TechRadar News

• An over-the-air update has "bricked" many Samsung HW-Q990D soundbars
• The update has been pulled by Samsung
• It's unclear what those affected can do

Owners of Samsung's five-star soundbar, the Samsung HW-Q990D, are having a less than five-star experience right now. In what appears to be a faulty firmware update, many users are reporting huge problems that are effectively turning their Samsung soundbars into little more than long black bricks.

If you have a Samsung soundbar and you're not currently affected by this, it would be wise to turn off its auto-update feature (if it has it) for the time being. According to The Verge, the same issues may be affecting the Samsung HW-Q800D and HW-S801D – albeit in significantly smaller numbers so far.

So far, all a Samsung spokesperson had to say about the issue is: "Samsung is investigating the cause of the issue, and is taking immediate action to resolve the firmware error."

That leaves a lot of unanswered questions, so here's what we know so far.

What's going wrong with Samsung's soundbar firmware?

The firmware update, version 1020.7, has rendered some soundbars completely unresponsive: you can turn them on, but they appear to freeze and cannot be adjusted via Samsung's SmartThings app. Right now there's a 31-page (and growing) thread on the Samsung support site, with unhappy owners also turning to AV forums and Reddit.

Towards the end of that thread there's a screenshot apparently of a UK Samsung moderator, who says that "we have received feedback from our AV product specialists that this issue is a result of an OTN (Over The Network) software update... the OTN is now suspended, and customers that have updated via USB are not affected."

The post continues: "if your issue is the result of an OTN update, then it will be necessary to arrange an engineer's inspection and repair in order to get this resolved." The customer is then asked to contact Samsung support directly.

As yet we don't know if that's the only solution, or if there's a way for Samsung to use an OTN update to fix the last one. We've followed up with Samsung to request instructions for those affected as soon as possible.

Whatever the fix, this is a bad look for Samsung: this is a soundbar with a launch price of $1,799 / £1,699 / AU$1,995, not some cheap buy from a no-name brand. And as we've seen from the Sonos app disaster, if issues aren't resolved quickly they can do serious damage to brands among the early adopters and big spenders they work so hard to attract. Especially with the Samsung HW-Q990F soundbar that replaces the Q990D due for release in a matter of weeks…

You might also like

• Sonos cancels its TV streamer amid the ongoing app issues
• The best soundbars you can buy
• 5 cheap soundbars to upgrade your TV audio

• Intel's Panther Lake lineup won't come to consumers until Q1 2026, according to a leaked presentation slide
• This comes after the expected full launch in late 2025, but that is only for the Early Enablement Program
• Rumors suggest the delay is due to Intel's 18A production process

Intel's Lunar Lake Core Ultra 200 processors are already here, present across powerful laptops (such as the MSI Prestige EVO AI+ lineup) and handhelds (notably the MSI Claw 8 AI+). Now, it's time to look forward to Team Blue's Panther Lake lineup - but it may not come as soon as initially promised.

As reported by VideoCardz, a leaked slide from a special event in China has seemingly revealed that Intel's Panther Lake processors won't launch until Q1 2026, while its EEP (Early Enablement Program) will start later this year in preparation. This is contrary to former interim CEO Michelle Holthaus' previous confirmation of the lineup's launch and availability in the second half of 2025, indicating a delay. Ultimately, it means consumers may have to wait until 2026 to get their hands on Intel Panther Lake-powered products.

Reasons for the delay reportedly stem from difficulties with managing Team Blue's 18A manufacturing process in terms of the original production schedule. It's not exactly clear why this delay is in place, but if the setback is significant enough, it could have an impact on Intel's placement in the market (especially up against AMD).

We know that AMD still has its Ryzen AI 400 series processors set to be announced, and we still don't know if it will launch around the same time as Intel's Panther Lake.

It's also worth noting that Team Red's Ryzen AI 300 series APUs have already made huge leaps in processing power, notably the Ryzen AI 395+ Max - so if its new lineup does arrive in Q1 2026, it will be interesting to see how it fairs against the Panther Lake chips using the Xe3 architecture.

(Image credit: Tester128 / Shutterstock) What does this mean for handheld gaming PCs?

With Intel's Core Ultra 7 258V processor, we've seen the new MSI Claw 8 AI+ (the same chip is also used in Dell's XPS 13 laptop) work wonders at providing huge improvements in gaming - especially in comparison to the previous MSI Claw A1M. Spoiler: it wasn't good.

With Panther Lake processors now on the (somewhat distant) horizon, I'm hopeful that we can get even stronger gaming ultrabooks and handheld gaming PCs based on these new chips. Intel's Technical Marketing Vice President, Robert Hallock, confirmed Team Blue's plans to support gaming handhelds (as reported by LaptopMag) and this could work perfectly for competition.

AMD's Ryzen Z2 Extreme is slated for 'early 2025' as a part of the Ryzen AI 300 series family - it will be used for handhelds like the Lenovo Legion Go 2 (still a prototype), but it's still not clear when that will launch. It also seems as though it will rival the Core Ultra 7 258V processor, which may end up being a good indication of how both Team Red and Team Blue's upcoming lineups could perform against each other.

I hope that both lineups provide drastic leaps in power for handhelds across the board, further placing handheld gaming within the limelight where it belongs - let's just hope that these delays don't make it a long wait...

You may also like...

• Intel in 2024: year in review
• Intel Core Ultra 9 285K and Intel Core Ultra 5 245K Review
• Can 18A save Intel from being devoured by its rivals – and Wall Street?

• A popular tool for automated software updates was compromised via GitHub
• A piece of malicious code was added, exposing user secrets
• Dozens of organizations were harmed already, researchers said

Tens of thousands of organizations, from SMBs to large enterprises, were at risk of inadvertently exposing internal secrets after a supply-chain attack hit a GitHub account.

A threat actor compromised the GitHub account of the person(s) maintaining tj-actions/changed files, a tool that is part of a larger collection called tj-actions, which helps automate software updates, and is reportedly used by more than 23,000 organizations.

Once in the account, the hacker silently modified the software so that instead of working as intended, it also stole sensitive information from the computers running it. Many developers apparently trusted the tool without checking for changes, executing the malicious code and exposing sensitive credentials. The report claims AWS access keys, GitHub Personal Access Tokens (PATs), npm tokens, private RSA Keys and more, were added to a plaintext log and thus exposed.

Dozens of victims

The stolen credentials could allow attackers to access private systems, steal data, or compromise the services mentioned above, which means that the effects of this attack are yet to be seen in the weeks and months to come.

GitHub addressed the incident, saying that the company and its platform were not compromised in the attack, but it still helped remedy the problem.

