Feed aggregator

• Anthropic has developed an AI-powered tool that detects and blocks attempts to ask AI chatbots for nuclear weapons design
• The company worked with the U.S. Department of Energy to ensure the AI could identify such attempts
• Anthropic claims it spots dangerous nuclear-related prompts with 96% accuracy and has already proven effective on Claude

If you’re the type of person who asks Claude how to make a sandwich, you’re fine. If you’re the type of person who asks the AI chatbot how to build a nuclear bomb, you'll not only fail to get any blueprints, you might also face some pointed questions of your own. That's thanks to Anthropic's newly deployed detector of problematic nuclear prompts.

Like other systems for spotting queries Claude shouldn't respond to, the new classifier scans user conversations, in this case flagging any that veer into “how to build a nuclear weapon” territory. Anthropic built the classification feature in a partnership with the U.S. Department of Energy’s National Nuclear Security Administration (NNSA), giving it all the information it needs to determine whether someone is just asking about how such bombs work or if they're looking for blueprints. It's performed with 96% accuracy in tests.

Though it might seem over-the-top, Anthropic sees the issue as more than merely hypothetical. The chance that powerful AI models may have access to sensitive technical documents and could pass along a guide to building something like a nuclear bomb worries federal security agencies. Even if Claude and other AI chatbots block the most obvious attempts, innocent-seeming questions could in fact be veiled attempts at crowdsourcing weapons design. The new AI chatbot generations might help even if it's not what their developers intend.

The classifier works by drawing a distinction between benign nuclear content, asking about nuclear propulsion, for instance, and the kind of content that could be turned to malicious use. Human moderators might struggle to keep up with any gray areas at the scale AI chatbots operate, but with proper training, Anthropic and the NNSA believe the AI could police itself. Anthropic claims its classifier is already catching real-world misuse attempts in conversations with Claude.

Nuclear AI safety

Nuclear weapons in particular represent a uniquely tricky problem, according to Anthropic and its partners at the DoE. The same foundational knowledge that powers legitimate reactor science can, if slightly twisted, provide the blueprint for annihilation. The arrangement between Anthropic and the NNSA could catch deliberate and accidental disclosures, and set up a standard to prevent AI from being used to help make other weapons, too. Anthropic plans to share its approach with the Frontier Model Forum AI safety consortium.

The narrowly tailored filter is aimed at making sure users can still learn about nuclear science and related topics. You still get to ask about how nuclear medicine works, or whether thorium is a safer fuel than uranium.

What the classifier attempts to circumvent are attempts to turn your home into a bomb lab with a few clever prompts. Normally, it would be questionable if an AI company could thread that needle, but the expertise of the NNSA should make the classifier different from a generic content moderation system. It understands the difference between “explain fission” and “give me a step-by-step plan for uranium enrichment using garage supplies.”

This doesn’t mean Claude was previously helping users design bombs. But it could help forestall any attempt to do so. Stick to asking about the way radiation can cure diseases or ask for creative sandwich ideas, not bomb blueprints.

You might also like

• Has ChatGPT-5's cold tone made you want to try alternative AIs? Claude just added a new memory feature
• You don’t have to explain everything to Claude anymore – it’s finally in your apps
• How Claude’s 3.7's new ‘extended' thinking compares to ChatGPT o1's reasoning

Siri's long-delayed overhaul could end up powered by Google's Gemini AI, a move that shows how urgently Apple is trying to close the gap with rivals.

• Bloomberg is reporting that Apple is looking at Gemini to power Siri
• Apple Intelligence's best bits are still delayed until possibly 2026
• No confirmation from either company, and the report put discussions at the earliest stages

Apple's efforts to deliver the smarter Siri and full Apple Intelligence we were promised "in the coming year" might get a boost from an unlikely third party if Bloomberg's latest report is true. The iPhone maker is reportedly in early-stage exploratory talks about integrating Gemini in Siri.

There aren't many details beyond that, though Bloomberg's Mark Gurman contends that the shift to these Google chats happened after Apple couldn't reach financial terms with Anthropic (maker of Claude AI).

