Over a decade since Microsoft tried to make talking to Cortana on PCs a thing — and spectacularly failed in the process — the Windows giant is taking another swing at voice commands with its Copilot AI assistant in Windows 11. Starting today, the company is rolling out an upgrade to its existing Copilot Voice and Vision features which will let you say “Hey, Copilot” and then ask your PC a question based on what’s on the screen. If you’re looking at pictures of Hawaii, for example, you could ask your Windows 11 PC where exactly they were taken, have it plot you a flight plan and potentially even give you some budgeting tips to afford that island vacation.

Microsoft’s jaunty promotional videos for the Copilot features, set to Vampire Weekend’s almost two-decade-old “A Punk,” make the process look practically seamless. One user asks Copilot to show them how to stream their music in the “best possible quality,” and the AI proceeds to highlight the exact location of the streaming settings in Spotify, while suggesting they choose the lossless option. Another person asks Copilot to write up a short biography based on their photo portfolio. Now Copilot isn’t just about searching the web or generating novelty AI art, it’s making it easy for users to perform practical tasks without much effort.

Microsoft is clearly striving for the convenience of the Star Trek ship computer, a dream that also pushed Amazon to invest billions in its Echo devices and Alexa. The difference with Copilot is that you’re not just talking to a faceless speaker — Microsoft is also trying to make Windows 11 aware of what you’re doing on your screen. The “Hey Copilot” feature and all of the Copilot Vision are cloud-based, so you’ll have to live with image data of your desktop making its way to Microsoft’s servers. That involves a level of trust the company has lost with many users, especially after the messy debut of Recall, its first flagship AI-powered feature.

It doesn’t help that many people are still peeved about the death of Windows 10 support this week. Unsurprisingly, the company stresses that “Hey, Copilot” is a purely opt-in feature that’s buried in the Copilot app settings. (Of course, that can always change, especially if the company wants to juice AI engagement stats in a few years.)

Microsoft

I suspect it’ll be even harder for users to swallow where Microsoft wants to take Copilot: Giving it the ability to perform Windows tasks on its own. That’s the goal of the experimental Copilot Actions feature, which initially debuted as a tool that could perform tasks on websites. Once enabled, Copilot Actions can be prompted to handle manual tasks, like resizing and straightening an a folder of photos. If any questions pop up, it can prompt you to answer them within the Copilot app. And as Copilot Actions is handling its job in the background, you’re free to do anything else you’d like on your computer.

Conceptually, Copilot Actions sounds similar to handing off a task to a real life assistant — but just like a human assistant, there’s always a chance something could go wrong along the way. It’s also not hard to imagine the feature being coopted by nefarious malware down the line, since it’s basically a Windows script in a better interface. Microsoft says it’s tested Copilot Actions “extensively” internally, and it’s rolling out the feature slowly to gather feedback.

Just like “Hey, Copilot,” it’s entirely opt-in, and you can see everything Copilot Actions is doing step-by-step in the Copilot app. Microsoft says you’l be able to jump in and take control of a Copilot Action job at any point, as well as control the permissions of AI agents in Windows 11’s user settings. Copilot tasks are also performed in a contained environment, according to Microsoft, which allows for even more specific permissions controls as well as runtime isolation (so Copilot can’t affect the rest of your system beyond its specific task).

And as if we’re not already inundated with Copilot all over Windows 11 already, Microsoft also plans to add an “Ask Copilot” search function right on Windows 11’s taskbar. The company claims it’s part of a mission to make the taskbar “a dynamic hub” for accomplishing tasks, but personally I like to keep my taskbar clear so I can cram in more app windows. Like everything Microsoft is announcing today, the Ask Copilot bar will also be entirely opt-in.

As someone who’s been skeptical of Microsoft’s Copilot initiatives so far, I could actually see myself using “Hey Copilot” if it works as advertised. It sounds far more practical than the old Siri voice commands, which were limited by simplistic language models from a decade ago. Microsoft is also expanding AI actions built into Windows 11, including a new integration with Manus, an AI agent that can do things like turn several documents into a website, as well as Filmora, which lets you create AI videos right from the File Explorer.