"Out of an abundance of caution, we suspended user accounts and removed the content in accordance with GitHub's Acceptable Use Policies," GitHub was cited saying.

"We reinstated the account and restored the content after confirming that all malicious changes have been reverted and the source of compromise has been secured."

Users should "always review GitHub Actions or any other package that they are using in their code before they update to new versions," GitHub concluded.

Ars Technica noted security researchers from Wiz already found “dozens of users” who were harmed in this attack.

“Wiz Threat Research has so far identified dozens of repositories affected by the malicious GitHub action, including repos operated by large enterprise organizations. In these repositories, the malicious payload successfully executed and caused secrets to leak in workflow logs,” they concluded

If your system is using tj-actions, make sure to inspect it thoroughly for any signs of compromise.

You might also like

• GitHub users targeted by dangerous new phishing threat
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

• Pebble is teasing a major smartwatch announcement
• A countdown timer on its website confirms there's a reveal coming on March 18
• The company has previously confirmed its iconic smartwatches could make a comeback

Just weeks after confirming its iconic smartwatch would make a return, Pebble is teasing a major announcement for March 18 with a countdown timer on its website.

9to5Google reports that the company sent an email to those who signed up for updates on the company's anticipated new project which states:

Time to get excited!

Set your alarm for March 18 (tomorrow!) at 9am PDT – 12 noon EDT – 4pm GMT – 9:30pm IST – Midnight CST – 1am JST

If you want to keep track , you can watch the countdown timer here.

Pebble teases major announcement

(Image credit: Pebble)

As to what Pebble is going to announce on March 18, it's impossible to say, but we do have some ideas based on what the company has already said this year.

In January, Google Made Pebble's source code available for download, and founder Eric Migicovsky confirmed he was working on a new smartwatch using PebbleOS soon after.

In February, he confirmed that his latest project would be "a Pebble and almost exactly as you remember it." He's also confirmed it'll feature the always-on e-paper display, long battery life, buttons, simple user experience, and hackability that made the original Pebble line so iconic.

"Please don’t get your hopes up that the new watch will have X/Y/Z new feature," he said in February. "It’s going to be a Pebble and almost exactly as you remember it, except now with open-source software that can you can modify and improve yourself. More hardware details will be shared in the future."

With Pebble's announcement just hours away, we don't have long to wait and find out what the company has in store.

You may also like

• 'It's going to be a Pebble and almost exactly as you remember it' – beloved smartwatch founder reveals major progress on new venture
• Pebble was one of the best smartwatches ever and now it's a miracle if you can get one to work
• Pebble is back! The iconic smartwatch is returning with its OG founder and 'hackable' software

Whether you’ve got the best Apple Watch on the market, a humble Apple Watch SE 2, or even an older model, watchOS is home to a plethora of options, buried in the nooks and crannies of the settings menus, that can make or break your Apple Watch experience.

There are some settings I always change when I’m using my Apple Watch, be it the notification sounds, raise-to-wake tweaks, and others. But there are also a few settings that people often overlook when using their Apple Watch.

I’ve been using Apple Watch for 10 years, both as a tech writer and an Apple employee, and these are some of the most common mistakes people make – myself very much included.

1. Not updating your fitness goals

(Image credit: Future)

Even beginner Apple Watch users will be familiar with Apple’s addictive Activity rings that track your move, exercise, and stand goals throughout the day. Closing your rings is a great way to stay healthy and active, with fun rewards and badges on offer to make taking care of yourself ever-so-slightly easier.

That being said, there’s one aspect you might have forgotten about, and that’s your fitness goals. You select your fitness goals whenever you set up a new Apple Watch for the first time.

If you don’t buy a new Apple Watch every year, or you restore your new model from a backup, there’s a chance that you haven’t looked at your goals in years, and as such it might be time to change them.

If you set relatively modest goals when you first got your Apple Watch, and find that you’re crushing them on the daily, perhaps it’s time to increase those targets and push yourself a bit more.

Suppose your move goal is a very low number of calories, or you set your stand goal to a paltry one hour because you just wanted to get on with using your Apple Watch. Why not revisit those goals and give yourself a boost? Conversely, if you're struggling to hit goals that are too ambitious, why not revise them to more modest levels to help increase your motivation?

To change your fitness goals, simply head to the Activity app on your watch, scroll down, and tap the -+ icon in the bottom right-hand corner to change your goals. Remember to apply the principles of progressive overload here. If you’re happily managing 10 minutes of exercise, maybe try 15 or 20, but don’t overdo it and go for 60-plus – even 15 is a 50% increase.

Not managing your notifications

(Image credit: Future)

Even the most basic notification management on your Apple Watch is going to save you a world of frustration. Having on-wrist dings and pings sounds great, but after 10 minutes you’re going to find Apple Watch notifications hilariously annoying.

Notifications from your Apple Watch, especially superfluous ones, are distracting and drain your battery. At the very least, use the Apple Watch Do Not Disturb to get some reprieve – just remember that this also silences your iPhone.

A more effective way to manage your notifications is to use the Focus Mode feature on your iPhone, which can help sift notifications while you’re at work, exercising, or just trying to get some personal downtime.

Finally, head to the Notifications section of the Apple Watch app on your iPhone and toggle which apps from your iPhone you want to see notifications from on your Apple Watch – I guarantee you’ll turn off half of them, and your wrist will thank you for it.

Not using your Emergency Contacts

(Image credit: Future)

Formulating this list, I settled on Emergency Contacts as my final pick with a chuckle. “Who doesn’t have Emergency Contacts set up on Apple Watch?” I scoffed to myself quietly.

Me. That's who. I didn’t have Emergency Contacts set up on my Apple Watch.

I thought I did, but I’m not sure why. Maybe I assumed it was automatic, or maybe I thought about doing it once and then never got around to it.

In your Apple Watch app head to Emergency SOS, scroll down to Emergency Contacts to check yours. If your Apple Watch automatically calls emergency services after a crash or fall, or if you do it manually by holding the Side Button, your Watch can also automatically notify relatives and friends of your plight, including information about your location.

You should absolutely check you have a list of contacts including friends and loved ones in there. You should also check that any contacts you've set up are still relevant, and that their numbers haven’t changed. The Apple Watch Emergency SOS feature is a vital safety tool even Tim Cook has used, so don’t be a plum like me.

You may also like

• How can we make wearables more sustainable, and is it even possible?
• I've been wearing an Apple Watch for 10 years – these are the 5 settings I change right out of the box
• New figures claim the smartwatch market just shrunk for the first time ever, and the Apple Watch Ultra 3 is to blame

• Researchers uncovered a brute-forcing tool called BRUTED
• It was used since 2023 against VPNs and firewalls
• BRUTED allows for automated brute-force and credential stuffing attacks

The infamous Black Basta ransomware actors created an automated framework for brute-forcing firewalls, VPNs, and other edge networking devices.

