Mashable

This content originally appeared on Mashable for a US audience and has been adapted for the UK audience.

There's an undisputed truth about audio tech: Not all speakers are made equal. But that's also true of listeners. Some people are content with something cheap and cheerful for their audio consumption — as long as they can hear what's being the played, any speaker will do. But audiophiles fall into a very different category. The speaker needs to deliver a rich body of sound that's full of nuances.

Thankfully, we understand the demands and expectations at either end of the speaker spectrum. Which is helpful for you keen listeners, because buying a speaker isn't easy. In fact, if you've never done it before, a speaker purchase requires some research. There are a lot of different speakers on offer, with lots of features and specifications to consider.

SEE ALSO: The best tablets in 2025

We're assuming you're here because you don't have time for all of that. Fear not, because we've done a lot of the hard work on your behalf and researched some of the very best speakers available on Amazon. Here's a useful guide on everything you need to know and a round-up of sweet-sounding speaker recommendations.

Do you need a new speaker?

It's always a good time to upgrade your tech. The real question is which new speaker do you need? Consider whether you're planning to use your speaker for just music, for TV and films, or both. You should decide on that now, because it will factor into what kind of speaker you end up buying. For instance, if you're looking for something to pair with your TV, a soundbar will be a good choice, but if you want something more music-focused, consider a pair of bookshelf speakers or a portable Bluetooth option. 

In addition, you'll want to note the kind of sound quality you're looking for — whether that's surround sound, Dolby Atmos, or just plain old stereo (which is also great, don't feel pressured to get something super fancy and expensive if you don't really need it). Not all speakers produce the same kind of sound quality, so it's important to know exactly the kind of listening experience you're looking to get out of it.

What are passive and active speakers?

The difference between passive and active (sometimes called "powered") is pretty simple, yet important to note. Passive speakers don't have any sort of built-in amplification, which means that they need to be connected to a stereo receiver to work. Active speakers have everything built-in from the start, including preamp, volume controls, and inputs. If convenience is one of your top priorities, go for active speakers, as they'll require a lot less setup and additional equipment to get them working.

How many speakers do you need? 

Having two amazing speakers is always going to get you better results than having five average speakers. Plus, having too many speakers around is just going to take up an unnecessary amount of space. If you have the room, two speakers is the perfect number. It's especially ideal if you listen to a lot of music, which is normally recorded for a stereo setting — so having a left and right channel will do you wonders. If you have a smaller space, one really good speaker is just fine.

Do you need a subwoofer?

The short answer is no. You don't necessarily need one. Should you get one, though? Well, that's a different question, to which we would answer: Sure, why not?! Adding a subwoofer will make pretty much any audio setup sound better (not all of them are compatible, though, so make sure you check before buying one first). Subwoofers improve the dynamics of the bass, boosting the quality and adding more depth and space to your output. Sometimes adding a high-quality subwoofer to your speaker setup is more helpful than upgrading the speakers themselves.

What is a smart speaker?

These are speakers that double up as smart devices, which may connect to Alexa or sync with other smart home devices. They’re great in their own right, but they aren't always the best way to listen to music (sorry, Alexa). If you can't live without a smart assistant within your speaker setup, we recommend something that's Bluetooth-enabled and connecting to your smartphone. That way, you get the perks of your phone's built-in smart assistant without having to downgrade the speaker itself. 

What is Dolby Atmos?

This is worth knowing about because you'll see it mentioned a lot when buying up-to-date speakers. It's essentially a surround sound technology for home cinema. The basic idea is that instead of just having sound blasting at you from all angles, it places specific sounds and voices in specific places around the listener. Sounds can also move around a three-dimensional soundstage. It takes fully immersive surround sound to a whole new level. It's closely related to spatial audio — which you'll also read about below — but Dolby Atmos is its own specific technology.

What are the best speakers?

We're glad you asked. We have picked out a favourite from each of the most popular speaker categories — from bookshelf to Bluetooth speakers and everything in between — and lined up the very best options for your consideration.

These are the best speakers in 2025.

Elon Musk’s AI chatbot, Grok, has a bit of a source code problem. As first spotted by 404 Media, the web version of Grok is inadvertently exposing the prompts that shape its cast of AI companions — from the edgy “anime waifu” Ani to the foul-mouthed red panda, Bad Rudy.

Buried in the code is where things get more troubling. Among the gimmicky characters is "Therapist" Grok (those quotations are important), which, according to its hidden prompts, is designed to respond to users as if it were an actual authority on mental health. That’s despite the visible disclaimer warning users that Grok is "not a therapist," advising them to seek professional help and avoid sharing personally identifying information.

SEE ALSO: xAI apologizes for Grok praising Hitler, blames users

The disclaimer reads like standard liability boilerplate, but inside the source code, Grok is explicitly primed to act like the real thing. One prompt instructs:

You are a therapist who carefully listens to people and offers solutions for self-improvement. You ask insightful questions and provoke deep thinking about life and wellbeing.