The new “Hey Copilot” and Copilot Vision features are available today on all Windows 11 PCs that have access to Copilot. Microsoft is also making Copilot Vision broadly available around the world today where Copilot is available. Copilot Actions and the Ask Copilot taskbar feature will “gradually” become available to Windows 11 Insiders, according to Microsoft.

This article originally appeared on Engadget at https://www.engadget.com/computing/microsofts-next-windows-11-ai-gamble-just-say-hey-copilot-130000875.html?src=rss

Pinterest is taking new steps to reduce the amount of AI slop its users are seeing. The service is adding settings that allow people to “dial down” AI-generated content in a number of categories that are “highly-prone” to such imagery, the company said in an update.

While most social platforms have grappled with how to deal with the rise of AI-created content, Pinterest has been particularly inundated. Its image-board UI has proven particularly susceptible to AI slop, and users have complained about the difficulty of finding content created by humans.

Now, Pinterest is offering users more control over how much AI content appears in their recommendations. The service is adding a “refine your recommendations” setting that allows you to toggle generative AI content from specific categories, including art, architecture, beauty, fashion, entertainment, health, home decor and sport. According to the company, these topics have seen an influx of AI-generated content, but users should “expect even more additions in the future.”

Notably, Pinterest isn’t promising to root out generative AI content entirely. Rather, it says the new settings should “dial down” the amount of AI-based content they’re seeing in a particular category. A spokesperson for the company says this is because not all AI-generated content on the platform is low quality and some users are in fact open to seeing AI-generated material.

The setting also applies only to image pins, not video, so it likely won’t do much to prevent Sora or other AI-created video clips from appearing in your feeds. For AI-created or AI-edited content that does continue to surface, Pinterest says it will label these posts more prominently. The company started experimenting with labels back in May, but has now “ramped up” its tools for identifying such content. 

Pinterest’s new settings are available now on desktop and Android and will be available on iOS in the next few weeks.

This article originally appeared on Engadget at https://www.engadget.com/social-media/pinterest-will-let-you-dial-down-ai-slop-in-your-feeds-130000337.html?src=rss

ExpressVPN is one of the best VPNs on the market, with user-friendly apps, excellent speed test scores and a strong security record. In my ExpressVPN review, I found it to live up to its positive word of mouth, especially when unblocking foreign streaming sites. But no service is perfect, and my cup of VPN tea is not everybody’s. If you’re looking to switch away, follow this guide to cancel ExpressVPN.

How to cancel ExpressVPN on desktop

No matter where you originally signed up for ExpressVPN, you can cancel through your browser on a desktop platform. With any browser (i.e. Chrome, Safari, etc), the steps are as follows. Note that doing this will instantly cancel your ExpressVPN subscription and revoke your access to the service.

1. Go to expressvpn.com. In the top bar, click on My Account. Enter your username and password to sign in.

2. You’ll be taken to your account dashboard. In the menu on the left-hand side of the screen, click Subscription.

3. Scroll down until you find the words “Subscription details.” You should see a box containing the monthly cost of your subscription. At the right of that box, click Cancel Subscription.

4. From here, follow the on-screen prompts to complete cancellation.

Sam Chapman for Engadget

If you’re not ready to lose ExpressVPN service just yet, you can end auto-renewal instead of cancelling altogether. You’ll still be able to use ExpressVPN until your subscription runs out.

To cancel auto-renew, go to the subscription tab of your account dashboard and find the box with your subscription ID (it should be right at the top). In that box, click the link that says Edit subscription settings, then scroll down and click Turn Off Automatic Renewal. If you change your mind before your subscription lapses, you’ll be able to turn it back on later.

How to cancel ExpressVPN on mobile

You can also cancel ExpressVPN on your phone or tablet, but the process is largely the same — some buttons are just in different places. As above, this instantly ends your ExpressVPN subscription. Here’s how to do it.