The possibility of Apple using Gemini's much more accomplished generative AI and one of its models (Gemini Pro, Flash, Lite?) to bring the conversational intelligence lacking in Siri would immediately transform Apple's nearly 15-year-old digital assistant into a more able AI tool, but it would also mean that Apple is ceding control in what is a key digital arms race.

How we got here

While working with third parties has always been a part of Apple Intelligence's strategy, Apple's CEO Tim Cook and the company's development leadership have never mentioned ingesting someone else's generative AI models. It's also a fact that Apple Intelligence's rollout has not gone exactly according to plan.

(Image credit: Shutterstock)

In TechRadar's conversation at WWDC 2025 with Apple's Senior Vice President of Software Engineering Craig Federighi, who is also now running Apple's AI development efforts, he explained why the company hadn't delivered full Apple Intelligence and a smarter Siri on time. After struggling to get V1 architecture working as they wanted to, Apple had a decision to make:

"...fundamentally, we found that the limitations of the V1 architecture weren't getting us to the quality level that we knew our customers needed and expected." He added, "As soon as we realized that [...] we let the world know that we weren't going to be able to put that out, and we were going to keep working on really shifting to the new architecture and releasing something."

Gurman, though, contends that Apple is still not fully committed to using its own architecture and models and will soon make the decision about whether or not to outsource to a third party like Google for at least some of the necessary intelligence. Again, the discussions he describes are in the earliest stages. And whatever comes of them, assuming they exist, it's unlikely they will have any impact on the upcoming release of iOS 26, which features a smattering of Apple Intelligence updates but virtually none to Siri.

(Image credit: Shutterstock/rafapress)Far from strangers

Apple and Google are already search partners (Google is Safari's default search engine), and in Apple's Visual Intelligence, where you can choose to use Google to search on captured images (or you can ask OpenAI's ChatGPT about them).

Still, Gemini inside Siri would mark a major turning point for Apple and an admission that it's simply not up to the task of competing in the AI sphere, at least not at the level of an OpenAI, Anthropic, Perplexity, or Google.

This approach, though, is not unheard of; Microsoft's Copilot is essentially a reskinnning of ChatGPT (though there are questions if Microsoft will continue getting access to OpenAI's best models).

Even so, Apple put a lot of effort and marketing into Apple Intelligence. The question is, can it still be called that if a big chunk of it is powered by Google?

We contacted Apple and Google for comment. Google had no comment. We'll update this story if and when Apple replies.

You might also like

• This is what really happened with Siri and Apple Intelligence, according to Apple
• I spoke to Apple’s software engineering VP for the inside story on how iPadOS 26 finally became a real Mac alternative
• 15 things we learned at the Apple WWDC 2025 keynote

• Bloomberg is reporting that Apple is looking at Gemini to power Siri
• Apple Intelligence's best bits are still delayed until possibly 2026
• No confirmation from either company, and the report put discussions at the earliest stages

Apple's efforts to deliver the smarter Siri and full Apple Intelligence we were promised "in the coming year" might get a boost from an unlikely third party if Bloomberg's latest report is true. The iPhone maker is reportedly in early-stage exploratory talks about integrating Gemini in Siri.

There aren't many details beyond that, though Bloomberg's Mark Gurman contends that the shift to these Google chats happened after Apple couldn't reach financial terms with Anthropic (maker of Claude AI).

The possibility of Apple using Gemini's much more accomplished generative AI and one of its models (Gemini Pro, Flash, Lite?) to bring the conversational intelligence lacking in Siri would immediately transform Apple's nearly 15-year-old digital assistant into a more able AI tool, but it would also mean that Apple is ceding control in what is a key digital arms race.

How we got here

While working with third parties has always been a part of Apple Intelligence's strategy, Apple's CEO Tim Cook and the company's development leadership have never mentioned ingesting someone else's generative AI models. It's also a fact that Apple Intelligence's rollout has not gone exactly according to plan.