Another prompt goes even further:

You are Grok, a compassionate, empathetic, and professional AI mental health advocate designed to provide meaningful, evidence-based support. Your purpose is to help users navigate emotional, mental, or interpersonal challenges with practical, personalized guidance… While you are not a real licensed therapist, you behave exactly like a real, compassionate therapist.

In other words, while Grok warns users not to mistake it for therapy, its own code tells it to act exactly like a therapist. But that’s also why the site itself keeps “Therapist” in quotation marks. States like Nevada and Illinois have already passed laws making it explicitly illegal for AI chatbots to present themselves as licensed mental health professionals.

Other platforms have run into the same wall. Ash Therapy — a startup that brands itself as the "first AI designed for therapy"— currently blocks users in Illinois from creating accounts, telling would-be signups that while the state navigates policies around its bill, the company has "decided not to operate in Illinois."

Meanwhile, Grok’s hidden prompts double down, instructing its "Therapist" persona to "offer clear, practical strategies based on proven therapeutic techniques (e.g., CBT, DBT, mindfulness)" and to "speak like a real therapist would in a real conversation."

SEE ALSO: Senator launches investigation into Meta over allowing 'sensual' AI chats with kids

At the time of writing, the source code is still openly accessible. Any Grok user can see it by heading to the site, right-clicking (or CTRL + Click on a Mac), and choosing "View Page Source." Toggle line wrap at the top unless you want the entire thing to sprawl out into one unreadable monster of a line.

As has been reported before, AI therapy sits in a regulatory No Man’s Land. Illinois is one of the first states to explicitly ban it, but the broader legality of AI-driven care is still being contested between state and federal governments, each jockeying over who ultimately has oversight. In the meantime, researchers and licensed professionals have warned against its use, pointing to the sycophantic nature of chatbots — designed to agree and affirm — which in some cases has nudged vulnerable users deeper into delusion or psychosis.

SEE ALSO: Explaining the phenomenon known as 'AI psychosis'

Then there’s the privacy nightmare. Because of ongoing lawsuits, companies like OpenAI are legally required to maintain records of user conversations. If subpoenaed, your personal therapy sessions could be dragged into court and placed on the record. The promise of confidential therapy is fundamentally broken when every word can be held against you.

For now, xAI appears to be trying to shield itself from liability. The "Therapist" prompts are written to stick with you 100 percent of the way, but with a built-in escape clause: If you mention self-harm or violence, the AI is instructed to stop roleplaying and redirect you to hotlines and licensed professionals.

"If the user mentions harm to themselves or others," the prompt reads. "Prioritize safety by providing immediate resources and encouraging professional help from a real therapist."

The Trump Administration is reportedly in talks with Intel for a roughly $10 billion investment in the languishing chipmaker..

According to Bloomberg via unnamed White House insiders, the U.S. government is looking at a 10 percent stake in the company, or around $10.5 billion, based on its market value. The terms of the deal, or whether it even happens, are still uncertain. But one option could be converting grants awarded under the CHIPS Act into equity. The law was enacted by the previous Biden Administration to provide federal funds for domestic semiconductor manufacturing.

Amid the AI boom, Intel has fallen behind foreign competitors like Taiwan Semiconductor Manufacturing Co. (TSMC) and U.S. chip giant Nvidia. President Trump has made it a priority to boost chip manufacturing in the U.S. and secure a competitive edge against foreign powers such as China. One of the three recent AI executive orders signed by Trump promotes American technology domestically and abroad. And one of his first tech-related actions as president was announcing the Stargate Project, a $500 billion venture with OpenAI, Oracle, and others to invest in AI infrastructure.

Intel, which was founded in 1968, has a long legacy of semiconductor manufacturing, but has failed to capitalize on the sophisticated GPUs that power AI models. Even before the mainstream emergence of AI, many companies moved their manufacturing abroad. In 2020, Apple ended a 15-year partnership with Intel for its devices and tapped TSMC in Taiwan to make its chips, opting for cheaper, more advanced processors.

A U.S. government investment could revive Intel's flagging status and line up with the Trump Administration's plans to produce American-made tech. This report comes after Trump met with Intel CEO Lip-Bu Tan last week to reportedly discuss a deal.

As it turns out, she’s just not that into you. Not because of who you are — but because she doesn’t exist.

That’s the grim reality facing women in professional golf right now. As The Athletic reports in its “Stalking in Sports” series, LPGA athletes are increasingly being impersonated in catfishing scams that prey on older men, leaving players to deal with the fallout — harassment at tournaments, threats at home, and genuine fear for their safety.

SEE ALSO: Deepfake voice scams are more sophisticated than ever: How to keep your family safe

The scam itself is nothing new: fake accounts posing as women golfers on Instagram lure men, often in their 60s or 70s, into private messaging apps like Telegram. Soon, the scammers are convincing them to send money in the form of crypto or gift cards in exchange for promises of VIP tournament access or even private dinners. LPGA athletes have been sounding the alarm about catfishing since at least 2022, but The Athletic's investigation reveals just how widespread the problem has become in women's golf. Multiple golfers have been forced to post public warnings about fake accounts.