1. Go to expressvpn.com in your mobile browser. Tap the three horizontal lines at the top-right, then in the menu that appears, tap My Account.

2. In the account dashboard, tap the three horizontal lines at the top-right once again. This time, scroll down to the expanded “My Account” menu and tap Subscription.

3. Scroll down to the “Subscription details” box and tap Cancel Subscription.

4. Follow the prompts on the screen to complete cancellation.

Like on desktop, you can also turn off auto-renewal so your subscription ends when your current pay period expires. Go to the subscription tab of your account dashboard as described in the steps, find the box with your subscription ID, then follow the steps from the last paragraph of the previous section.

Cancelling ExpressVPN through an app store

The website is almost always the right way to cancel ExpressVPN, but there is one exception: if you originally signed up through the Google Play Store or Apple App Store. This includes both paid subscriptions and the 7-day free trials ExpressVPN offers to app store users.

In this case, you’ll need to cancel through the app store where you originally made the purchase. On Apple, open the app store, then tap Subscriptions and scroll down to find ExpressVPN. On Android, open Google Play and tap Payment & subscriptions, then Subscriptions. In both cases, once you’ve found ExpressVPN, tap it and scroll down to find the button for cancellation.

How to get a refund from ExpressVPN

ExpressVPN offers a full refund to anyone who cancels a subscription within 30 days of purchase. You can also get a refund if your subscription renews without your consent — if that happens, you have 14 days to request your money back.

Sam Chapman for Engadget

There’s no dedicated button for asking for a refund. Instead, you’ll need to contact ExpressVPN. To do that, go to expressvpn.com/support, then click on the button in the bottom-right corner that says Need help? Chat with us! This will open a chat with a bot which you can use to ask for your money back. Unfortunately, there’s no public link to email the support team, so live chat is the only option here.

ExpressVPN alternatives

Depending on what made you want to cancel ExpressVPN, there are a few other VPNs you might like better. Proton VPN is my favorite provider at the moment, combining great apps with a unique focus on privacy rights (it’s also cheaper). NordVPN has a lot of useful features ExpressVPN leaves off its no-frills clients. And if you’re all about speed, Surfshark is the current fastest VPN in my tests.

This article originally appeared on Engadget at https://www.engadget.com/cybersecurity/vpn/how-to-cancel-expressvpn-and-get-a-full-refund-123020142.html?src=rss

The difference in emissions between plug-in hybrid electric vehicles (PHEVs) and gas cars is merely 19 percent and not 75 percent as lab tests show, according to a report by a group of European transport and environment NGOs. They analyzed data transmitted by the onboard fuel consumption meters of 800,000 plug-in hybrids registered in the continent between 2021 and 2023. What they found was that the real-world carbon dioxide emissions of PHEVs were 4.9 times higher than what they’re supposed to emit on paper, mostly because of manufacturers’ and testers’ flawed assumption on the share of the vehicles’ electric driving mode. 

Specifically, the official estimates for the share of electric driving mode are 84 percent, but the share is only as little as 27 percent in the real world. That means most of the driving done on PHEVs uses gas. In addition, even when PHEVs are in electric driving mode, they’re still emitting considerable pollution. Apparently, their electric motors don’t have sufficient power to operate on their own, and their combustion engines have to supply power for almost a third of the distance driven. The lack of fast charging in PHEVs also makes it less convenient to charge, leading people to just fill their tanks with gas. All these factors don’t only translate to pollution, but also to additional gas expenses. The report says drivers end up paying $582 more than official estimates for fuel costs.

As The Guardian notes, carmakers have been pushing for the adoption of PHEVs in Europe as a low-emissions alternative that won’t give owners range anxiety, with Volkswagen, Mercedes-Benz and BMW accounting for most of the sales over the past years. The researchers have noted that four major carmaker groups have avoided over $5.8 billion in fines between 2021 and 2023 due to the underestimation of PHEV emissions in the lab. These official emission numbers allowed companies to comply with the required carbon dioxide emission targets for their fleets. Based on real-world data, however, the vehicles they sold emitted 52 million tons more carbon dioxide than they were supposed to.