(Image credit: Shutterstock)

In TechRadar's conversation at WWDC 2025 with Apple's Senior Vice President of Software Engineering Craig Federighi, who is also now running Apple's AI development efforts, he explained why the company hadn't delivered full Apple Intelligence and a smarter Siri on time. After struggling to get V1 architecture working as they wanted to, Apple had a decision to make:

"...fundamentally, we found that the limitations of the V1 architecture weren't getting us to the quality level that we knew our customers needed and expected." He added, "As soon as we realized that [...] we let the world know that we weren't going to be able to put that out, and we were going to keep working on really shifting to the new architecture and releasing something."

Gurman, though, contends that Apple is still not fully committed to using its own architecture and models and will soon make the decision about whether or not to outsource to a third party like Google for at least some of the necessary intelligence. Again, the discussions he describes are in the earliest stages. And whatever comes of them, assuming they exist, it's unlikely they will have any impact on the upcoming release of iOS 26, which features a smattering of Apple Intelligence updates but virtually none to Siri.

(Image credit: Shutterstock/rafapress)Far from strangers

Apple and Google are already search partners (Google is Safari's default search engine), and in Apple's Visual Intelligence, where you can choose to use Google to search on captured images (or you can ask OpenAI's ChatGPT about them).

Still, Gemini inside Siri would mark a major turning point for Apple and an admission that it's simply not up to the task of competing in the AI sphere, at least not at the level of an OpenAI, Anthropic, Perplexity, or Google.

This approach, though, is not unheard of; Microsoft's Copilot is essentially a reskinnning of ChatGPT (though there are questions if Microsoft will continue getting access to OpenAI's best models).

Even so, Apple put a lot of effort and marketing into Apple Intelligence. The question is, can it still be called that if a big chunk of it is powered by Google?

We contacted Apple and Google for comment. Google had no comment. We'll update this story if and when Apple replies.

You might also like

• This is what really happened with Siri and Apple Intelligence, according to Apple
• I spoke to Apple’s software engineering VP for the inside story on how iPadOS 26 finally became a real Mac alternative
• 15 things we learned at the Apple WWDC 2025 keynote

Before buying supplements, give these 11 vitamin-D-rich foods a try.

DCI Jessica James and DI Sunny Khan return for more crime solving on the streets of London.

This month's departures from HBO Max have a color in common: red. It's the color of Hellboy and the color of Mars – and it's also the color of that famous scene in the horror classic Carrie.

My three catch-em-while-you-can recommendations for HBO Max this month have something else in common too: stunning central performances. Ron Perlman is a wisecracking delight as a hellishly powerful superhero, adding a very welcome dose of grumpiness to a genre that was starting to feel somewhat stale. Matt Damon is utterly believable and completely compelling as a scientist stranded millions of miles from home. And Sissy Spacek in Carrie is truly exceptional, delivering a performance that's heartbreakingly fragile and truly terrifying.

These are very different movies, but they're all exceptional. If you've seen them already they're well worth revisiting. And if you haven't, you're in for a movie masterclass on one of the best streaming services.

Carrie

Two Carries are leaving HBO soon: the 1976 original and the 2013 remake. The older film is vastly superior to the newer one – the remake scored just 51% with the critics on Rotten Tomatoes and has variously been called "remarkably redundant", "terrible pointless junk" and "one of the worst remakes ever made". But the original film based on Stephen King's horror classic is tremendous, with an astonishing central performance by Sissy Spacek as the titular teen who starts to suspect she has supernatural powers. It's currently sitting with a whopping 94% rating from the critics.

Carrie "is a terrifying lyrical thriller," legendary New Yorker critic Pauline Kael wrote. "The director, Brian De Palma, has mastered a teasing style – a perverse mixture of comedy and horror and tension." Looking back from the 2020s, Total Film wrote: "Brian De Palma transcends the pulpy horror feel by emphasizing the awakening-sexuality metaphor, and using some glorious trickery," while The Fright File called it "One of cinema's ultimate operatic teenage melodramas. I have seen "Carrie" more times than I can count, and yet it never loses its uncommon heartbreak and blood-curdling dramatic power."