The “BRUTED” tool has apparently been in use for years now, according to cybersecurity researchers EclecticIQ, who have been sifting through the recently-leaked Black Basta chat logs, which were leaked and subsequently uploaded to a GPT for easier analysis.

Besides being used to analyze the group’s structure, organization, and activities, researchers used it to identify the tools, too. Apparently, BRUTED was in use since 2023 in large-scale credential stuffing and brute-force attacks. The endpoints being targeted include SonicWall NetExtender, Palo Alto GlobalProtect, Cisco AnyConnect, Fortinet SSL VPN, Citrix NetScaler (Citrix Gateway), Microsoft RDWeb (Remote Desktop Web Access), and WatchGuard SSL VPN.

High confidence often leads to victimization

The tool first identifies potential victims by enumerating subdomains, resolving IP addresses, and appending prefixes such as “vpn”, or “remote”. It then pulls a list of potential login credentials and combines them with locally generated guesses, executing as many requests as possible.

To narrow the list down, BRUTED extracts Common Name (CN) and Subject Alternative Names (SAN) from the SSL certificates of targeted devices, as well, the researchers said.

Finally, to remain under the radar, BRUTED uses a list of SOCKS5 proxies, although its infrastructure is apparently located in Russia.

To protect against brute-force and credential stuffing attacks, businesses should make sure all their edge devices and VPN instances have strong, unique passwords, consisting of at least eight characters, both uppercase and lowercase, numbers, and special characters. They should also enforce multi-factor authentication (MFA) on all possible accounts, and apply the zero-trust network access (ZTNA) philosophy, if possible.

Ultimately, monitoring the network for authentication attempts from unknown locations, as well as for numerous failed login attempts, is a great way to spot attacks.

Via BleepingComputer

You might also like

• Many workers are overconfident at spotting phishing attacks
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

• Google unveils new AI initiatives, including expanding UK data residency for Agentspace and AI upskilling projects
• Chirp 3 will also be coming to Vertex AI
• Follows Google’s $1 billion commitment to building a new data center in 2025

Google has announced the expansion of UK data residency for Agentspace, offering extra security by allowing British businesses to store their information locally.

Speaking at the launch in its DeepMind’s London HQ, Google Cloud CEO Thomas Kurian outlined the opportunities Google’s UK data residency extension brings for organisations that use Agentspace.

“It's an important element to give companies confidence that they can use our systems and their data will be private to their own use and that they have full control to keep the data where they need it," he noted.

Agentspace inclusion

The existing data residency programme in the UK gives Google Cloud users and customers the ability to conduct machine learning processing for Gemini 1.5 Flash, but Google has strengthened its commitment to UK AI by expanding this to include Agentspace.

“Recognizing the importance of the United Kingdom, we already have a cloud region here," Kurian added. "We're expanding our program with a new large scale data center, and we host these solutions and models here in the UK so that you can keep your data in the country and meet regulations to train and serve models right here.”

Agentspace is a business-specific AI agent platform, which combines Gemini AI, Google search, and company data all in one place, using emails, documents, tables, and even third party services like Confluence and ServiceNow within a multi-model search agent, designed to help answer complex questions and boost productivity.

That’s why its data residency is “super important”, said BT Group Chief Executive, Allison Kirkby, praising the move for offering “a bit more sovereignty and protection” for UK users.

“Because our overall mission as a company is to be the most trusted connector of people, business, and society. And considering the world we live in, certainly have customers that want to ensure that their data is safe and secure and it's on platforms that are encrypted in the right way with the right people handling it”

Skills investments

The news follows Google’s $1 billion commitment to building a new data centre, set to open later in 2025.

This comes alongside significant investment into skills and training initiatives, and a Google Cloud Startup Program, offering up to £280,000 for AI startups across the country - all part of a Google’s reinforced commitment to the UK market.

Google is looking to expand AI skill initiatives in the UK, having trained over a million students, educators, and developers in the last ten years. This includes 35 free Skills Boost credits per month for the Google Cloud Skills Boost programme, with a focus on generative AI in the workplace.

When polled, over half of TechRadar Pro readers (54%) said they use AI daily in their work, with a further 22% saying they use AI tools at least once or twice a week. This outlines just how crucial AI safety and skills training is, especially given only 17% of respondents told us they never use AI at work. If you want to take a look at our polls, or participate in the next one, you can do so by signing up to our Whatsapp here.

“We're investing heavily in startups,” said Kurian. ”For example, smaller companies who use models to run the business. There's a whole range of these that we're doing. We have programs at universities to teach computer science programs on how to use models as part of their work. There's a broad range of things.”

Over 60% of startups in the UK are already working with Google Cloud, and the “Gemini for UK” event confirmed cloud credits for UK startups, alongside technical resources and training. These startups will also gain a community of Google Cloud experts, as well as investors, partners, and other startups.

Chirp 3

Also announced at Google’s event was the introduction of Chirp 3 on Vertex AI - an audio generation model, available in 31 different languages. This ‘groundbreaking’ model will offer 248 distinct voices and eight speaker options, with detailed speech functionality that “captures the nuances of human intonation.”

Whilst impressive, this does raise some questions about its potential for harm. Chirp3 does currently come with significant restrictions in order to mitigate the risk of abuse, but Google is, "working through some of those things with [our] safety team,” Kurian confirms.

Deepfake AI calls have been on the rise thanks to the advancement of similar technologies, and although AI is frequently used in threat detection, accessible tools like these could lower the barriers for cybercriminals looking to impersonate trusted figures to trick victims into handing over personal or financial information.

Researchers recently hijacked AI agents into writing and sending phishing emails for them, proving that AI agents and tools can be a useful accessory for cyberattackers - so Google will have to ensure strict guardrails and safety controls are in place.

You might also like

• Take a look at our picks for the best AI tools around
• I've tested all the best AI agents including ChatGPT Deep Research and Gemini - these are the 5 top automated artificial intelligence tools you can try right now
• EU tech companies push for digital sovereignty, reducing reliance on US and others

• Nintendo Switch 2 likely to have AI upscaling
• A patent has been updated with new information showcasing a 540p image upscaled to 1080p output
• New info on the console is scarce but this could mean older games look better on Nintendo Switch 2

The Nintendo Switch 2 is likely to have AI Upscaling which could drastically improve performance and graphics on Nintendo's upcoming console.

A patent that previously used vague language hinting that the Switch 2 would be able to "convert images through the use of a trained neural network." has now been updated to showcase how the Switch 2 would go about upscaling 540p to 1080p.