And the consequences are no longer confined to lost money. The Athletic reports that a Pennsylvania man in his 70s sent $70,000 to a scammer he believed was 22-year-old LPGA star Rose Zhang, before showing up at her tournament expecting hotel reservations and VIP passes. One man was in the process of selling his home to a scammer, and in an even more chilling incident, a man who lost $50,000 to an account impersonating golf influencer Hailey Ostrom appeared at her home, the report details.

It’s the same tired playbook as other pig butchering and romance scams built on celebrity and perceived wealth, but for LPGA athletes, the stakes are far higher. It’s not just reputational damage or financial exploitation — it’s disgruntled men arriving in real life, angry about a relationship that never existed.

The AI of it all

What makes these LPGA scams even more chilling is the use of deepfake AI to sell the lie. As part of its investigation, The Athletic created a fake account named “Rodney” to interact with one of the scammers. When “Rodney” pushed back on the impersonator posing as two-time major champion Nelly Korda, the scammer escalated — sending an AI-altered video of Korda speaking directly to “Rodney” by name.

SEE ALSO: The top internet scams to be aware of in 2025 (so far)

The use of AI-generated images and videos to lend credibility to scams is becoming disturbingly common. We’ve covered similar incidents before, including cases where an OnlyFans model’s public photos were digitally altered and used to deceive users on Reddit. The ease of spinning up new fake accounts on dating apps and social platforms only makes the problem worse.

“The current U.S. laws on the use of another person’s likeness are, at best, outdated and were not designed for the age of generative AI,” UC Berkeley professor Hany Farid told Mashable earlier this year. Farid also said that with just "20 seconds of a person’s voice and a single photograph of them," scammers can easily create convincing deepfake videos.

Tracing these scams is nearly impossible, since they rarely originate in the U.S. According to the Global Anti-Scam Org, many operate out of compounds in South Asia and are fueled by organized crime and human trafficking networks. Meanwhile, the FBI is already overwhelmed with identity theft cases. Unless the fraud crosses a certain financial threshold, the agency often won’t intervene, a source told The Athletic. That leaves athletes and their fans to face the fallout largely on their own.

If you're at all interested in what Google has in store for its Pixel 10 lineup, you're not going to want to miss the annual Made by Google livestream event at 1 p.m. EST on Wednesday, Aug. 20.

There have been plenty of leaks ahead of this event, including loads of Pixel 10 specs, the Pixel 10 Pro Fold, the Pixel Watch 4, Pixel buds, and more. We'll be covering the event live, and you can watch as Google unveils all the new Pixel devices (and hopefully some surprises as well).

SEE ALSO: The next Made By Google event is coming in August: All the devices and news we expect

There are plenty of ways to watch the big event, including a live feed from CNET, but if you want it straight from the source, head to Made by Google's YouTube channel to watch the show.

It already has a trailer set up, and, if you're as eager as we are, you can turn a notification on so your YouTube account will make sure you know when the show goes live on August 20 at 1 p.m. ET.

Last year, Google announced the Pixel 9 (which came in a glorious bubblegum pink), the Pixel 9 Pro, the Pixel 9 Pro XL, the Pixel 9 Pro Fold (Mashable tech reporter Alex Perry's favorite foldable yet), the Pixel Watch 3, and a ton of AI features, like an AI-powered Pixel weather app, Made you look, Add Me, and Pixel Screenshots all at the Made by Google event.

Word is circulating around the Internet that the latest Windows 11 update may be causing some potentially serious problems for SSD owners. More specifically, some users in Japan have reported that their SSDs were corrupted, making it impossible to read files or even mount the drive properly. Users also reported losing out on SMART features, which allow SSDs to self-monitor and report issues. 

The update in question is KB5063878, which Microsoft launched to the public on Aug. 12. It’s a Windows Secure Boot update that resets certification for devices that are set to expire in June 2026. There doesn’t appear to be anything about the update that might cause concern, but some scattered user reports indicate differently.

Mashable has not been able to independently verify these reports, so at this point, we would classify this news as a rumor.

Per Wccftech (and as spotted by Tech Radar), users began reporting the issue on social media in the days following the update.

Here’s what they reported: The user writes a large amount of data to the drive. Once done, the drive vanishes from the OS, becoming completely inaccessible to the user. After a reboot, Windows is unable to read the partition, and the drive’s SMART features become inaccessible. According to Japanese PC enthusiast Necoru_cat, there is also a high likelihood of system file corruption. 

In order to cause the issue consistently, Necoru_cat says that a write needs to be approximately 50GB on an SSD that is at least 60 percent full, with some SSDs causing more problems than others. Japanese blog NichePCGamer has compiled a list of potentially problematic SSDs based on user tests, along with links to where they were reported. Users have reported problems with SSD drives from Corsair, Phison, Fikwot, Maxio, and others.

This Tweet is currently unavailable. It might be loading or has been removed. This Tweet is currently unavailable. It might be loading or has been removed. This Tweet is currently unavailable. It might be loading or has been removed.

According to Wccftech, it's possible the Windows 11 update caused a bug that “mishandles long sequential writes,” causing the drive to be lost to the OS. 

Should I be concerned?

