Using the antivirus trial that came with your PC? Or perhaps you’ve just been sitting it out with Windows Security? You may be able to do better. Why only may do better? It all depends on your needs—factors like your household, level of tech savvy, and willingness to directly manage your protection. You could work in tech but might not have time to constantly monitor your family’s fleet of devices, for example. Or maybe you’re an individual with interest and time to stay abreast of the latest security news. With that in mind, these three antivirus recommendations are the ones I most often suggest. They cover a broad range of situations, no matter your circumstances or budget. Option 1: Windows Security The former Security Center is now called “Windows Security” and summarizes the Windows 11 security modules in one overview. Foundry This may be the option you’ve already got—or perhaps you don’t know it exists. Either way, Microsoft’s built-in security can be an adequate defense against threats, if you already practice good security habits. (Not installing software from questionable sources or purposely hitting up sketchy sites, etc.) You will, however, have to play with the settings to get the most out of the app. Ransomware protection may not automatically be on by default, for example. The same goes for blocking unwanted apps by turning on reputation-based protection in Windows Security’s App & browser control settings. If you’re on a fresh PC, I recommend also turning on Smart App Control while in the App & browser control settings, for an extra boost. (Note: These settings can mildly interfere with select apps, but most people won’t notice.) By the way—if your PC has a free trial for another antivirus service already installed, you’ll need to disable it (or better, uninstall it) to allow Windows Security to take over. Pros: Free, unobtrusive Cons: Bare-bones user interface with minimal explanations. Management of passwords and privacy must be done separately. Option 2: Full-featured antivirus suite Mattias Inghe Paid antivirus subscriptions may not be needed by everyone, but they do offer some advantages over Windows Security and other free options. Good ones have three key strengths: First, they’re more proactive about guarding users from dangers. For example, plans may include screening tools in the apps to detect scam language or links in messages. Second, they make understanding features and notifications much easier, by giving context and explanations in the interface and alerts. Finally, they simplify management of online security by offering a more comprehensive set of services. At minimum, I like to see a password manager, VPN, and basic data breach monitoring. But more robust subscriptions will include features like parental controls, cloud storage to help back up vital files (as extra protection against ransomware), or PC utilities that scan for outdated drivers. Currently, our top pick for best paid antivirus is Norton Deluxe 360, as this particular plan ticks these boxes. But there are plenty of other options that could be equally (or better) fits for your circumstances—our roundup of the best antivirus suites has multiple recommendations. Pros: Broader, centralized protection that’s easier to understand Cons: Such suites require a regular subscription fee Option 3: Free antivirus Foundry Perhaps you don’t want to pay for antivirus, but you would like higher protection, more information, or even additional features than found in Windows Security. Fortunately, you have options—and some of them are excellent. Want some of the strongest offline and online safeguards? Bitdefender’s free version is no-frills, but its antivirus engine consistently gets high marks across the board from independent testing organizations. Or perhaps you’d like strong defense and more features? You can lean on Avast One’s no-cost plan, which also can monitor your email and browser activity. These free plans rely on the same scanning engines as paid subscriptions, so you’re not compromising on protection. The main catch is that advanced features and services will require upgrading–and some apps will nag you periodically to consider forking over that money. But with some tweaks, you usually can make the software leave you in peace. Pros: Free, but with stronger security or more features than Windows Security (or both) Cons: Still limited feature sets. Periodic notifications about upgrading to paid plans. Read...