Initially reported by Mike Odyssey on X, the updated patent all but confirms Nintendo Switch 2 to be capable of upscaling for much-improved graphics when playing games. With the ability to double resolution from 540p to 1080p, handheld games could look significantly better on the Switch 2 than its predecessor.

Backward-compatible games like Super Mario Odyssey or The Legend of Zelda: Tears of the Kingdom could run much better on this new hardware with AI upscaling. This feature would be a huge reason to upgrade for current Nintendo Switch owners.

Nintendo just released an update for Nintendo Switch 2 on AI resolution Upscaling. It's happening, folks! pic.twitter.com/6GZUEXeBliMarch 16, 2025

Better graphics, better performance

With the Nintendo Switch 2 likely to have AI upscaling similar to Nvidia's DLSS or PS5 Pro's PSSR there's a huge reason to be excited for the upcoming Nintendo Switch 2 reveal on April 2. At the time of writing, official information on the Nintendo Switch's successor is scarce following Nintendo's trailer reveal in February.

That said, we expect to hear more about performance and the console's hardware in April when Nintendo pulls the cover on one of the most-anticipated console launches of all time.

As a huge Nintendo fan, and a sucker for good performance, I think Switch titles that have been released over the last few years have struggled with the aging hardware. One of the main reasons I'm excited to upgrade to the Nintendo Switch 2 is to play my old Switch games because I hoped that Nintendo would find a way to improve them with better hardware.

Now, thanks to this patent there's a real possibility that Nintendo Switch games will look and play better on the new console. For me, even if there were no new games at launch, that would be enough to part with my hard-earned cash.

You might also like...

• Xbox is reportedly teaming up with a mystery manufacturer to launch a PC gaming handheld this year
• The next Xbox console is reportedly now 'fully in production' and targeting a 2027 release
• Lego just celebrated Mario Day in the best way possible, with an incredible Mario Kart set that's up for preorder now

Looking for a different day?

A new NYT Strands 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 Monday's puzzle instead then click here: NYT Strands hints and answers for Monday, March 17 (game #379).

Strands is the NYT's latest word game after the likes of Wordle, Spelling Bee and Connections – and it's great fun. It can be difficult, though, so read on for my Strands hints.

Want more word-based fun? Then check out my NYT Connections today and Quordle today pages for hints and answers for those games, and Marc's Wordle today page for the original viral word game.

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

NYT Strands today (game #380) - hint #1 - today's theme What is the theme of today's NYT Strands?

• Today's NYT Strands theme is… Grrr

NYT Strands today (game #380) - hint #2 - clue words

Play any of these words to unlock the in-game hints system.

• SOUTH
• REED
• ROVE
• SCAR
• SIRE
• SANK

NYT Strands today (game #380) - hint #3 - spangram letters How many letters are in today's spangram?

• Spangram has 9 letters

NYT Strands today (game #380) - hint #4 - spangram position What are two sides of the board that today's spangram touches?

First side: left, 6th row

Last side: right, 5th row

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

NYT Strands today (game #380) - the answers

(Image credit: New York Times)

The answers to today's Strands, game #380, are…

• TOUCHY
• GRUMPY
• PEEVED
• IRKED
• SURLY
• CRANKY
• VEXED
• SPANGRAM: CROSSWORD

• My rating: Easy
• My score: Perfect

The words featured in today’s search perfectly sum up my feelings when I’m struggling with Strands.

Although not today, as I cheerfully breezed through the puzzle – with the only stumbling block being when I thought CROSS was one of the answers, before seeing it was the Spangram: CROSSWORD. Top punning.

I like to think of all of these words as being generational, with babies being VEXED by the complications of the world, children PEEVED by their parents’ invasions and teenagers SURLY without really knowing why before becoming TOUCHY if they’re misunderstood.

Then, as we age, there is the slow, almost inevitable progression from feeling IRKED, then GRUMPY, before the final era of Grrr – when your loved ones forgive you but your neighbors dislike you for being CRANKY.

It’s best to accept these emotions, but the more fleeting they are the better life is.

How did you do today? Let me know in the comments below.

Yesterday's NYT Strands answers (Monday, 17 March, game #379)

• CROW
• CHIPS
• WELLS
• BEDDING
• BLUSHING
• STRICKEN
• SPANGRAM: SPOONERISMS

What is NYT Strands?

Strands is the NYT's not-so-new-any-more word game, following Wordle and Connections. It's now a fully fledged member of the NYT's games stable that has been running for a year and which can be played on the NYT Games site on desktop or mobile.

I've got a full guide to how to play NYT Strands, complete with tips for solving it, so check that out if you're struggling to beat it each day.

Looking for a different day?

A new Quordle 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 Monday's puzzle instead then click here: Quordle hints and answers for Monday, March 17 (game #1148).

Quordle was one of the original Wordle alternatives and is still going strong now more than 1,100 games later. It offers a genuine challenge, though, so read on if you need some Quordle hints today – or scroll down further for the answers.

Enjoy playing word games? You can also check out my NYT Connections today and NYT Strands today pages for hints and answers for those puzzles, while Marc's Wordle today column covers the original viral word game.

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

Quordle today (game #1149) - hint #1 - Vowels How many different vowels are in Quordle today?

• The number of different vowels in Quordle today is 4*.

* Note that by vowel we mean the five standard vowels (A, E, I, O, U), not Y (which is sometimes counted as a vowel too).

Quordle today (game #1149) - hint #2 - repeated letters Do any of today's Quordle answers contain repeated letters?

• The number of Quordle answers containing a repeated letter today is 1.

Quordle today (game #1149) - hint #3 - uncommon letters Do the letters Q, Z, X or J appear in Quordle today?

• No. None of Q, Z, X or J appear among today's Quordle answers.

Quordle today (game #1149) - hint #4 - starting letters (1) Do any of today's Quordle puzzles start with the same letter?

• The number of today's Quordle answers starting with the same letter is 0.

If you just want to know the answers at this stage, simply scroll down. If you're not ready yet then here's one more clue to make things a lot easier:

Quordle today (game #1149) - hint #5 - starting letters (2) What letters do today's Quordle answers start with?

• N

• T

• C

• S

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

Quordle today (game #1149) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle, game #1149, are…

• NAVEL
• TREAT
• COPSE
• SLINK

When I said yesterday that three of that game's answers would make good start words, I was totally wrong.

Most of my problems came from not using any words containing O or A, which left me with some tricky gaps.

SLINK and COPSE came easily, but I struggled to get TREAT and it took a great deal of gazing before I finally got NAVEL.

How did you do today? Let me know in the comments below.