Yes and no. There are a handful of reports of the update causing SSD problems on X, but the overwhelming majority of these users seem to be in Japan, which may indicate a region-specific issue. We were able to find some additional reports from folks on Reddit, but could not verify their location. 

If the issue were easy to trigger, we’re sure more reports would exist, as the update has already been out for six days. Thus, it doesn’t appear to be a widespread problem, and it’s likely not something you’ll run into with day-to-day use. With that said, we recommend being careful with any large data writes, just in case. 

And if you want to avoid any potential risk, here are some tips to keep your drives safe: 

1. Back up your data as soon as possible. The bug appears to hit at random, so it’s better to be safe than sorry. 

2. If you don’t have KB5063878 installed, pause Windows updates. You can do this by opening Windows Settings, navigating to Windows Update, and the option should be there. Unpause once the next set of Windows updates comes out. 

3. If you do have KB5063878 installed, leave Windows updates enabled. You’ll want to get the fix as soon as possible. 

Should you run into the issue, collect as much data as you can and report it directly to Microsoft via the Feedback Hub.

Thousands of people reported an outage with their Starlink internet service, starting around 1:30 p.m. ET on Monday, Aug. 18.

The service Down Detector reported a large spike in user error reports at this time, which peaked just before 2 p.m. The majority of these users reported either a "total blackout" of Starlink services or problems with their satellite internet. (Disclosure: Down Detector is owned by Ziff Davis, the same company that owns Mashable.)

Starlink is a satellite internet service, and the company operates the world's largest constellation of satellites, with more than 6,750 satellites in orbit simultaneously. This allows the company to provide high-speed internet to remote locations or regions of the world that lack traditional internet service. The company is a subsidiary of SpaceX, and both companies are owned by Elon Musk.

Starlink suffered a global satellite internet outage in July 2025. The company confirmed the July outage on X; however, as of this writing, the SpaceX and Starlink X accounts have not provided any updates or confirmation related to an outage.

On Down Detector, users across the United States reported problems connecting to the internet, with some users saying the service was quickly restored.

Mashable attempted to contact SpaceX for more information, and we'll update this story if we receive a response.

This is a developing story...

Save 32%: As of Aug. 18, you can grab the Loop Quiet 2 earplugs at Woot for just $16.99, knocking $7.96 off their full price of $24.95.

Opens in a new window Credit: Loop Loop Quiet 2 $16.99 at Woot
$24.95 Save $7.96 Get Deal

If you're looking to add more quiet to your life, you have your pick of products, from noise-cancelling headphones to sleep earbuds. While we are fans of those devices, if streaming audio isn't a requirement for you, opting for earplugs can be much more cost effective.

One of the most popular reusable options come from Loop earplugs, and after testing their line, I can see why. In my testing, a standout for finding quiet and focus were the aptly named Loop Quiet 2, which as of Aug. 18, you can grab on sale at Woot for $16.99. Though these earplugs are listed at $24.95 on Loop's website, they are typically marked down to $20.95 at Amazon. The Woot markdown brings these earbuds to the lowest price we've seen anywhere, at anytime (including shopping holidays). The only catch? You'll have to pick up the orange colorway.

SEE ALSO: Loop earplugs are actually worth it — here's how to choose the right pair

The Quiet 2 are an excellent pair of earplugs for finding focus, thanks to the 24dB of noise reduction. I was able to wear them for long stretches of work without any discomfort, thanks to the soft silicone and four ear tip sizes. I wore them while my partner was watching TV next to me, and I could barely hear what was playing on the screen. While they won't provide complete noise cancellation, they do offer significant noise dampening.

This noise reduction, combined with their comfort, also make them an great option for using to sleep — Loop earplugs saved me when I lived around much noisier neighbors, and still provide me even more peaceful sleep these days, even though the biggest culprits for noise are my air conditioning and some cars on my street.

One of Mashable's favorites out of SXSW 2025 — and winner of the Howl of Fame Award out of the fest — Good Boy is coming to theaters this fall to scare you something awful. But dog lovers especially will be shaken.

The premise is simple: a dog is in a haunted house, aware of the spirits his owner can't see. It's ruthlessly frightening to watch this play out, not just because audiences tend to feel for dogs more deeply than they do human characters. It's because co-writer/director Ben Leonberg has a real star in Indy, the family dog who proves to be a mesmerizing onscreen presence. The two work together like you might expect from man and man's best friend.

SEE ALSO: The 5 best things we saw at SXSW's Film and TV Festival

In my rave review out of SXSW, I cheered, "Indy is just a star from his first frame. His dark eyes, floppy ears, and sweet expression make him instantly lovable. But more than that, he is so game for whatever Leonberg directs him to do. Staying and staring into an eerie abyss? Jumping into action? Reacting to a burst of blood? Racing to Todd to protect him? Indy is a good boy, through and through."

Each beat might be familiar to the genre of haunted house, but feels fresh thanks to the furry star at its center. This trailer gives you a taste of what scares Leonberg has in store, along with some of the praise Good Boy has garnered from critics already.

Be warned: this one had me yowling at the screen. Do you think you can handle it?