Google says that it’s bringing a major change to Chrome users today: a mainstream agentic AI that will “autobrowse” the web for you, performing tasks that you assign it. Chrome’s also getting a sidebar — and yes, with Gemini AI. Autobrowsing capabilities will arrive today for Chrome users who subscribe to either Google AI Pro ($19.99 per month) or Google AI Ultra ($249.99 per month), while the Chrome sidebar will roll out today for all Chrome users. The move comes a day after Google announced the Google AI Plus plan for $7.99 per month, which will not have access to the autobrowse features. Essentially, Google sees the autobrowsing as the evolution of its autofill capabilities, which aren’t unique to Chrome. Autofill stores details like your credit card and applies them when necessary, such as to complete a purchase. Autobrowsing simply takes a task and goes out and completes it to the best of Chrome’s ability, leaving the final step — a confirmation of purchase — for you to approve. Agentic browsing is nothing new; Microsoft showed off an agentic shopping demonstration a year ago, as well as Copilot Mode for Edge last summer. OpenAI has done something similar with the Atlas agentic browser, and others have followed suit. Nevertheless, Statcounter reports that Chrome has about 65 percent of the desktop browser market share for North America, far and away the most dominant browser. Mainstreaming agentic actions, even for a paid subscription, is a significant move. Google You’ll access auto browse the same way that you’ll interact with Google’s Gemini AI: as a browser sidebar, accessible by clicking the small Gemini icon at the top of the browser. That will open up the sidebar and a text box, and allow you to ask Gemini to start filling out tasks. Google said that auto browse tasks can include filling out PDFs, renewing drivers’ licenses, but also researching trips and other tasks, including scheduling and booking reservations. However, executives said information “that’s happening on the right-hand side is not shared back with the site.” Google said that the new Gemini sidebar will tap what it’s referred to as “personal intelligence,” remembering past conversations and information you’ve shared with it, which is now included in AI Mode. If you’ve allowed Google access to apps like Gmail, this information will be used, too. Google An interesting addition will be support for Nano Banana, Google’s image rendering algorithm. Nano Banana will be accessible by Chrome, so that it will be able to pull in and edit an image that you have in your browser — not just one that you own. Google executives couldn’t say whether there would be any copyright protections in place, or whether users would simply be able to tell the algorithm to edit the image. Yes, autobrowsing is reserved for paying subscribers — for now. But it will likely move down to cheaper tiers over time, if it proves successful. Google Read...

$1,000 is a lot to pay for a monitor. But this one might be right up your alley, if you’re looking for a combination of OLED brilliance, massive size, and relatively manageable resolution. LG’s 45-inch UltraGear OLED gaming monitor is going for $999.99 on Amazon, an impressive $700 off the retail price. Let me break it down for you. With 3440×1440 ultrawide resolution and a 240Hz refresh rate, this 45-inch monitor has the same requirements as most 34-inch gaming monitors on the market right now, including premium OLEDs. But its physical size is massive, almost double the total screen space of a 34-inch ultrawide monitor. You actually get more physical screen space than 49-inch, super-mega-ultrawides, which stretch the screen out to a wider 5120×1440 resolution (double QHD) and a 32:9 aspect ratio. Generally more pixels is better… unless you’re trying to hit high game framerates on anything less than a top-of-the-line machine. So PC gamers who want a big, beautiful picture without sacrificing performance might find this LG monitor right up their alley. The 45GS95QE includes lots of other premium features, including support for Nvidia G-Sync, AMD FreeSync, and HDR-10. It’s using a dramatic 800R curve on the panel, which makes it more immersive for gaming and video (and perhaps less ideal for regular office-style work, depending on your setup). The only downer on the spec list is USB-C, or rather, the lack thereof. The only data expansion in the monitor is a pair of USB-A ports, making this an awkward companion for laptops, which will need to use HDMI or DisplayPort. This monitor can be used with a standard VESA mount, but if you want to add a monitor arm, make sure it’s reinforced. The screen without its stand weighs 19.6 pounds or 8.9 kilograms. If you’re looking for something else, be sure to check out PCWorld’s roundup of the best gaming monitors. Get a 45-inch LG OLED ultrawide gaming monitor for $999.99View Deal Read...

Any time you leave your home to go out and about for a while, it’s always a good idea to have a power bank with you, whether in your bag or backpack or even your pocket. Yes, a tiny little power bank is better than nothing, like this compact Anker Nano power bank. You can now get it for just $46 on Amazon and it has several features that you’ll love. View this Amazon deal This power bank only has a 10,000mAh capacity, which might be on the lower side but is still more than enough to charge your smartphone about two times over (give or take depending on your model). For a day trip or several hours running errands, that should suffice with plenty to spare. And with its 45-watt high-speed charging capabilities, you’ll be able to keep your devices charged without waiting too long. The best feature of this power bank, however, is the retractable USB-C cable that pulls out to a maximum length of 2.3 feet. Not having to carry around a separate charging cable makes it all worthwhile! And the fact that it’s retractable means it stays compact and portable for maximum convenience. But it also has USB-C and USB-A ports that you can use to charge a second and third device at the same time. Get this Anker Nano 10K power bank for $46 before this hot deal expires because this one is worth every penny! Save 23% on this travel-friendly power bank with retractable USB-C cableBuy now at Amazon Read...