Daily Sequence today (game #1149) - the answers

(Image credit: Merriam-Webster)

The answers to today's Quordle Daily Sequence, game #1149, are…

• ILIAC
• BLAST
• SPLAT
• DEBUG

Quordle answers: The past 20

• Quordle #1148, Monday 17 March: INFER, FLINT, CHUMP, SHEER
• Quordle #1147, Sunday 16 March: INPUT, RUMBA, ELUDE, BLANK
• Quordle #1146, Saturday 15 March: SPEED, CLASH, RISKY, GRATE
• Quordle #1145, Friday 14 March: BELIE, SPEAR, GROWL, ADOBE
• Quordle #1144, Thursday 13 March: AMISS, METAL, TEARY, SHORT
• Quordle #1143, Wednesday 12 March: AWASH, PUSHY, RALPH, PROOF
• Quordle #1142, Tuesday 11 March: LINGO, JUICE, SHARP, REVEL
• Quordle #1141, Monday 10 March: AVAIL, THESE, TITAN, GAUGE
• Quordle #1140, Sunday 9 March: SKILL, IDEAL, WHERE, REHAB
• Quordle #1139, Saturday 8 March: WATCH, LAPSE, WREST, HEFTY
• Quordle #1138, Friday 7 March: GLEAN, SHINY, DECRY, MANGA
• Quordle #1137, Thursday 6 March: TRICE, EIGHT, BELCH, RIPER
• Quordle #1136, Wednesday 5 March: PLEAD, TWANG, MAJOR, RISKY
• Quordle #1135, Tuesday 4 March: FLOWN, ETUDE, TOPAZ, JOKER
• Quordle #1134, Monday 3 March: FRILL, KAYAK, REBAR, WORDY
• Quordle #1133, Sunday 2 March: TUNIC, GLOSS, DEBUG, LARGE
• Quordle #1132, Saturday 1 March: DENSE, BONEY, KITTY, SMEAR
• Quordle #1131, Friday 28 February: PASTA, RAZOR, PLUMB, DROOL
• Quordle #1130, Thursday 27 February: THIEF, CHAIR, ETHER, GRIME
• Quordle #1129, Wednesday 26 February: LOVER, SPIED, VAPOR, METER

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 Monday's puzzle instead then click here: NYT Connections hints and answers for Monday, March 17 (game #645).

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 #646) - today's words

(Image credit: New York Times)

Today's NYT Connections words are…

• KINDER
• DICE
• SINGULAR
• DIE
• OXEN
• SPECIAL
• DEER
• LICE
• UNIQUE
• SHEEP
• SQUID
• ANGST
• MICE
• WURST
• REMARKABLE
• SHRIMP

NYT Connections today (game #645) - hint #1 - group hints

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

• YELLOW: One of a kind 
• GREEN: European language 
• BLUE: One sounds the same as some
• PURPLE: Some sounds different to one

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 #646) - hint #2 - group answers

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

• YELLOW: EXCEPTIONAL 
• GREEN: GERMAN WORDS 
• BLUE: PLURAL ANIMALS IDENTICAL TO THEIR SINGULAR FORMS 
• PURPLE: PLURAL WORDS THAT ARE VERY DIFFERENT FROM THEIR SINGULAR FORMS 

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

NYT Connections today (game #646) - the answers

(Image credit: New York Times)

The answers to today's Connections, game #646, are…

• YELLOW: EXCEPTIONAL REMARKABLE, SINGULAR, SPECIAL, UNIQUE
• GREEN: GERMAN WORDS ANGST, DIE, KINDER, WURST
• BLUE: PLURAL ANIMALS IDENTICAL TO THEIR SINGULAR FORMS DEER, SHEEP, SHRIMP, SQUID
• PURPLE: PLURAL WORDS THAT ARE VERY DIFFERENT FROM THEIR SINGULAR
• FORMS DICE, LICE, MICE, OXEN

• My rating: Hard
• My score: 2 mistakes

Maddening is the only way to describe today’s Connections.

I roared through the yellow and green groups, at first thinking “this couldn’t be as straightforward as just being GERMAN WORDS could it”? EXCEPTIONAL was equally unchallenging. But then I hit the wall.

Before I finally realized the connections were plurals, I initially thought it must be something unique to the animals, so I put SHRIMP, SQUID, LICE and MICE together, thinking they all had flexible bodies or something. Yes, fair enough, I was reaching.

Then, when I realized it was about linguistics, I still didn’t fully get it and included LICE instead of DEER.

As infuriating as this was (I’d argue OXEN is not that different to Ox) I admire how SINGULAR and DIE (the singular of DICE) were also in the puzzle.

How did you do today? Let me know in the comments below.

Yesterday's NYT Connections answers (Monday, 17 March, game #645)

• YELLOW: FILAMENT FIBER, STRAND, STRING, THREAD
• GREEN: UNINHIBITEDNESS ABANDON, FREEDOM, SPONTANEITY, UNRESTRAINT
• BLUE: KINDS OF TRUCKS DUMP, GARBAGE, MONSTER, PICKUP
• PURPLE: BANDS MINUS THE NUMBER FIVE BEN FOLDS, JACKSON, MAROON, MC

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.

• Seagate has announced a partnership with MiHoYo
• It has revealed the new Seagate Genshin Impact Limited Edition External SSD
• It is available now in the US, with a UK launch later this year

Storage giant Seagate has announced a partnership with MiHoYo to introduce a new Genshin Impact limited edition external SSD.

The 1TB SSD drive is an officially licensed piece of Genshin Impact merchandise and is designed to commemorate the upcoming version 5.5, or 'Day of the Flames Return', game update. It features an eye-catching purple print with some cute art of the character Cyno, plus bright matching LED lighting.

It's a great looking drive and definitely one of the coolest models I've seen from Seagate yet. With a 1TB capacity, it's far from the biggest model around but should be more than sufficient for most game libraries - especially considering its blazing fast PCIe Gen4 transfer speeds.

Image 1 of 2

A closer look at front of the new drive... (Image credit: Seagate)Image 2 of 2

and the back! (Image credit: Seagate)

The Seagate Genshin Impact Limited Edition External SSD is compatible with PC and mobile in addition to PlayStation 5, PlayStation 4, Xbox Series X, and Xbox Series S - in short, all major platforms where you are able to play Genshin Impact.

It is currently available as a Seagate web store exclusive for $139.99 in the US and is set to launch in the UK for £139.99 in May this year.

In addition to the limited edition SSD, an 18in / 45.7cm USB-C 3.2 cable is included in the box, plus a quick install guide, 3-year warranty, and 3-year Rescue Data Recovery Service.

Seagate is known for making not only some of the best Xbox Series X hard drives, but also a number of the best SSDs for PS5 and best PS5 external hard drives. While the design is definitely a winner, hopefully the performance of this new product will live up the company's lofty reputation.