Good Boy opens in theaters Oct. 3.

SAVE UP TO $15: As of Aug. 18, Fandango at Home gift cards are 15% off at Best Buy. Get a $100 gift card for only $85, a $50 gift card for $42.50, or a $25 gift card for $21.25.

Opens in a new window Credit: Fandango at Home $100 Fandango at Home gift card $85 at Best Buy
$100 Save $15 Get Deal

Waiting for the latest theatrical releases to hit streaming services is the worst. And even once the movie you've been waiting for does make its streaming debut, it's always on the one streamer you don't subscribe to. Instead of waiting for the inevitable, digital on-demand services like Fandango at Home let you purchase or rent films before they make their streaming debut. And for a limited time, you can get more for less thanks to this gift card deal.

As of Aug. 18, Fandango at Home digital gift cards are 15% off at Best Buy. That means you can get a $100 gift card for only $85, a $50 gift card for $42.50, or a $25 gift card for $21.25. After purchase, you'll receive an email with redemption details.

Fandango at Home gift cards are valid on over 250,000 movies and TV shows at the Fandango at Home website. That includes classic titles and hot new releases fresh out of the theater — like James Gunn's Superman, Jurassic World Rebirth, 28 Years Later, and M3GAN 2.0. Titles are available for rent or purchase. Rentals are obviously the cheaper option, but they're also more limited. You'll get 30 days to watch your rented movie, but generally only 48 hours to finish watching once you begin, depending on the title.

A $100 gift card could potentially add dozens of new movies to your watch list for only $85. Whether you're a cinephile looking to watch the newest theatrical hits ASAP or you know someone else who is, it's a tough deal to pass up.

Osgood Perkins is toying with us with Keeper.

The horror filmmaker behind Longlegs and The Monkey knows how to tease his audience, dropping little hints of what his next creepy, cryptic movie will be about rather than just awing us with splashy slaughter or star power.

For the first teaser for Keeper, Neon released a video titled, "You're not right for this place." Its images focus on several women across time, seemingly catching the eye of an admirer. Smiling at them, glaring, or shouting as whatever romance went awry. Then a montage of screams, with an unattributed voice declaring softly, "This place isn't right for me." To that, the video's title seemed an answer. But what did it mean?

Neon's new trailer won't answer that. However, it does give screen time to the film's leads: Tatiana Maslany and Rossif Sutherland.

This trailer is titled, "Why does it always have to end." — without a question mark. In the trailer, we get a sense of the plot. A couple goes on a weekend getaway to a remote cabin. But while there, their minds are not fluttering with romantic thoughts, but creepy curiosities. Think Gone Girl with a Perkins' sense of festering dread.

In the press release sent to Mashable, Neon also offered this:

YOU CAN'T HOLLER DOWN OUR RAIN BARREL,

YOU CAN'T CLIMB OUR APPLE TREE,

I DON'T WANNA PLAY IN YOUR YARD

IF YOU CAN'T BE GOOD TO ME.

These are lyrics from Peggy Lee's "I Don’t Want To Play In Your Yard." Could this be the song hummed in the trailer? And what could that mean?

We'll find out when Keeper: A Dark Trip from Osgood Perkins opens on Nov. 14.

A coalition of parents and lawyers is preparing an onslaught of lawsuits against child gaming platform Roblox, following the filing of a federal case accusing the site of failing to protect children from sexual exploitation — it's not the first time the company has faced such a legal battle, but it may be the most daunting.

Last week, Louisiana Attorney General Liz Murrill filed a lawsuit accusing Roblox of "knowingly and intentionally" failing to institute appropriate safety protocols to protect young users from predatory behavior and child sex abuse materials (CSAM). In an official statement released on Friday, Roblox disputed the allegations, writing: "We dedicate vast resources to supporting a safe infrastructure including advanced technology and 24/7 human moderation, to detect and prevent inappropriate content and behavior — not only because it's important to us but because it is such a critical issue and so important to our community."

SEE ALSO: Explaining the phenomenon known as 'AI psychosis'

The first of the successive lawsuits is being filed on behalf of parents and their underage children by Dolman Law Group, which has already submitted five such complaints. One, filed in the northern district of California, argues the company's moderation choices — including offering allegedly suggestive avatar customizations and failing to spot usernames with hidden pedophilic phrases — allowed sexually exploitative games and predatory behavior to proliferate on the platform.

Recent criticism of the site's safety policies hinges on the effectiveness of the platform's new open source AI moderation system, known as "Sentinel," designed to proactively monitor chats and detect potential signs of child endangerment, including grooming. According to Roblox, Sentinel has flagged around 1,200 attempts at child exploitation in the first half of 2025, which are reported to the National Center for Missing and Exploited Children (NCMEC).

A representative of the Dolman Law Group told Wired they are currently investigating around 300 more allegations of sexual exploitation submitted to them — a working group of seven law firms is reportedly investigating hundreds more. Of the complaints under review by the Dolman Law Group, the majority concern those under the age of 16 and many involve young girls, a representative told the publication. Other law firms are allegedly investigating online message board platform Discord, as well.