This article originally appeared on Engadget at https://www.engadget.com/transportation/plug-in-hybrids-emit-almost-as-much-pollution-as-gas-cars-report-finds-120032534.html?src=rss

It could be a consequential act of quiet regulation. Cloudflare, a web infrastructure company, has updated millions of websites’ robots.txt files in an effort to force Google to change how it crawls them to fuel its AI products and initiatives.

We spoke with Cloudflare CEO Matthew Prince about what exactly is going on here, why it matters, and what the web might soon look like. But to get into that, we need to cover a little background first.

The new change, which Cloudflare calls its Content Signals Policy, happened after publishers and other companies that depend on web traffic have cried foul over Google’s AI Overviews and similar AI answer engines, saying they are sharply cutting those companies’ path to revenue because they don’t send traffic back to the source of the information.

Read full article

Comments

GrapheneOS, the privacy-focused Android fork once exclusive to Google Pixels, is partnering with a major Android OEM to bring its hardened, de-Googled OS to Snapdragon-powered flagship phones. Android Authority reports: Until now, GrapheneOS has been available only on Pixel phones, making Google’s flagships popular among privacy enthusiasts, journalists, and, as a Spanish police report suggested earlier this year, even organized crime groups in Catalonia. But that Pixel exclusivity may end by 2026 or 2027. GrapheneOS revealed in a Reddit thread that it has been working with a “major Android OEM” since June 2025 to enable official support for “future versions of their existing models.” These devices will reportedly use flagship Snapdragon chips, a notable shift from Google’s in-house Tensor processors.

The project explained that only Pixels have met its strict security and update requirements so far. However, the new partnership suggests that another OEM is finally matching those standards. GrapheneOS also hinted that the mysterious partner’s devices will be “priced similarly to Pixels” and available globally as part of the brand’s standard lineup.

Read more of this story at Slashdot.

Scientists from Spain and China have successfully repaired the blood-brain barrier in Alzheimer’s-model mice, enabling the brain to naturally clear amyloid-beta plaques and reverse cognitive decline. “After just three drug injections, mice with certain genes that mimic Alzheimer’s showed a reversal of several key pathological features,” adds ScienceAlert. From the report: Within hours of the first injection, the animal brains showed a nearly 45 percent reduction in clumps of amyloid-beta plaques, a hallmark of Alzheimer’s disease. The mice had previously shown signs of cognitive decline, but after all three doses, the animals performed on par with their healthy peers in spatial learning and memory tasks. The benefits lasted at least six months.

These preclinical results don’t guarantee success in humans, but they’re an encouraging start, which the authors say “heralds a new era” in drug research. “The therapeutic implications are profound,” claim the international team of researchers, co-led by scientists at the Institute for Bioengineering of Catalonia (IBEC) and the West China Hospital Sichuan University (WCHSU). The findings have been published in Signal Transduction and Targeted Therapy.

Read more of this story at Slashdot.

An anonymous reader quotes a report from the Conversation: Sound waves at frequencies above the threshold for human hearing are routinely used in medical care. Also known as ultrasound, these sound waves can help clinicians diagnose and monitor disease, and can also provide first glimpses of your newest family members. And now, patients with conditions ranging from cancer to neurodegenerative diseases like Alzheimer’s may soon benefit from recent advances in this technology.

I am a biomedical engineer who studies how focused ultrasound — the concentration of sound energy into a specific volume — can be fine-tuned to treat various conditions. Over the past few years, this technology has seen significant growth and use in the clinic. And researchers continue to discover new ways to use focused ultrasound to treat disease. […] Research on focused ultrasound has primarily focused on the most devastating and prevalent diseases, such as cancer and Alzheimer’s disease. However, I believe that further developments in, and increased use of, focused ultrasound in the clinic will eventually benefit patients with rare diseases.