You might also like...

• The best mobile controllers 2025: upgrade your portable play
• These new Genshin Impact gaming mice from Logitech G could be one seriously cute addition to your PC setup
• The best SSD in 2025: top solid-state drives for your PC

• Criminals are using stolen email addresses to distribute malicious OAuth Apps
• These apps steal sensitive data and redirect people to phishing pages
• The pages steal login credentials and deliver malware

Hackers are spoofing popular cloud and productivity apps to steal people’s Microsoft 365 login credentials and deliver malware, experts have warned.

Cybersecurity researchers Proofpoint detailed their findings in an X thread, revealing unidentified cybercriminals used compromised Office 365 accounts and email addresses belonging to charity organizations or small businesses to launch the attacks.

It is unclear what the contents of the emails are, but apparently, the goal is to get victims to install malicious Microsoft OAuth apps pretending to be Adobe Drive, Adobe Drive X, Adobe Acrobat, and DocuSign.

"Highly targeted" attacks

Those that install these apps are asked to grant specific permissions: ‘profile’, ‘email’, and ‘openid’. Alone, these aren’t that destructive, since they only grant access to the user’s name, user ID, profile picture, username, and the primary email address (no access, just information about the account). The ‘openid’ permission also allows the attackers to confirm the victim’s identity and retrieve their Microsoft account details.

While these aren’t enough to steal data or install malware, they can be used in more personalized phishing attacks, the researchers said. The campaign itself was “highly targeted”, Proofpoint said, going after organizations in different industries across the US and Europe, including government, healthcare, supply chain, and retail.

After granting these permissions, the apps redirect the victims to phishing landing pages, collecting login credentials, and distributing malware. Proofpoint could not confirm the strain of the malware being distributed this way, but stressed that the attackers used the ClickFix social engineering attack.

Nowadays, ClickFix has grown quite popular. It starts with a browser popup, informing the victim that they cannot view the contents of the web page unless they update their browser (or something similar). The popup shares steps on how to “fix” the issue, tricking the victims into downloading malware instead.

Via BleepingComputer

You might also like

• A flaw in Google OAuth system is exposing millions of users via abandoned accounts
• We've rounded up the best password managers
• Take a look at our guide to the best authenticator app

• Snap's fifth-gen Spectacles are getting a major update
• The update includes new location-based tools and better hand-tracking
• There are also new AR dog-walking and basketball-coaching apps

While Meta, Google, and Samsung take up a lot of the air in the next-gen AR glasses space with their lofty plans, they aren’t the only players – and Snap is making sure we don’t forget that, with a suite of new health-related updates and apps rolling out to its fifth-generation Snap Spectacles to celebrate six months since their launch.

The Snap Spectacles are developer-focused, so the changes focus around app-making tools, such as unlocking the ability to build new software that incorporates GPS, Global Navigation Satellite System (GNSS), and compass heading data, or is built for specific locations.

As an example of this in use, there's NavigatAR from Utopia Lab, which facilitates AR-enhanced navigation between two places, and Path Pioneer, which allows you to create customized AR walking courses to help keep people active.

If you want to see these tools used in a more finished state there’s Doggo Quest, which brings your furry friend into AR. SnapML (Snap’s machine learning) can recognize your dog and create a visual overlay that tracks metrics like the route you’ve taken and your pup’s step count – even gamifying the experience with virtual rewards.

Then there's Basketball Trainer which, as the name implies, is an AR basketball coach that can provide you with shooting drills, and track your shots using SnapML to provide you with a live score update. This tool sounds ideal for fledgling sports stars who want to improve their game between their in-person training sessions.

Beyond fitness, Snap is also adding new hand-tracking capabilities such as phone detector – so the glasses can recognize when you have a phone out – and an improved AR keyboard for more streamlined typing.

An exciting tease for the future

(Image credit: Future)

For now, Snap’s latest specs are only available through its developer program, which costs $99 plus tax per month (or €110 including VAT in the EU, around £90 / AU$190) unless you get a 50% educational discount. But that doesn’t mean we can’t get excited about the software being made for them today.

Consumer smart glasses are coming (for example, Meta’s Orion AR glasses are expected to launch in 2027, and I’m sure they won’t be alone), but it’s no use having impressive next-gen XR tech if it can’t do anything exciting.

These developer kits and early apps should mean that when the tech is rolled out to the public there will be a host of uses for our smart glasses out the box on day one that’ll make them feel like a worthwhile wearable – and not an overpriced fashion accessory.

We’ll obviously have to wait and see what Snap, Meta, Google, and the rest have up their sleeves when next-gen AR glasses make their consumer debut, but this update has me excited for an XR future which can’t come soon enough.

You might also like

• Samsung, Apple and Meta want to use OLED in their next VR headsets
• Mark Zuckerberg just teased next-gen Ray-Ban smart glasses
• Meta's new smart glasses come with some much-needed upgrades

• With Alpine’s input, the new R5 Turbo packs 540hp
• It will be one of the first production cars with in-wheel motors
• Renault’s halo model is limited to just 1,980 units

We are calling it: the award for the wildest electric vehicle of 2025 has already be scooped up by Renault’s 5 Turbo 3E - an all-electric homage to a 1980s icon that tore up race tracks and rally circuits, while simultaneously adorning the walls of hot hatch and car-tuning fanatics around the world.

But despite creating a serious amount of buzz when it was first teased at the end of last year, the full extent of its performance, innovation and utterly jaw-dropping looks have been kept under wraps until now.

We now know that it will feature an 800V electrical architecture that will allow it to charge from 15 to 80% in around 15 minutes from a 350kW charger.

Phillipe Varet, one of the chief engineers on the project, says that it will manage around 20-minutes of track time, before requiring a 20 minute charge before it will happily head back out on circuit again.

In terms of performance, the Renault 5 Turbo 3E is the dictionary definition of a pocket rocket, with the two-seat hot hatch using two innovative in-wheel motors that package most of the drivetrain inside the rear wheels.

Image 1 of 6

(Image credit: Renault)Image 2 of 6

(Image credit: Renault)Image 3 of 6

(Image credit: Renault)Image 4 of 6

(Image credit: Renault)Image 5 of 6

(Image credit: Renault)Image 6 of 6

(Image credit: Renault)

Motorcycle-style brake calipers provide the stopping power, while the twin 200kW (540hp) motors have enough power to propel the sub 1,450kg hatchback to 62mph from a standstill in just 3.5 seconds... and then on to 168mph.

But more than simply a straight-line performance pony (aren’t most EVs?), Renault has leaned on the expertise of sister company Alpine to create a true driver’s vehicle.