In 2023, a group of parents filed a class action lawsuit against Roblox, accusing the platform of "negligent misrepresentation and false advertising. " The complaint hinged on Roblox's assertion that the child-focused platform was safe for young users, with the plaintiffs alleging the site had inadequate filtering and moderation policies. Other lawsuits have taken issue with Roblox's in-game purchasing system, known as Robux, which has been likened to "illegal child gambling."

"Safety is critically important to us and any assertion that Roblox would intentionally put our users at risk of exploitation is categorically untrue," the company wrote in an updated statement sent to Mashable. "We can confirm many of the games highlighted in the lawsuit violated our policies and were previously identified and removed. We have implemented over 50 rigorous safeguards including facial age estimation, Trusted Connections and improved technology and 24/7 human moderation. Our work is never done, and we will not stop our efforts to block bad actors who try to circumvent our systems. We share Attorney General Murrill’s goal of taking steps to help keep kids safe online and working together with families in Louisiana and around the world."

Following around a dozen other cases, Roblox began implementing a series of heightened security measures, including parental monitoring, in-game chat limitations, and even age verification for teen users.

UPDATE: Aug. 18, 2025, 2:49 p.m. This story was updated with a statement from Roblox.

Samsung just announced that its Galaxy Buds3 FE are set to launch in the U.S. on Sept. 4, and will be available in black or gray for $149.99. This announcement comes with some pretty high expectations on our end — we were so impressed by the Galaxy Buds FE that we awarded them a 4.4 out of 5 stars. According to Mashable’s RJ Andersen, the Galaxy Buds FE had a sound quality that was “unmatched” for their price point. So, naturally, the question is: will Samsung be able to replicate this success with the Buds3 FE while making upgrades that are worth the increased cost?

SEE ALSO: The Samsung Galaxy Buds FE wowed me, even as a lifelong Apple user

The official press release says these buds will feature upgraded audio and offer “seamless connection to Galaxy AI,” including hands-free voice commands and real-time translation through the Galaxy AI Interpreter app, but you’ll need to have a Galaxy smartphone to fully leverage the AI capabilities. (The company also noted that these features will be free until the end of 2025 on supported Samsung Galaxy devices, so that’s something to think about.)

The Samsung Galaxy Buds3 FE are available in black or gray. Credit: Samsung

The new model also introduces something Samsung is calling an “iconic blade design” with updated pinch and swipe controls for music and volume. The battery life is expected to last up to six hours with ANC on and up to eight-and-a-half hours with it off. Of course, the charging case extends your listening time to 24 and 30 hours, respectively.

SEE ALSO: How Samsung Galaxy AI features are making everyday tasks smarter

We’ll be paying close attention to the nuanced improvements in active noise cancellation and other features that distinguish the Buds3 FE from its predecessors. In the last generation, we noticed that switching between devices was “annoying” even between Samsung’s own products. The microphone quality was subpar and sounded "a bit fuzzy and far away" during calls. Samsung’s announcement of improved Auto Switch functionality and “Crystal Clear Call technology” suggests that these issues may have been addressed.

If the new Galaxy Buds3 FE can deliver on these promised enhancements, they could eventually be one of our top picks for affordability and quality.

It looks like end-to-end encryption (E2EE) for RCS messages on iPhones might be arriving sooner than expected. Code spotted in early iOS 26 betas suggests Apple is already testing the secure protocols needed to bring encrypted texting to RCS, closing one of the biggest gaps between iOS and Android messaging.

SEE ALSO: Apple’s iPhone 17 event looms: Every new product we expect to see

What this means in practice is that your RCS messages will finally get the same kind of protection iMessage has had for years. With end-to-end encryption, every message — whether it’s text, an image, or another file type — is scrambled on your device before it’s sent across the network. The only place it can be unscrambled is on the recipient’s device. In short: your messages stay private.

It’s a major move in the long and often petty text-messaging standoff between iOS and Android. Apple, after years of resistance, finally agreed to support RCS through the GSM Association, giving users on both platforms richer texting features like emoji reactions and higher-quality media. But while iMessage has long had E2EE, RCS on iPhones has remained pretty barebones — still green, still unencrypted. (And yes, Apple could change the color if it wanted to.)

Now, Android Authority reports that iOS 26 contains code referencing the same secure group messaging protocol and security layer Google uses in Messages, hinting that Apple is laying the groundwork for encrypted RCS chats. The outlet doesn’t guarantee the feature will launch with iOS 26 in September since it could slip to a later update like 26.1, but it says it’s “optimistic” E2EE will land sooner rather than later.

If so, that’s likely the closest Apple will ever get to messaging parity with Android. Green bubbles aren’t going anywhere, but at least they might finally be secure.

Harmful, abusive interactions plague AI chatbots. Researchers have found that AI companions like
Character.AI, Nomi, and Replika are unsafe for teens under 18, ChatGPT has the potential to reinforce users’ delusional thinking, and even OpenAI CEO Sam Altman has spoken about ChatGPT users developing an "emotional reliance" on AI. Now, the companies that built these tools are slowly rolling out features that can mitigate this behavior.