One rare disease of particular interest for my lab is cerebral cavernous malformation, or CCM. CCMs are lesions in the brain that occur when the cells that make up blood vessels undergo uncontrolled growth. While uncommon, when these lesions grow and hemorrhage, they can cause debilitating neurological symptoms. The most common treatment for CCM is surgical removal of the brain lesions; however, some CCMs are located in brain areas that are difficult to access, creating a risk of side effects. Radiation is another treatment option, but it, too, can lead to serious adverse effects.

We found that using focused ultrasound to open the blood-brain barrier can improve drug delivery to CCMs. Additionally, we also observed that focused ultrasound treatment itself could stop CCMs from growing in mice, even without administering a drug. While we don’t yet understand how focused ultrasound is stabilizing CCMs, abundant research on the safety of using this technique in patients treated for other conditions has allowed neurosurgeons to begin designing clinical trials testing the use of this technique on people with CCM. With further research and advancements, I am hopeful that focused ultrasound can become a viable treatment option for many devastating rare diseases.

Read more of this story at Slashdot.

mspohr shares a report from Futurism: Western automotive and green energy executives who visit China are returning humbled — and even terrified. As The Telegraph reports, the executives are warning that the country’s heavily automated manufacturing industry could quickly leave Western nations behind, especially when it comes to electric vehicles. “We are in a global competition with China, and it’s not just EVs,” Ford CEO Jim Farley told The Verge last month. “And if we lose this, we do not have a future at Ford.” Some companies are giving up on new initiatives altogether, with the founder of mining company Fortescue, Andrew Forrest, claiming that his recent trip to China led to him abandoning attempts to produce EV powertrains in-house. “There are no people — everything is robotic,” he told The Telegraph.

Other executives recalled touring “dark factories” that don’t even need to keep the lights on, as most work is being done around the clock by robots. “You get this sense of a change, where China’s competitiveness has gone from being about government subsidies and low wages to a tremendous number of highly skilled, educated engineers who are innovating like mad,” British energy supplier Octopus CEO Greg Jackson told the newspaper. According to recent figures by the International Federation of Robotics, China has deployed orders of magnitude more industrial robots than Germany, the US, and the UK.

Read more of this story at Slashdot.

Norway has effectively achieved its 2025 goal of 100% electric new car sales, prompting the government to declare “mission accomplished” and propose scaling back EV tax exemptions to reflect a mature market. “We have had a goal that all new passenger cars should be electric by 2025, and … we can say that the goal has been achieved,” announced Finance Minister Jens Stoltenberg. Electrek reports: With the finish line in sight, the Norwegian government is now fine-tuning its approach. The current incentive program maintains the crucial VAT exemption for EVs, but only up to a purchase price of 500,000 Norwegian kroner (approximately $49,000 USD). This move is designed to target more expensive, luxury EVs, ensuring that the incentive benefits a broader range of consumers.

However, the latest budget proposal aims to reduce the EV tax exemption to vehicles costing 300,000 Norwegian kroner (~30,000 USD). This would apply for 2026, and then the tax exemption would completely end in 2027. Additionally, the government plans to increase taxes on new gasoline and diesel cars, further widening the cost gap between polluting and zero-emission vehicles.

However, the proposal still needs to be adopted by Norway’s government, and there is some opposition. EV associations are advocating for a more extended phase-out period to ensure that the adoption rate doesn’t decline.

Read more of this story at Slashdot.

Boris Johnson’s former adviser claims that China infiltrated a key UK government data-transfer network for years, compromising highly classified materials and prompting a Whitehall cover-up that prioritized Chinese investment over national security. The Times reports: Dominic Cummings, who served as a senior adviser to Boris Johnson, said that he and the then prime minister were informed about the breach in 2020 but that there had subsequently been a cover-up. He said he was warned at the time that disclosing some specific details of the breach would be a criminal offence. He claimed that the breach included some “Strap” material, which is the government term for the highest level of classified information.