The Martian

Matt Damon spent a lot of time in spacesuits in the 2010s: there was (mild spoiler alert) Bad Space Matt in Interstellar, and Good Space Matt in this impressive solo performance. Damon is Mark Watney, left behind on Mars after a fierce storm leads his fellow explorers to think he's dead and leave the red planet without him. But he's not dead, and he'd really like to get home.

The 91% critic rating is well deserved. Empire Magazine gave the film four stars: "Instantly joining E. T. and Bruce Dern’s Freeman Lowell (Silent Running) in the pantheon of cinema’s greatest space gardeners, Damon’s Watney is the actor at his most engaging, by turns flip and desperate... The Martian mixes smarts, laughs, weird character bits and tension on a huge canvas. The result is (Ridley) Scott’s most purely enjoyable film for ages."

Hellboy

Like Carrie, there's more than one Hellboy movie and the original is vastly better than the remake: the 2019 reboot of Hellboy got a frankly embarrassing 17% critic rating. That's partly because it didn't have Guillermo Del Toro in the director's chair or Ron Perlman in the Hellboy prosthetics. The film is "a unique romp," The New Yorker said, "with an exciting yet vulnerable superhero at the center who just happens to be the spawn of Satan."

NPR raved about it too. "Anyone can send an immense, computer-generated vegetable monster rampaging at the base of the Brooklyn Bridge, but it takes a special kind of imagination to do it in a way that's thrilling, emotionally complex, and rapturously beautiful all at once." Time Out agreed. "Del Toro, in love with his source but never overawed by it, keeps things moving; Perlman ties it together with some of the driest witticisms this side of Indiana Jones."

You may also like

• When is James Gunn's Superman movie coming to HBO Max and other streaming services?
• IT: Welcome to Derry – everything you need to know about the new Max horror series
• Peacemaker season 2: everything we know so far about the HBO Max show's return (release date, trailer, cast, and plot)

This month's departures from HBO Max have a color in common: red. It's the color of Hellboy and the color of Mars – and it's also the color of that famous scene in the horror classic Carrie.

My three catch-em-while-you-can recommendations for HBO Max this month have something else in common too: stunning central performances. Ron Perlman is a wisecracking delight as a hellishly powerful superhero, adding a very welcome dose of grumpiness to a genre that was starting to feel somewhat stale. Matt Damon is utterly believable and completely compelling as a scientist stranded millions of miles from home. And Sissy Spacek in Carrie is truly exceptional, delivering a performance that's heartbreakingly fragile and truly terrifying.

These are very different movies, but they're all exceptional. If you've seen them already they're well worth revisiting. And if you haven't, you're in for a movie masterclass on one of the best streaming services.

Carrie

Two Carries are leaving HBO soon: the 1976 original and the 2013 remake. The older film is vastly superior to the newer one – the remake scored just 51% with the critics on Rotten Tomatoes and has variously been called "remarkably redundant", "terrible pointless junk" and "one of the worst remakes ever made". But the original film based on Stephen King's horror classic is tremendous, with an astonishing central performance by Sissy Spacek as the titular teen who starts to suspect she has supernatural powers. It's currently sitting with a whopping 94% rating from the critics.

Carrie "is a terrifying lyrical thriller," legendary New Yorker critic Pauline Kael wrote. "The director, Brian De Palma, has mastered a teasing style – a perverse mixture of comedy and horror and tension." Looking back from the 2020s, Total Film wrote: "Brian De Palma transcends the pulpy horror feel by emphasizing the awakening-sexuality metaphor, and using some glorious trickery," while The Fright File called it "One of cinema's ultimate operatic teenage melodramas. I have seen "Carrie" more times than I can count, and yet it never loses its uncommon heartbreak and blood-curdling dramatic power."

The Martian

Matt Damon spent a lot of time in spacesuits in the 2010s: there was (mild spoiler alert) Bad Space Matt in Interstellar, and Good Space Matt in this impressive solo performance. Damon is Mark Watney, left behind on Mars after a fierce storm leads his fellow explorers to think he's dead and leave the red planet without him. But he's not dead, and he'd really like to get home.