Windows 11 is increasingly evolving from a classic operating system to an AI-centric platform. Microsoft is integrating Copilot not as a standalone application, but as a permanent system function. For private users, however, it is not so much the strategic orientation that is decisive as the concrete added value in everyday life. This is precisely where a conflict arises between genuinely helpful functions, technical overload, and limited user acceptance to date. We have already provided an overview of the topic of AI in Windows 11 and the new functions in a separate article. Getting started, visibility, and control in the system Copilot starts directly from Windows. On supported systems, the assistant opens using the key combination Windows + C or via the icon in the taskbar. Copilot can be activated or hidden in the settings under “Personalization” and “Taskbar.” New AI PCs and AI laptops also have a physical Copilot button on the keyboard. This button is only available on AI devices with the appropriate hardware. Classic Windows 11 computers do not have it. The button makes for access easier, but does not add any functional value compared to the keyboard shortcut. The button lowers the barrier to entry, but does not replace understanding or meaningful use. Michael Crider / Foundry Copilot as an everyday tool In everyday use, Copilot primarily assists with tasks that can be described in language. Content can be summarized, texts can be roughly formulated, and longer web pages can be reduced to key messages. Explanations of Windows settings or programs can also be provided quickly. The benefit arises when you use Copilot as preliminary work. Results need to be checked, adjusted, and put into context. Copilot does not replace research and your own evaluation. Analyze files without opening them Windows 11 integrates Copilot in several places in File Explorer. A new option allows you to transfer Office files directly to Microsoft 365 Copilot to obtain summaries or content analyses. However, you need a corresponding subscription to do this. There is a similar function in the right-click menu. Both look the same but work differently. It is precisely this dual integration that causes confusion. Users often do not recognize which Copilot is active. The benefits are real, but the operation remains confusing. Those who don’t know the difference will get different results than expected. Some practice is required here. Microsoft itself obviously doesn’t yet know exactly how and where AI should be integrated into the operating system. That’s why development is dynamic: New features are added, while others are dropped. Microsoft System-wide writing assistant A universal writing assistant is now available in more and more text input fields. It corrects grammar, adjusts style, and shortens texts. This can save a noticeable amount of time for short answers, comments or forms. However, this feature requires a Copilot PC with an integrated NPU, which means that many personal computers are not compatible. In addition, the suggestions are often very neutral and standardized. Without manual editing, the texts can therefore quickly appear interchangeable. Those interested should carefully examine the assistant’s capabilities, for example directly in Notepad. It should be noted that a subscription to Microsoft 365 or even Microsoft 365 Copilot is sometimes required. In such cases, the system will indicate this accordingly. Voice input and voice mode Copilot supports voice input and tests activation via voice command. This works well in quiet living environments. In multi-person households or in the evening, voice input remains impractical. Many users still prefer text input. Microsoft is responding to this with parallel text interaction. The vision of a talking PC does not fit into everyone’s everyday life. Foundry AI PCs, NPUs, and local processing AI PCs have a neural processing unit (NPU) that processes selected AI tasks locally. These include live subtitles, studio effects for cameras, and simple classifications. These functions are energy-efficient and do not require a cloud connection. However, many Copilot functions still rely on online services. For private users, the practical difference is therefore less than the marketing suggests. The NPU primarily improves battery life and local effects — but not automatically the quality of Copilot responses. Recall as an example of overambitious features Recall stores screen snapshots to find past content via voice. Recall remains optional and requires active consent, device encryption and Windows Hello login. In practice, the picture is mixed. The quality of the hits remains unreliable. Many users disable Recall for privacy reasons or because of its limited usefulness. Recall clearly shows that technical feasibility does not guarantee everyday value. Microsoft AI agents and new system architecture Windows 11 is currently testing AI agents that perform tasks independently in the background and display their progress directly in the taskbar. One example is a research agent that creates comprehensive evaluations and displays the current status transparently. For private users, this approach theoretically promises additional convenience. At the same time, however, skepticism is growing: Autonomous actions in the file system require a high degree of trust. Microsoft is addressing these concerns with isolated workspaces, explicit approvals, and clear handover mechanisms. Nevertheless, the