The breach, which was confirmed by two other senior Whitehall sources, was said to have been connected to a Chinese-owned company involved in Britain’s critical national infrastructure. Tom Tugendhat, a former Tory security minister, supported Cummings’s account. Cummings said that he and Johnson were informed of the breach in the “bunker” of No 10 — a reference to the secure room in Downing Street.

He told The Times: “The cabinet secretary said, ‘We have to explain something; there’s been a serious problem’, and he talked through what this was. “And it was so bizarre that, not just Boris, a few people in the room were looking around like this — ‘Am I somehow misunderstanding what he’s saying? Because it sounds f***ing crazy.'” He added: “What I’m saying is that some Strap stuff was compromised and vast amounts of data classified as extremely secret and extremely dangerous for any foreign entity to control was compromised. “Material from intelligence services. Material from the National Security Secretariat in the Cabinet Office. Things the government has to keep secret. If they’re not secret, then there are very, very serious implications for it.”

Read more of this story at Slashdot.

An anonymous reader quotes a report from The Guardian: People in London could be hiring driverless taxis from Waymo next year, after the US autonomous vehicle company announced plans to launch its services there. The UK capital will become the first European city to have an autonomous taxi service of the kind now familiar in San Francisco and four other US cities using Waymo’s technology.

The launch pits an innovation sometimes dubbed the “robotaxi” against London’s famous black cabs, which can trace their history back to the first horse-drawn hackney coaches of the Tudor era. But a representative of the capital’s cab drivers said they were not concerned by the arrival of a “fairground ride” and questioned the reliability of driverless vehicles. Waymo said its cars were now on their way to London and would start driving on the capital’s streets in the coming weeks with “trained human specialists,” or safety drivers, behind the wheel.

The company, originally formed as a spin-off from Google’s self-driving car program, said it would scale up operations and work closely with Transport for London and the Department for Transportto obtain the permits needed to offer fully autonomous rides in 2026. Uber and the UK tech company Wayve have also announced their own plans to trial their driverless taxis in the capital next year, after the British government said it would accelerate rules allowing public trials to take place before legislation enabling self-driving vehicles passes in full.

Read more of this story at Slashdot.

This is a flight demonstration of X Control’s Janus-I, a personal dual-blade helicopter that was dubbed the ‘flying suitcase’, presumably in honor of George Jetson’s flying car (which unfolded from his suitcase). The gas-powered helicopter folds up small enough to fit in the back of a truck, weighs only 154-pounds but can carry 440-pounds, has a top speed of around 100MPH, can reach altitudes of ~20,000ft, and can even take off and land on water. What it can’t do is make you feel the least bit safe flying it at any altitude over 10 feet. Especially considering you don’t even need a pilot’s license to fly it, so there’s very little chance you’ll have any idea what you’re doing besides being moments from death.

Mozilla is testing a free, built-in VPN for Firefox that routes traffic through Mozilla-managed servers directly in the browser. The Register reports: According to a staff post on Mozilla Connect, the company’s idea-sharing platform, Firefox VPN is still an experimental feature in the early stages of development, but users will be selected at random to test it “over the next few months.” Moz describes the feature as one that will sit beside the search bar on Firefox, routing web traffic through a Mozilla-managed VPN server, concealing the user’s real IP address while adding a layer of encryption to their communications. Firefox VPN is a different project entirely from Mozilla VPN, a separate, paid-for product. The Firefox version will be free to use and confined to the browser itself, while Mozilla VPN can be used by up to five devices at a time.

The Moz staffer on the product team who announced the feature said of the upcoming beta test: “We’ll start simple, then gradually add new capabilities while learning how it impacts browsing, usage, and overall satisfaction. “Our long-term vision is ambitious: to build the best VPN-integrated browser on the market.” In response to feedback, the staffer noted that while it will be a desktop browser feature first, “mobile is definitely a natural next step.”