The carbon fibre superstructure helps keep structural rigidity high and the weight down, the battery packs are mounted as low as possible for an optimal center of gravity, while aerodynamic splitters, spoilers and side scoops help channel the air to aid downforce and improve aerodynamic efficiency.

Sneak a look through the windows and you’ll find an interior that joyously plays on Renault’s quirky legacy, yet also propels the brand into the future. There’s even a vertically-mounted handbrake that helps driver’s initiate drifts and other rallying maneuvers.

“We didn’t want to just look back and create something retro,” Gilles Vidal, Design Vice President at Renault told us at the exclusive reveal event in France. “This had to appeal to our kids as much as it did to fans of the original cars."

It nails that brief and then some, taking plenty of styling cues from the even more banzai concept vehicle of 2022, but pairing it with a modern and practical interior... if you look past the roll cage.

There is a 10.1-inch driver display and a 10.25-inch OpenR infotainment system that’s found in the standard R5 E-Tech electric hatchback, which houses a bespoke digital dial design but still offers Renault's AI assistant.

Analysis: Renault ramps up the EV excitement

(Image credit: Renault)

The French marque has already won some serious accolades for its R5 E-Tech model, which I raved about after driving.

More importantly, it has gone on to win numerous global awards for its looks, affordable price and everyday practicality, while snatching the sales crown from Tesla's Model Y in its home country of France last year.

Building on this buzz, Renault is now flexing its design and innovation muscle, proving that electric platforms have the potential to produce truly exciting and interesting vehicles that have the power to bust genres.

The Renault 5 Turbo 3E is the same length as a practical city car, but sits as wide as most modern supercars. Its innovative in-wheel electric motors provide blistering performance, yet its engineers say it can still manage an electric range of almost 250-miles and be driven to the shops every day.

In that sense, it is the epitome of the coveted hot hatch of yesteryear - a pragmatic "shopping trolley” that could also prove an absolute blast on the right roads. But, seeing as this is a demonstration of Renault's technological prowess, it packs performance figures that would make most modern sports cars weep.

(Image credit: Renault)

Alas, only 1,980 will be available when the order books open later this year, with first deliveries expected to land in 2027. It’s also going to be a lot more expensive than the standard model upon which it is loosely based.

That said, Renault’s design chief Gilles Vidal claims that his team will be on hand to assist with creating truly unique cars for customers, with completely bespoke paint jobs, ‘Gentleman Racer’ liveries and wild interior finishes up for discussion. “If you want a Hello Kitty scheme, we can help with that,” he joked.

Above all else, it feels like Renault, and its current CEO Luca de Meo, have unwavering faith in electrification, to the point where it is comfortable experimenting. Just look at the R17 concept or the Filante Record 2025 rolling laboratory as examples.

At a time when a number of manufacturers are busy backtracking on pure EV plans, it’s a bold stance.

You might also like

• I’ve driven the new Renault 5 – and it's finally the first small, fun EV for the masses
• Renault’s striking new Emblème concept car runs on hydrogen – could this signal a resurgence for the alternative fuel?
• Renault's hyper-efficient concept car swaps pedals for fighter jet controls – and it could catch on

Welcome to the live blog for Adobe Summit 2025 in Las Vegas, where I will be bringing you all the news, updates, and announcements in real time. I'm getting ready for everything to kick off on March 18, so bookmark this tab and be sure to come back to see all the biggest news and announcements as they happen.

Google’s plans to completely retire its Google Assistant software is officially under way following last week’s announcement that its AI-advanced Google Gemini will replace Google Assistant in Android smartphones by the end of 2025. This news doesn’t come as a huge shock to us, as Google has been phasing out some of its Google Assistant tasks since October 2024, however we’ve come to learn that it’s not just the best Android phones that are getting the Gemini treatment.

As well as smartphones, Google’s line of smart home assistants which includes the best smart speakers and best smart displays are also going through some changes, implying that they too will also receive a Gemini facelift. Since Gemini started supercharging Google Assistant on smart displays and speakers late last year, we could only guess that it would be a matter of time before this evolved into a complete Gemini takeover, and now it’s looking that way in light of Google’s recent decision to disable certain Google Assistant features from its smart home devices.

In order to get the ball rolling with Google’s big transition, the company is eradicating traces of Google Assistant bit by bit to make room for its more advanced Gemini software, which we guess will be fully rolled out by the end of 2025 - at least to Android smartphones. It may take a little longer for the case of Nest devices, but now that Google has removed functions such as Live Translation and Family Bell, it’s already got a leg up on making smart devices more Gemini-friendly. This isn’t to say that Gemini will have less functionality than Google Assistant, it’s simply a case of Google slowly removing one to make room for the other.

Google Photos capabilities

You’ll no longer be able to use voice commands to ask Google Assistant to favorite or share your photos, or ask specific questions regarding when and where they were taken. With that said, you’ll still be able to share them through the Google Photos app, and use your smart display to see where they were taken.

Smart display ambient screen settings

(Image credit: Google)

To enable photo frame settings or ambient screen settings on your Google smart display, you’ll have to refer to the Smart Display app to change these manually as Google Assistant will no longer support the voice command.

Live Translations

(Image credit: Google )

Since Google Assistant introduced its useful Interpreter Mode function in 2019, users have been able to translate live conversations from one language to another. This is no longer the case following Google’s decision to axe it from Assistant’s functions, however it’s still possible to translate singular words or phrases.

Birthday reminders with Routines

Routines is Google’s version of a personal task assistant which can automate tasks and actions by voice commands such as turning the lights off, playing music, or sending you notifications of important birthdays in your calendar. Google has now disabled the latter, but you can still use the Assistant voice control to set reminders about upcoming birthdays.

Family Bell announcements

Google’s Family Bell function was helpful for the whole family, allowing you to schedule reminders or announcements so that everyone in your household would be in the loop with family events.

As part of Google’s transition to integrating Gemini with smart home devices, you can no longer use Assistant to create or read your previously scheduled announcements. This was a favorite among Google smart home users, now Routines is its alternative.

Daily updates

(Image credit: Google)

This is a basic function, but a rather useful one that Google has just terminated. Just like you can set Alexa to give you daily updates, Assistant will no longer support voice requests such as ‘send me the weather everyday’. Instead, you’ll have to manually schedule updates when you create a custom Routine.

Google Assistant on car accessories

Google has disabled the ability to use Assistant on car accessories that support Bluetooth connection or have an AUX plug.

