin in early May 2025. It’s priced at $469 (28% off the MSRP of $649) and there’s a welcome offer on the site that will give you an extra $10 off ($459), plus two free tempered glass screen protectors (worth $69.98), and free worldwide shipping.

Aura Displays says early demand will determine its wider retail rollout later in 2025.

You might also like

• These are the best business laptops you can buy right now
• At almost 24 inches each, this is the largest dual-screen monitor I've ever seen
• At 11lbs, this double 24-inch 'portable' monitor is a bit too much for me

• Chatbot interactions could soon be serving you related Google ads too
• ChatGPT and others are snaffling Google's search market dominance
• Its dominance has already come under fire by the DOJ and others

Google is now showing ads in chatbot conversations via its AdSense for Search network, marking a shift in its digital advertising strategy as generative AI becomes not only more commonplace, but increasingly used in place of traditional search engine tools.

Previously, ads were shown within the search results of other websites, however Google added them to conversations with chatbots operated by AI startups earlier in 2025 (via Bloomberg).

The rollout comes after the tech giant conducted tests last year, and earlier this year, with AI search apps like iAsk and Liner.

Google Ads are now being shown in chatbot interactions

Google's ads business has been an incredibly lucrative area for the California company, which maintains a 90% share of the global search market (via Statcounter).

However, with the emergence of rival AI chatbots like OpenAI's ChatGPT, consumers are now able to obtain information more succinctly and more quickly, putting Google's business under threat, and therefore its ads network.

A company spokesperson has confirmed: "AdSense for Search is available for websites that want to show relevant ads in their conversational AI experiences."

Google's collaboration with startups is a strategic move to stay profitable even if the search market as we know it is shrinking, but it hasn't gone unnoticed.

Two US senators recently questioned Google about its partnership with Anthropic, expressing concerns about competition in the sector. Microsoft's deal with OpenAI was also under fire.

The Department of Justice also recently noted Google’s search engine market monopoly is helping its AI strategy, and that its use of AI if further enhancing its search market monopoly.

Earlier in April 2025, a US judge also deemed Google violated antitrust laws by “willfully acquiring and maintaining monopoly power” in the advertising technology market. Though found guilty, a further hearing will be required to determine potential remedies.

You might also like

• Google CEO hopeful Gemini will be integrated into Apple Intelligence in time for iPhone 17 launch
• We've listed the best browsers, including Chrome alternatives
• Reach more people with the best online marketing services

• World Password Day is here
• Most TechRadar Readers are using Google Password
• Password manager's are considered best practice for password hygiene

We polled our TechRadar Pro readers on WhatsApp (you can join our channel here) to ask them which password manager they use to keep their accounts safe, and the most popular was Google password manager, a free to use service easily accessible and built into Chrome and Android devices, with 39% of respondents choosing this service. You can check out our full review of Google Password Manager here.

Worryingly, following closely behind were the readers who don’t use a password management service at all, with 20% of our readers choosing to manage their accounts by themselves. Take a look at our picks for best password manager here if you need any recommendations.

Trusting a third party with your passwords and credentials naturally comes with an element of risk, especially given that security attacks on password managers have soared in just the last year. If you choose to handle your password hygiene yourself, make sure to rotate them often, never reuse them between sites, and always change them immediately if they’re exposed in a breach.

A close second

Of those who do use a password manager, Apple Passwords came just behind Google Passwords with 23% choosing the service. This is a pretty basic manager, free and built in with Apple’s OSes, which helpfully reports on data breaches and password health.

Many of our readers (17%) use a different third party service, if that’s you, feel free to let us know which you recommend and why. No matter which you use, password management is definitely important, especially given how much of our lives is digital;

"Good password hygiene and management is crucial in today's world, where we all have many different accounts,” says Joel Latto, Threat Advisor at F-Secure

”If a hacker gains access to an account that uses a similar password to another, it's easy for them to also obtain information from the other account. This makes it easier for them to obtain more information about you, putting you at higher risk of identity theft, which can have devastating and distressing consequences including financial loss."

You might also like

• Take a look at our picks for the best malware removal software around
• Check out our choice for best antivirus software
• Microsoft RDP apparently lets you log in with expired passwords - and it apparently doesn't have plans to fix the issue