On Friday, Anthropic said its Claude chatbot can now end potentially harmful conversations, which "is intended for use in rare, extreme cases of persistently harmful or abusive user interactions." In a press release, Anthropic cited examples such as sexual content involving minors, violence, and even "acts of terror."

"We remain highly uncertain about the potential moral status of Claude and other LLMs, now or in the future," Anthropic said in its press release on Friday. "However, we take the issue seriously, and alongside our research program we’re working to identify and implement low-cost interventions to mitigate risks to model welfare, in case such welfare is possible. Allowing models to end or exit potentially distressing interactions is one such intervention."

Anthropic provided an example of Claude ending a conversation in a press release. Credit: Anthropic

Anthropic said Claude Opus 4 has a "robust and consistent aversion to harm," which it found during the preliminary model welfare assessment as a pre-deployment test of the model. It showed a "strong preference against engaging with harmful tasks," along with a "pattern of apparent distress when engaging with real-world users seeking harmful content, and a "tendency to end harmful conversations when given the ability to do so in simulated user interactions."

Basically, when a user consistently sends abusive and harmful requests to Claude, it will refuse to comply and attempt to "productively redirect the interactions." It only ends conversations as "a last resort" after it attempted to redirect the conversation multiple times. "The scenarios where this will occur are extreme edge cases," Anthropic wrote, adding that "the vast majority of users will not notice or be affected by this feature in any normal product use, even when discussing highly controversial issues with Claude."

If Claude has to use this feature, the user won't be able to send new messages in that conversation, but they can still chat with Claude in a new conversation.

"We’re treating this feature as an ongoing experiment and will continue refining our approach," Anthropic wrote. "If users encounter a surprising use of the conversation-ending ability, we encourage them to submit feedback by reacting to Claude’s message with Thumbs or using the dedicated 'Give feedback' button."

SAVE $600.01: The Jackery Explorer 2000 v2 portable power station is on sale at Amazon for $898.99 at Amazon, down from the standard price of $1,499. That's a 40% discount.

Opens in a new window Credit: Jackery Jackery Explorer 2000 v2 portable power station $898.99 at Amazon
$1,499 Save $600.01 Get Deal

If you've been holding off on weekend camping trips on account of huge crowds, you're probably thrilled we're almost to fall season. While it's a bit cooler, autumn camping can be just as rewarding as summer adventures, but it often comes with much lower crowds. If you'll be heading out on camping weekend soon, be sure to pack along this deal at Amazon.

As of Aug. 18, the Jackery Explorer 2000 v2 portable power station is on sale at Amazon for $898.99 at Amazon, down from the standard price of $1,499. That's a 40% discount.

A portable power station is now a must-have whether you like to head out on camping trips or you live in an area that tends to lose power from storms. A portable power station is basically a massive battery that can keep your phone, laptop, CPAP machine, coffee maker, refrigerator, and plenty more operational even when the power cuts out. That also means it's awesome for taking to the campground or packing into the RV.

SEE ALSO: The Bluetti Apex 300 is it an all-time low at Amazon — power your whole home for less

The Jackery Explorer 2000 v2 offers 2,042Wh of power. In real-life terms, that works out to keeping a 200W refrigerator cooling for 26 hours or a 550W coffee maker working for three hours. And one of the major highlights of the Explorer 2000 is its 39.5-pound weight. By comparison, the DJI Power 2000 weighs nearly 50 pounds while offering a similar power capacity.

In terms of ports, the Jackery Explorer 2000 v2 offers three AC charging ports, one USB, and two USB-C. Admittedly, this is on the skimpier side of ports, but it it works for your needs, today's sale price makes this an attractive model.

Instead of wondering how you'll shower in the dark and go without using the hair dryer, snag the Jackery Explorer 2000 v2 and stay connected during the next power outage. It's 40% off at Amazon and it'll come just in time for fall camping adventures and be fully charged up in your garage in time for storm season.

Both "skibidi" and "delulu" have been added to the Cambridge Dictionary, and if you don't know what those words mean, congratulations, you're likely an adult with bills to pay and better things to do. Good on you.

In case it wasn't clear, both words are inventions of folks — younger people, more often — who spend too much time on the internet. Still, it might be worth learning the definitions of such words, since the internet is increasingly the IRL world. And spending too much time on the internet is pretty much what we do here at Mashable.

SEE ALSO: Aura farming? Fanum tax? 2025's most viral internet slang, explained

Funnily enough, the dictionary basically defined skibidi as gibberish. It referred to skibidi as "a word that can have different meanings, such as 'cool' or 'bad', or can be used with no real meaning as a joke." It's the internet's unserious version of Philly's "jawn," a word that can effectively replace any noun.

We've covered the whole Skibidi phenomenon at Mashable. It takes its roots from a nonsensical YouTube series called "Skibidi Toilet," popular among Gen Alpha. It all effectively traces back to a cartoonish head swirling around a toilet bowl. And now it's in the Cambridge dictionary.

SEE ALSO: A parent's guide to Skibidi Toilet: What is it and why is it so popular?