technology remains a preview feature with correspondingly limited acceptance. Agent Launchers, MCP, and On-Device Registry Windows registers AI agents system-wide with Agent Launchers. These can be launched via Ask Copilot, the taskbar, or the search function. An on-device registry manages capabilities and access rights locally. The Model Context Protocol enables collaboration between agents and tools. For private users, this means more automation in the future, but also more complexity. More entry points increase the learning curve and the risk of overload. Ask Copilot as a replacement for search In new insider versions of Windows 11, Microsoft is testing replacing the classic search with “Ask Copilot.” Files, settings, and apps can then be found using natural language. This works reliably for general queries. For precise file paths or known names, the classic search is often faster. Many users switch depending on the situation. A complete replacement seems unrealistic in the short term. Microsoft Accessibility and side effects AI functions are also reaching classic system areas. The screen reader receives customizable output via natural language. Voice Access simplifies setup. These functions offer real added value, regardless of the AI hype. They show that AI is convincing when it solves specific problems. These functions are being rolled out gradually and are partly reserved for AI PCs with NPUs. Low usage despite maximum presence Despite massive integration, many users rarely use Copilot. Microsoft has already reduced its sales targets for AI. Many users feel that AI in the operating system is imposed on them and is not yet optimally integrated. Creative results are difficult to reproduce. Users often continue to use AI via browsers because they can work more specifically there. The physical Copilot button increases visibility but does not generate acceptance. Hardware constraints and acceptance issues A large proportion of existing PCs do not meet the requirements for AI functions in Windows 11. Many users feel that the switch is being forced upon them. AI notebooks cost significantly more. At the same time, the practical added value of AI functions remains limited. This explains the reluctance to switch, despite the expiry of support for Windows 10. Sam Singleton Practical recommendations for private users Use Copilot selectively: Summaries, short explanations, and text drafts can save you a lot of time. However, always check the results yourself and deactivate functions that do not offer you any added value. Windows 11 remains fully functional even without active AI use. The Copilot button on AI notebooks simply makes access easier — it does not oblige you to do anything. Windows 11 is increasingly evolving into an agent-enabled system. Further AI functions, greater automation, and higher visibility are foreseeable. For users, it is not the sheer number of functions that counts, but their reliability in everyday use. In the short term, Copilot remains a tool for selected scenarios. In the long term, it remains to be seen whether Microsoft will turn mere presence into actual relevance. Microsoft Copilot enhances Windows 11, but it does not replace independent thinking. The greatest benefit comes from conscious, selective use. Many features still seem experimental, some even overly ambitious. Private users would be well advised to view Copilot as an option rather than an obligation. This way, the ubiquitous AI becomes a tool that provides support at the right moment — and otherwise remains discreetly in the background. Read...

Sonos’s first new product since 2024’s Arc Ultra soundbar has finally arrived, but it’s not something you’ll ever find in Best Buy.  Slated to arrive “soon,” the Sonos Amp Multi is like a super-changed version of the Sonos Amp, a wireless device that lets you connect traditional home audio components like turntables, wired speakers, and other devices to your Sonos setup.  While the $800 Sonos Amp can only handle a single zone of audio, the 8-channel Amp Multi can juggle up to four zones, and it also boasts pro-level room calibration technology plus a 2U rack mount for installation in an A/V closet.  No question, the “bespoke” Amp Multi is a serious piece of hardware—so serious that it’s intended only for the professional installer market. Indeed, the usual “buy now” button on the Sonos website has been replaced by a “find an installer” button.  Sonos Even if the Sonos Amp Multi isn’t for you, the arrival of the new—and very much “audio-first” — device signals that Sonos may be back on track after 15 months of turmoil.  The last big hardware release for Sonos was in October 2024, when the Arc Ultra arrived. A follow-up to the Sonos Arc, the newer soundbar packs 14 drivers, 9.1.4 channels of audio, and a special ingredient dubbed Sound Motion, a component that allows the Arc Ultra to deliver impressive low-frequency performance without a separate subwoofer.  The Arc Ultra received warm reviews, in stark contrast to the drubbing Sonos endure following the disastrous reception to the revamped Sonos app in mid 2024, a debacle that left the company reeling and spurred the resignation of its longtime CEO Patrick Spense.  Riddled with bugs and bereft of such basic features as an editable music queue or support for local music sources, the redesigned Sonos app was greeted with near universal derision, with longtime Sonos users threatening to bail on the once-acclaimed wireless speaker platform.  