The 91% critic rating is well deserved. Empire Magazine gave the film four stars: "Instantly joining E. T. and Bruce Dern’s Freeman Lowell (Silent Running) in the pantheon of cinema’s greatest space gardeners, Damon’s Watney is the actor at his most engaging, by turns flip and desperate... The Martian mixes smarts, laughs, weird character bits and tension on a huge canvas. The result is (Ridley) Scott’s most purely enjoyable film for ages."

Hellboy

Like Carrie, there's more than one Hellboy movie and the original is vastly better than the remake: the 2019 reboot of Hellboy got a frankly embarrassing 17% critic rating. That's partly because it didn't have Guillermo Del Toro in the director's chair or Ron Perlman in the Hellboy prosthetics. The film is "a unique romp," The New Yorker said, "with an exciting yet vulnerable superhero at the center who just happens to be the spawn of Satan."

NPR raved about it too. "Anyone can send an immense, computer-generated vegetable monster rampaging at the base of the Brooklyn Bridge, but it takes a special kind of imagination to do it in a way that's thrilling, emotionally complex, and rapturously beautiful all at once." Time Out agreed. "Del Toro, in love with his source but never overawed by it, keeps things moving; Perlman ties it together with some of the driest witticisms this side of Indiana Jones."

You may also like

• When is James Gunn's Superman movie coming to HBO Max and other streaming services?
• IT: Welcome to Derry – everything you need to know about the new Max horror series
• Peacemaker season 2: everything we know so far about the HBO Max show's return (release date, trailer, cast, and plot)

• More range, faster charging and a lower price
• NACS support means it can top up at Tesla Supercharger stations
• The Leaf arrives at a time when demand for EVs is waning in the US

Nissan has revealed pricing for its US-bound Nissan Leaf model and it claims it will be the cheapest EV on sale when it hits roads in 2026.

The updated Leaf, which the Japanese marque hopes will prove even more popular than the ground-breaking original, will start at $29,990 for the Leaf S+ and rise to $38,990 for the top spec Platinum+ trim, which manages 259 miles on a single charge.

The entry-level model is around $3,000 cheaper than the 2011 original and undercuts the outgoing 2025 version, but Nissan will also reveal pricing for the cheapest S model later this year, which could well start at under $28,000.

Redesigned from the ground up and sharing its platform with the Nissan Ariya, the US-spec Leaf S+ features a 75kWh battery pack that is capable of 303 miles on a single charge. The outgoing 2025 model managed a max range of 212 miles.

Despite now adopting the more popular SUV/crossover body shape, the new Leaf is actually a bit shorter than the outgoing hatchback and only 10mm taller, but engineers have somehow managed to declutter to the interior so there is more space to comfortably transport passengers.

Tesla’s North American Charging Standard (NACS) charging port has been added, allowing users to gain access to the vast Supercharger network. Charging from 10% to 80% takes around 35 minutes from the faster chargers.

Other notable new features include an electronically-dimming panoramic roof, which Nissan says is a first in the segment, a Google-based infotainment system and advanced camera technology that provides a 360-degree view of the vehicle and offers an 'Invisible Hood' view to make parking easier.

Nissan needs a big hitImage 1 of 6

(Image credit: Nissan)Image 2 of 6

(Image credit: Nissan)Image 3 of 6

(Image credit: Nissan)Image 4 of 6

(Image credit: Nissan)Image 5 of 6

(Image credit: Nissan)Image 6 of 6

(Image credit: Nissan)

Despite rapid growth over the past few years, EV sales have slowed in the US in recent months, with Inside EVs reporting that America’s EV market share dropped from 7.4% to 6.6% in April of this year.

Consumer confidence has been shaken by the US government’s decision to remove subsidies, while funding for EV-related industries continues to be attacked. The support to ensure the technology goes mainstream just isn’t in place.

Nissan is also facing a crisis of its own, as its share price continues to tumble due to the continuing losses it has been posting. As a result, it has already announced deep cuts to the workforce and the closure of several plants.

The new Leaf needs to be a big hit in the US, while the upcoming all-electric Micra also hopes to have a successful run in Europe – seeing as it shares most of its components with the hugely popular Renault 5 E-Tech.