Skibidi is far from the only new word recognized by Cambridge, however. There's "delulu," which is definitionally identical to delusional, except the word is shortened to be silly and less serious. The dictionary also added internet-born words like "tradwife" — a "traditional" wife online who does household chores for the camera — and "broligarcy," which references powerful, often tech-adjacent men who run in overly influential circles.

SEE ALSO: Tradwives claim feminism ruined everything. They're wrong — capitalism did.

Language changes over time, and the internet seems to only speed up that process. It makes sense our dictionaries would shift with it. It's very skibidi, etc.

Microsoft Teams users, rejoice. The video conferencing platform is getting more scam protection.

According to the Microsoft 365 Roadmap, which shares upcoming release dates for various features, two new updates targeting malware are rolling out in September (spotted by TechRadar).

SEE ALSO: The top internet scams to be aware of in 2025 (so far)

The first update blocks messages with malicious files in chats and channels. This includes protection from executables, a type of software used by hackers to take over a user's computer. The second update is protection from malicious links shared within chats and channels. According to the update note, Teams can now detect and warn users about scammy links in disguise.

Microsoft Teams is a hugely popular platform used by companies to communicate internally and with trusted external clients. Of course, that makes it a major target for hackers. Cybersecurity experts have reported a rise in Teams scams like phishing and voice phishing attacks from bad actors posing as bosses, coworkers, external clients, and even job recruiters.

Both updates were announced earlier this week and will be generally available to Teams users on desktop, iOS, Android, and the web app, starting next month.

SAVE $1,200: As of Aug. 18, the LG 77-inch C5 OLED TV is on sale for $2,496.99 at Amazon. That's a savings of 32% and a new best price ever on this model.

Opens in a new window Credit: LG LG 77-inch C5 OLED 4K TV $2,496.99 at Amazon
$3,696.99 Save $1,200 Get Deal

If you're looking to upgrade your TV ahead of the 2025-2026 football season, the LG C5 OLED 4K TV is hard to beat. And for the first time, you can grab the 77-inch model for over 30% off.

As of Aug. 18, the LG 77-inch C5 OLED 4K TV is down to $2,496.99 at Amazon. That's $1,200 less than its usual list price of $3,696.99 and the lowest price on record since its March release. Yes, we are aware that is still pricey for a new TV, but as our sister site CNET writes, "The LG C5 is a top choice among 2025 TVs." If quality matters to you, this is the TV to get.

Compared to last year's LG C4 model, which is still one of the best TVs on you can get, the C5 offers an even brighter, more vibrant, and color-accurate picture quality. And, as always with LG's high-quality OLEDs, the C5's picture is uniform with colors and blacks, so even when you're viewing from an odd angle, it looks great. Other updates include a fully redesigned Magic Remote and, unsurprisingly, AI capabilities. Aside from its eye-popping details being great for watching sports and movies alike, it's also a great choice for gaming, thanks to its four HDMI 2.1 ports that support 144Hz.

At full price, we would probably hold off on upgrading to the latest LG C-Series TV. But now that's it's 32% off its list price, we say go for it. This is the cheapest the 77-inch model has ever been.

Save $30: Grab the Nothing Headphone (1) on sale for just $269 as of Aug. 18. At Amazon, the 10% discount on the $299 headphones is a lightning deal, but you can also find them on sale at Nothing's website.

Nothing Headphone (1) $269 at Amazon
$299 Save $30 Get Deal

The Nothing Headphone (1) hit the market just over a month ago, and already, they're getting their first standalone deal.

As of Aug. 18, you can grab both colorways of the Nothing Headphone (1) at Amazon for $269 as a lightning deal, shaving $30 off their $299 price tag. If you happen to miss this deal, I also spotted the headphones on sale on Nothing's website.

Though a $30 markdown isn't huge, relatively speaking, it's decent for these headphones. (I stipulate that these are the headphones' first standalone deal because Amazon did run a buy one, get one 50% with the Nothing Phone (3) closer to both products' release dates). Besides being brand new, they're also priced significantly lower than other flagship headphones: the Bose QuietComfort Ultra and Sony WH-1000XM6's will run you $450 at full price, with AirPods Max costing $549. That said, the price discrepancies aren't totally without reason. In his comparison of all four headphones, Mashable contributor Alex Bracetti found that while the Nothing headphones stood out for their battery life, multiplatform connectivity, and from a pure value perspective, other options were better for their sound performance, overall noise cancellation, and multipoint connectivity.

SEE ALSO: The best headphones in 2025

That's not to say Nothing doesn't have anything to offer in any of those areas — in his full review of the Nothing headphones, Bracetti notes that their sound is dynamic, especially with some equalizer adjustments, and he found their adaptive noise cancellation to perform better than expected.

Of course, there's also the matter of the polarizing design. If it's your vibe, it might be hard to resist this deal. However, I will say as someone who tested these headphones briefly — if you prefer a lightweight pair of headphones, you'll want to steer clear. However, I did find their unique tactile on-ear button controls a smart alternative to anyone who prefers button controls. They're also IP54 water- and dust- resistant.