You might also like

• Multiple hands-on Google Pixel 9a videos have emerged, days ahead of the likely launch
• Google is finally rolling out a fix for broken Chromecasts – just as new bugs appear on the Chromecast with Google TV
• Google Messages could soon follow WhatsApp with an upgrade that makes it much easier to join group chats

• EU tech leaders want to move away from relying on foreign-owned infrastructure
• Establishing a Euro-stack would increase competition and innovation
• EU companies want common standards and less red tape

Several major European tech companies are pushing for greater action from the European Union to reduce the bloc’s reliance on foreign-owned infrastructure by buying and building locally.

In an open letter to the European Commission president, Ursula von der Leyen, and the EU’s digital chief, Henna Virkkunen, around 100 organizations called for greater adoption of EU-grown apps, platforms, AI models, chips, computing, storage, and connectivity.

The move comes as the Trump administration seemingly positions itself to burn an 80 year relationship that the United States has held with Europe.

Developing the ‘Euro-stack’

The European digital sovereignty movement has been gradually gaining momentum over the past few years, but has seen a massive increase in support and traction following the re-election of President Trump. Companies such as Airbus, Element, OVHCloud, Murena, Nextcloud, and Proton have all put their mark on moving towards EU digital sovereignty.

A paper [PDF] published in January 2025 by a number of prominent business heads and tech experts has given some foresight to how a “Euro stack” would work: “The objective is to reduce Europe’s current total dependency on non-European actors for service to European citizens, businesses, and institutions, to enhance security, create redundancy and resilience, improve opportunities for innovation and digital competitiveness while establishing European governance rules.”

Essentially, the European Union has become overly reliant on foreign-owned infrastructure - especially US Big Tech - and if nothing is done soon, EU countries will become subservient to foreign tech companies.

The solution therefore is to foster growth at home, with the letter stating, “Industry will invest if there are adequate demand prospects. Prioritizing areas where Europe can already deliver will be key to shifting resources fast to European suppliers, creating value and market in a virtuous circle.”

Among suggestions made on how to achieve a Euro stack are plans for the development of common standards to help EU companies compete and defend against US hyperscalers, establishing a ‘Sovereign Infrastructure Fund’ to publicly fund new digital infrastructure in the EU, and promoting innovation by cutting through the burdensome red tape of EU lawmakers.

Via TechCrunch

You might also like

• Take a look at our guide to the best cloud storage on offer
• We've listed the best antivirus around to keep you safe
• EU reveals sites for major AI factories across Europe

• Google Gemin 2.0 Flash's image generator in AI Studio can remove watermarks from images
• It seems to lack the guardrails found in the full Gemini app
• Gemini 2.0 Flash in AI Studio can also generate images of celebrities

Google Gemini 2.0 Flash's new image generation capabilities enable it to remove watermarks from images so well that it could enable large scale copyright violation and that’s very concerning for all of us.

Lots of users on the X platform have posted examples of how they’ve been able to use the AI tool to easily remove watermarks.

New skill unlocked: Gemini 2 Flash model is really awesome at removing watermarks in images! pic.twitter.com/6QIk0FlfCvMarch 15, 2025

Digital watermarks on images enable digital creators to show a preview of their work before somebody decides to purchase it, at which point the watermarks are removed. AI tools that remove watermarks on images for free are nothing new but are deeply unethical, and can land you in legal hot water if you use the stolen images they create in any way.

Other users on X have noted that Gemini 2.0 Flash accessed via AI Studio can add celebrities into images, although in testing I found the results seem a little bit random, with Gemini often refusing the request.

left: my deskright: elon musk sitting at my deskI usually tend to think worries re:photoshop etc are overblown, but goddam, the internet is about to get a lot weirder pic.twitter.com/O9gXBaWgkYMarch 14, 2025

The power of AI Studio

Google recently released a new version of its Gemini 2.0 Flash AI model that can generate its own images. This sounds slightly confusing when you realize that we’ve all been able to go to gemini.google.com or fire up the Gemini app and ask it to create an AI image for well over a year now.

However, what the mobile app and browser-based versions of Gemini do is call upon Imagen 3, Google’s flagship AI image generator, to generate the image for them. Imagen 3 will not remove watermarks from images if asked.

Recently Google has enabled Gemini 2.0 Flash to create and modify images itself. You can’t access the image generation power of Google 2.0 Flash from the regular Gemini interface, instead you need to access it via Google’s AI Studio, a developer-focussed interface for interacting with its latest AI models.

An image created using Gemini 2.0 Flash (with images) in AI Studio. (Image credit: Created with Google Gemini) Cause for concern

AI Studio is free to access and just requires a normal Google account for registration. To test it out I uploaded a watermarked image and was able to easily remove the watermark.

The results weren’t perfect - the image became slightly degraded and it added its own Gemini star logo to the bottom left corner, but what is especially concerning with the new Gemini 2.0 Flash watermark removal ability is how well it does it.

It’s clearly not the purpose that Gemini 2.0 Flash was created for, but the fact that Google hasn’t put sufficient guardrails in place with its image generator is a concerning development and something that needs addressing urgently.

TechRadar has reached out to Google for comment on this issue and will update this story when we hear back.

You may also like

• It's official: Google Assistant will be retired for phones this year, with Gemini taking over
• Now everybody gets Gems as part of Google Gemini for free, you can start making your own custom Gemini chatbots
• Gemini just got physical and you should prepare for a robot revolution

• Saber Interactive has issued a small update on its upcoming projects
• Its games are still in development, which indicates the Knights of the Old Republic remake is still being worked on
• It was initially teased all the way back in 2021

A brief flash of potentially good news for those of you excited for the Star Wars: Knights of the Old Republic remake.

As reported by IGN, Saber Interactive CCO Tim Willits has shared an update on the developers' upcoming projects via X / Twitter. Willits doesn't mention the KOTOR remake outright, but insists that the studio's upcoming projects (which would presumably include the remake) are all still in development.

"Saber Interactive is one of the largest independent developers in the world," Willits wrote. "We are working on numerous games across many different genres. Everything that we have talked about is still in development. We will share information on upcoming games when we have something cool to share."

Becoming independent from Embracer Group as of last year, Saber Interactive managed to keep hold of its projects including the Star Wars: Knights of the Old Republic remake. However, given the tumultuous nature of Embracer's acquisitions and subsequent mass layoffs, things were looking fairly grim for awhile there.

While Willits doesn't outright state it's still a go-ahead, it's possible that we'll receive a more constructive update on the remake sometime in the future. When that'll be, we're not sure, and indeed we haven't had any concrete updates since that initial 2021 teaser.

You might also like...

• I was already sold on Atelier Yumia as an RPG, but I wasn’t expecting it to have my favorite crafting system in all of gaming
• Sony has unveiled some goodies to celebrate God of War’s 20th anniversary, but it's not the remaster I was hoping for
• Silent Hill f will present players with 'a beautiful yet terrifying choice', and I can't wait to see what it is