Priced as it is, the new Nissan Leaf comes about as close as the US will get to the mythical $25,000 EV (with a useable range) that the internet has been pining for.

Seeing as Tesla looks to have killed that idea, Nissan might be in the perfect position win over new customers.

You might also like

• Let the cat out of the bag there’: Elon Musk says the new, affordable Tesla coming in 2025 will just be a cheaper Model Y
• Volkswagen wants you to pay a Netflix-style subscription to unlock extra EV power – and that could be a slippery slope
• Honda's tiny EV makes official debut – and the impossibly cute microcar can power your home too

There are some very interesting Philips Hue products on the way, including new Gradient Strips, Festavia string lights, and a Philips Hue Bridge Pro that can support up to 150 devices, but one rumored release has had me scratching my head - Philips Hue Essential.

These appear to be slightly lower-spec versions of standard Hue smart light bulbs, and will presumably have a more modest price tag to match – but why? Who, exactly, is Philips Hue Essentials for?

The budget end of the smart lighting market is already dominated by the likes of Govee, Ikea, and WiZ – and that’s where things get interesting. You see, WiZ lights are made by a company called Signify – the same company that makes Philips Hue products under license. Why would it pit its two brands against one another with a low-cost version of Hue?

WiZ smart lights (such as this WiZ Gradient Floor Lamp) are made by the same parent company as Philips Hue products, but for different users (Image credit: Abigail Shannon)

Well, it isn’t – not quite. Aside from price, one of the key differences between WiZ and Philips Hue is connectivity. WiZ lights connect directly to Wi-Fi, meaning there’s no need for a hub, and they’re easier to set up. They’re a good option if you just want one or two bulbs and aren’t planning to build a sophisticated lighting system with switches and sensors.

Although you can control individual Philips Hue lights from your phone using Bluetooth, most people will use them with a Philips Hue Bridge, which plugs into your router and uses the Zigbee wireless protocol to create a mesh network that links all your devices together. No Wi-Fi necessary.

Using Zigbee rather than Wi-Fi means you can have a lot more smart lights in your home (a typical home Wi-Fi router isn’t really intended to communicate with more than a couple of dozen devices), with better range without the need for a Wi-Fi extender. (Zigbee devices also use less power, though LED bulbs aren’t huge energy-hogs to begin with, so you’re not likely to notice a huge difference on that front.)

Philips Hue Essential and WiZ bulbs, therefore, aren’t filling quite the same niche. If you only ever intend to own a couple of smart lights, then WiZ would be just fine, but Hue Essential will give you the option to expand your setup much further in the future, should you choose to.

A bright idea?

"Hue stretches across indoor and outdoor, and has different variants of products in all those categories," Giuliano Ghidini, Business and Marketing leader at Signify, told me in a recent interview.

"Hue offers a more comprehensive range so you can cover all rooms, and thanks to the technology it’s based on, Zigbee, with a Hue hub, you have very good coverage indoor and outdoor, and very high reliability without putting too much stress on your Wi-Fi network."

Pick up a couple of Philips Hue Essential bulbs with a Hue Bridge, and you'll have the option to extend your setup much further at a future date if you want to (Image credit: Philips Hue)

That makes Hue appealing if you have a generous budget and can afford to deck out your entire house, but the high entry price can be off-putting otherwise. A starter kit of two White & Color Ambiance bulbs bundled with a Hue Bridge might cost as much as $140 / £130 / AU$190. When you can pick up four Govee bulbs for a quarter of the price, with no hub necessary, it’s easy to see why homeowners with more modest needs would write off Hue.

The Philips Hue Essential series, depending on how it’s priced, sounds like it could offer the best of both worlds, with a palatable asking price and the ability to expand your smart lighting setup as much as you like later on, or just keep it simple and not feel like you’re under-utilizing it.

Essential? Perhaps not quite, but certainly more tempting.