At the same time, Sonos launched the Ace, and pair of wireless headphones that could connect directly to Sonos soundbars but lacked Wi-Fi support, a feature that many eager Sonos fans had expected to ship with the cans. Reviews were middling, and sales figures were said to be disappointing.  The sinking Sonos ship slowly began to right itself with the arrival of interim—and now permanent—CEO Tom Conrad, who oversaw the much needed revisions to the new Sonos app while reportedly shooting down what could have been yet another debacle for the company: a rumored $400 streaming video player that would have potentially seen Sonos jumping into an unfamilar and oversaturated market with a wildly overpriced device.  Instead, the Amp Multi shows Sonos returning to its roots as an audio-first wireless speaker company. Hopefully Sonos’s next product—made for everyday consumers, one would hope—will follow the same path.  This news story is part of TechHive’s in-depth coverage of the best smart speakers. Read...

Meta is considering new paid subscription models for Instagram, Facebook, and WhatsApp, reports TechCrunch. Citing a spokesperson for the company, the goal here is to offer paying users additional features and advanced AI tools. The basic versions of these services, however, will remain free of charge. In practical terms, this means no one will be “forced” to pay to continue using Instagram, Facebook, or WhatsApp. The subscriptions are for users who want to unlock extra features. Meta has previously experimented with alternative monetization models for WhatsApp, giving users a choice between paying money or accepting advertisements. A “premium” experience for a fee According to TechCrunch, Meta is talking about a “premium experience” that will include exclusive features that offer more control over content, interactions, and new creative possibilities, among other things. It’s still unclear which specific features will be subject to fees. Meta wants to test different models and incorporate user feedback into their design. These paid subscriptions could vary significantly depending on the platform. For Instagram and Facebook, additional features related to content, reach, or analysis are conceivable. As things stand at present, WhatsApp is primarily concerned with supplementary offers and not the basic use of the messenger service. AI features are in the spotlight AI-driven features are likely to be a central component of the planned paid subscriptions. Meta has been investing heavily in artificial intelligence for years and continues to provide its own large language models free of charge with the Llama series. Unlike competitors (e.g., OpenAI and Google), Meta has not yet introduced a comprehensive AI subscription for end users. These new subscription plans could change that. Paid add-on features for creating, editing, or optimizing content are conceivable. This includes the AI-powered short video platform Vibes, which allows users to generate and edit videos. The basic version will remain free while advanced features could end up behind a paywall. Note: The planned paid subscriptions are separate from Meta Verified, which is a different payment model that’s been in place since 2023 and is primarily aimed at creators who want a verification badge, extended support, and protection against identity theft. The new paid subscriptions are intended to appeal to a broader target group and unlock additional other features. What does this mean for you? In the short term, little will change. Instagram, Facebook, and WhatsApp will continue to be free to use. In the long term, however, users may be faced with the choice of paying up for additional features and AI extras or sticking with a barebones version for free. One thing is clear: Meta is looking for new revenue sources beyond advertising, with increasing focus on paid add-ons. Meta has not yet announced how expensive these plans will be. Read...

These might be the absolute cutest version of the Amazon Echo Pop Kids! You can now get the Anthro Pup and Unicorn models (as well as the older models) for just $30 each on Amazon, which is a hefty 40% discount and a cost-effective way to introduce your kids to Alexa. View this Amazon deal The blue Anthro Pup model features a pup wearing headphones and glasses while the pink Unicorn features a… well, a poppy unicorn. They’re both adorable, though you might get the sense that these are AI-generated images. Regardless, you can get your kid whichever they like—and if they like neither, there are also the Disney Princesses and Marvel’s Avengers versions. Those are sure to be hits! The purchase isn’t just for an adorable smart speaker, by the way. It also comes with 6 free months of Amazon Kids+, a digital subscription for kids aged 3 to 12, providing unlimited to all kinds of kid-friendly audiobooks, podcasts, ebooks, games, and other cool features. Of course, this speaker also comes with Alexa+, Amazon’s next-gen AI assistant with better smart home controls, natural language processing, and more. When your kids ask Alexa something, they’ll get a better answer than they would’ve with the older Alexa, and this can come in handy for things like homework and stray questions. Go ahead and grab this $30 kid-friendly Amazon Echo Pop speaker and cheer up their rooms with these adorable new Echo Pop Kids covers. Get a super-cute and kid-friendly Amazon Echo speaker for 40% offBuy now at Amazon Read...