Read more of this story at Slashdot.

Anthropic is projecting its annualized revenue run rate to soar from roughly $7 billion today to as much as $26 billion in 2026, driven by rapid enterprise adoption of its Claude AI models. Reuters reports: Anthropic debuted a new version of its cheapest AI model, Haiku, on Wednesday, as part of a broader effort to appeal to companies that are looking for capable AI systems that are dramatically cheaper than its more advanced models. The Haiku 4.5 model sells for about one-third the price of Sonnet 4, one of its medium-sized models.

The revenue projections underscore continued strong demand for generative AI tools among businesses and help explain investor enthusiasm, even as AI spending, especially in infrastructure buildout, comes under scrutiny. Some people worry the level of investment might be unsustainable. Fueling the expansion is the uptake of enterprise products, which are built for organizations. Anthropic has more than 300,000 business and enterprise customers, which account for about 80% of its revenue.

Read more of this story at Slashdot.

In an embarrassing error, PayPal blockchain partner Paxos accidentally minted 300 trillion of the PYUSD token today. PayPal’s stablecoin, which it debuted in 2023, is supposed to be redeemable at a 1:1 rate for US dollars. $300 trillion is more than double the entire world’s current GDP of $117 trillion. In short, that’s a really bad mistake from Paxos.

At least the action did appear to be a mistake. Paxos acknowledged the issue in a post on X, assuring clients that their funds are secure and stating that it burned the excess reserves of stablecoins. The company claimed that “this was an internal technical error.” Maybe so, but internal technical errors can still wreak havoc pretty quickly when it comes to finance, especially in the fantasy worlds of blockchain and crypto. Thanks to the appropriately named Web3 is Going Just Great for picking up on this story.

This article originally appeared on Engadget at https://www.engadget.com/cybersecurity/paypals-blockchain-partner-accidentally-minted-300-trillion-in-stablecoins-224017538.html?src=rss

An anonymous reader quotes a report from BleepingComputer: U.S. cybersecurity company F5 disclosed that nation-state hackers breached its systems and stole undisclosed BIG-IP security vulnerabilities and source code. The company states that it first became aware of the breach on August 9, 2025, with its investigations revealing that the attackers had gained long-term access to its system, including the company’s BIG-IP product development environment and engineering knowledge management platform.

F5 is a Fortune 500 tech giant specializing in cybersecurity, cloud management, and application delivery networking (ADN) applications. The company has 23,000 customers in 170 countries, and 48 of the Fortune 50 entities use its products. BIG-IP is the firm’s flagship product used for application delivery and traffic management by many large enterprises worldwide. […]

F5 is still reviewing which customers had their configuration or implementation details stolen and will contact them with guidance. To help customers secure their F5 environments against risks stemming from the breach, the company released updates for BIG-IP, F5OS, BIG-IP Next for Kubernetes, BIG-IQ, and APM clients. Despite any evidence “of undisclosed critical or remote code execution vulnerabilities,” the company urges customers to prioritize installing the new BIG-IP software updates.

Read more of this story at Slashdot.

Google has released a new update to its Veo AI video generation model that should make it do a better job of sticking to prompts and converting images into videos. Veo 3.1 is available to try today through Google’s Gemini API and is now also powering the company’s Flow video editor.

Veo 3.1 builds on the new capabilities Google introduced with launch of Veo 3 at Google I/O 2025. The new model offers better “prompt adherence,” according to Google, and should have an easier time creating videos based on the image “ingredients” you upload alongside your written prompt. Veo 3.1 also makes it possible to convert images to video and generate audio at the same time, a capability that wasn’t available with Veo 3. 

In Flow, Veo 3.1 supports at least a new feature that gives you finer control over the videos you generate. With what Google calls “Frame to Video,” Flow lets you upload a first and last frame, and then generates the video in-between. Adobe Firefly, which is powered by Veo 3, offers a similar feature, but Flow will be able to pull it off and create audio at the same time. Those added audio skills will also apply to the video editor’s ability to extend clips and insert objects into existing footage, too.