You might also like

• I swapped my alarm clock for this Philips Hue light, and I've never felt more refreshed in the morning
• If Philips Hue is too expensive, the HDMI Sync Box and TV Backlight set from WiZ is a cheap and effective alternative
• Phillips Hue lights get a free AI upgrade – and it's a feature that could take your smart home to the next level

The partnership is one of the first big moves from Meta's newly revamped AI team. Here's what we know so far.

In a clumsy T1 preview, Trump Mobile teases two entirely different phones on social media.

An announcement of famine — as has now happened regarding Gaza — is a complicated decision. Here's what must be considered before such a declaration is made.

(Image credit: Eyad Baba/AFP)

• Hackers claim to be selling millions of PayPal logins, but experts suspect foul play
• The dataset allegedly includes passwords, emails, and URLs for automated attacks
• Experts say the leaked sample is too small to confirm authenticity, and its low pricing casts doubt about its legitimacy

Hackers recently announced on a well-known forum that they were selling a dataset of 15.8 million stolen PayPal credentials, allegedly including login emails and plaintext passwords.

The cybercriminals claim the information was stolen in May 2025, and the dataset contains not just emails and passwords but also associated URLs, making it easier for criminals to automate credential stuffing attacks and launch identity theft scams.

They also claim that while many of the leaked passwords appeared unique and “strong-looking,” a large portion were reused. If true, the value of the dump may be smaller than suggested.

Doubts over the breach claims

However, experts who examined the small sample released to the public concluded it was insufficient to verify the attackers’ claims, noting if the breach really occurred in May 2025, much of the usable data might already have been exploited.

Interestingly, the price set for the alleged database is surprisingly low, raising further doubts about its authenticity.

Historically, high-quality stolen data commands far higher prices on the dark web.

However, PayPal quickly denied any new breach, instead pointing to a “security incident” from 2022, which involved credential stuffing attacks and resulted in regulators fining the firm earlier this year.

That event saw only 35,000 accounts exposed, a far cry from the millions now claimed by attackers.

Skeptics argue the resemblance between the alleged PayPal dataset and the structure of infostealer malware logs from an older event suggests foul play.

Infostealers quietly harvest passwords, cookies, and other details from infected devices, often packaging the data with a URL followed by login information.

It is quite common to find credentials listed in stealer logs that circulate on dark web marketplaces, but these are not directly from PayPal’s system; they are from compromised user devices.

Regardless of whether this new claim proves genuine, the situation underscores how easy it is for user information to circulate once stolen.

Leaked login details can enable identity theft and financial fraud long after the original compromise.

Users who have reused PayPal credentials on other platforms remain vulnerable to attack.

How to stay safe

• Change your PayPal password and avoid reusing it across other services.
• Enable multi-factor authentication to add an extra layer of security.
• Monitor accounts regularly for signs of identity theft or unusual activity.
• Use a strong internet security suite with firewall protection.
• Be cautious with links and attachments that may carry infostealer malware.
• Consider dedicated identity theft monitoring services for added protection.

Via Cybernews

You might also like

• Take a look at our roundup of the best business VPNs available now
• These are the best cloud document storage solutions to choose from
• Asus new laptops come with 5-year BIOS update warranty and 1-year McAfee+ premium subs

The president's highly unusual announcement underscores the Trump administration's desire to take control over U.S. businesses.

(Image credit: David Paul Morris)

The case of Abrego Garcia, a Maryland man originally from El Salvador, raised basic questions of due process in Trump's crackdown on undocumented immigrants after he was arrested and sent to a maximum security prison in El Salvador, violating a U.S. immigration judge's 2019 order prohibiting his deportation to his home country.

(Image credit: Press Office Senator Van Hollen)

Family members of the victims of the 2022 Uvalde school shooting where 19 students and two teachers died, recently got a look at newly released files from the Uvalde Consolidated School District and Uvalde County from the day of the shooting. More than three years after the tragedy, their suffering lingers without answers to their questions about how the security protocols failed that day.

Here are hints and the answers for the NYT Connections: Sports Edition puzzle for Aug. 23, No. 334.

Here are some hints and the answers for the NYT Connections puzzle for Aug. 23, #804.

Here are hints and answers for the NYT Strands puzzle for Aug. 23, No. 538.