The Steam Deck is a remarkable gadget for a lot of reasons, but Valve’s work on the handheld’s Linux-based SteamOS software (and its support of the same) is its real special ingredient. For example, the Steam Deck Verified system lets you know that a game will work on its somewhat pokey hardware, tested and checked by Valve itself. The system recently passed 25,000 confirmed titles. That’s according to the third-party database SteamDB, which has cataloged 25,055 total games as either Verified (7,528 titles that are running with zero issues) or Playable (17,527 titles that are running with few issues). Meanwhile, GamingOnLinux says a little over 6,000 games have been designated as “Unsupported” by Valve. That could be for any number of reasons, but it generally means a lack of controller support, too-high technical requirements for the AMD Zen 2 architecture, or third-party software incompatible with SteamOS (e.g., kernel-level anti-cheat). The Steam Deck Verified system is so successful that Microsoft has basically copied it. The Xbox Handheld Compatibility Program mirrors many of the same features, but for handhelds like the Asus ROG Xbox Ally, which runs Windows out of the box. With over 30,000 tested titles, Valve is approximately one quarter of the way through Steam’s total catalog, now estimated at around 120,000 games. With 20,000 new games or more per year at this point, it might seem like an impossible task to get to everything. But presumably Valve prioritizes new releases and games with high player counts, so the vast majority of Steam Deck owners have an answer for whether a game they actually want to play will work with the system. Next month will be four years since the Steam Deck’s initial release, and the impact it’s made on PC gaming is hard to overstate. It blew up the handheld PC gaming space almost singlehandedly, and competitors like the Asus ROG and Lenovo Legion series apparently haven’t been able to touch it. There’s no sign of a second-generation Steam Deck arriving anytime soon, but Valve is expanding its hardware offerings with a desktop/console hybrid Steam Machine and a new VR headset, the Steam Frame. Read...

DuckDuckGo has become the most visible search alternative to Google…and right now, we could use one. Google’s Gemini “AI” system, chopping up and spitting out results with a high degree of error and uncertainty, is what pushed me to switch to the privacy-focused search engine last year. It looks like I’m not alone: In a recent poll, 90 percent of DuckDuckGo users responded “Search without AI.” This is just a poll, and it would be fundamentally wrong to say that DuckDuckGo isn’t using any “AI” systems. In fact it has a whole-ass “AI” chatbot product, Duck.ai, which is bundled in with DuckDuckGo’s privacy VPN for $10 a month. DuckDuckGo distinguishes its self-branded setup as a way to have private conversations with powerful LLMs including ChatGPT, Claude, Sonnet, and Llama. If you’re trying to get “AI” out of every facet of your online presence, DDG seems like it’s waffling on the issue, in stark contrast to other niche tech products like Vivaldi. DuckDuckGo founder Gabriel Weinberg seems to be leaning into user choice on the topic. “At DuckDuckGo our approach to AI is to only make AI features that are useful, private, and optional,” [emphasis his], he writes on his personal site. DuckDuckGo also has two new flavors of URL for those who want to opt out or opt in to “AI” features, at noai.duckduckgo.com and yesai.duckduckgo.com, respectively. Here it should be pointed out that privacy-focused users who don’t want “AI” in their search are more likely to use DuckDuckGo, and those users are even more incentivized to vote in this sort of poll. I’m not super thrilled that DuckDuckGo is making and promoting “AI” features, while watching Google and other create so many problems. But I do appreciate that the company lets me turn off the “AI” features, flag and hide slop results (with varying degrees of success), and do it all without needing to create an account. Read...