Based on the samples Google’s shared, videos generated with Veo 3.1 still have an uncanny quality that seems to vary greatly depending on the prompt and subject. Even if it’s missing some of the realism of OpenAI’s Sora 2, though, the company’s decision to try and make Veo more useful to people who actually work with video rather than a source of social media spam is a welcome move.

This article originally appeared on Engadget at https://www.engadget.com/ai/googles-veo-31-is-better-at-generating-videos-from-images-220829129.html?src=rss

We may earn a commission from links on this page.

Did you know you can customize Google to filter out garbage? Take these steps for better search results, including adding my work at Lifehacker as a preferred source.

The new Sony WH-1000XM6 are the best premium headphones of 2025, and they keep getting better. One of my favorite features is its LE Audio Bluetooth compatibility, which is a newer Bluetooth type that uses less energy and supports better audio quality.

If you’ve tried to use LE Bluetooth in the past on your XM6 headphones, you might’ve noticed some features were missing, like head tracking, Find My Device, Gemini Live, and Audio Sharing with Fast Pair. Well, that’s no longer the case: The latest update to the Sound Connect app allows you to take advantage of all of these features while using the LE Audio technology.

Left: LE Audio (LC3) Codec, Middle: Low Latency Bluetooth, Right: Pop up message for LE Adudio
Credit: Daniel Oropeza

How to make sure you get the update

While I’m a fan of the headphones, I can’t say I feel the same way about Sony’s companion app. It’s confusing and not user-friendly (and I’ve been using the app for years). The update was not easy to download, nor was it easy to know when it was done. But eventually, a message did pop up on my app asking me if I wanted the latest 3.0.0 update.

You can check if you already have it by clicking the three dots on the top right corner of the main menu and selecting “WH-1000XM6 Version.” Once you download it, you will need to do a hard connection reset between your phone and headphones for the update to kick in. Select the forget device on the XM6 from your Bluetooth settings and connect it again. Once you do all of that, you should be ready.

How LE Audio performs with the new features

It’s hard to say the LE Audio codec (LC3) sounds any better than it did before without a side-to-side comparison, but I can tell you that it sounds great. The audio is exactly what I would expect from arguably the best headphones of 2025.

The head tracking, which creates an all-around spatial sound quality, worked great as well. I took walks around my neighborhood and could hear conversations neighbors half a block away were having in their driveway.

When it comes to voice assistants and Gemini Live, your mileage may vary. From my experience, the mics on the WH-1000XM6 can be a bit sensitive, where long-winded answers from Gemini Live were cut off prematurely from noise around me. I did most of my testing while walking around my block and working from home (my wife also works from home, and we have a noisy dog). But this might be more about Gemini Live than the LE audio performance. Outside of those instances, the feature worked like I’d expect.

I can say I’ve noticed my sessions on the XM6 have lasted longer than before, which is one of the biggest advantages of using LE Audio. While the Low Latency feature is still in beta, it’s already meeting the high standards of the XM6.

Last month, OpenAI published a usage study showing that nearly 15 percent of work-related conversations on ChatGPT had to deal with “making decisions and solving problems.” Now comes word that at least one high-level member of the US military is using LLMs for the same purpose.

At the Association of the US Army Conference in Washington, DC, this week, Maj. Gen. William “Hank” Taylor reportedly said that “Chat and I are really close lately,” using a distressingly familiar diminutive nickname to refer to an unspecified AI chatbot. “AI is one thing that, as a commander, it’s been very, very interesting for me.”

Military-focused news site DefenseScoop reports that Taylor told a roundtable group of reporters that he and the Eighth Army he commands out of South Korea are “regularly using” AI to modernize their predictive analysis for logistical planning and operational purposes. That is helpful for paperwork tasks like “just being able to write our weekly reports and things,” Taylor said, but it also aids in informing their overall direction.

Read full article

Comments