The Information reported that an AI agent within Meta took unauthorized action that led to an employee creating a security breach at the social company last week. According to the publication, an employee used an in-house agentic AI to analyze a query from a second employee on an internal forum. The AI agent posted a response to the second employee with advice even though the first person did not direct it to do so.
The second employee took the agent’s recommended action, sparking a domino effect that led to some engineers having access to Meta systems that they shouldn’t have permission to see. A representative from the company confirmed the incident to The Information and said that “no user data was mishandled.” Meta’s internal report indicated that there were unspecified additional issues that led to the breach. A source said that there was no evidence that anyone took advantage of the sudden access or that the data was made public during the two hours when the security breach was active. However, that may be the result of dumb luck more than anything else.
Many tech leaders and companies have touted the benefits of artificial intelligence, this is just the latest incident where human employees have lost control over an AI agent. Amazon Web Services experienced a 13-hour outage earlier this year that also (apparently coincidentally) involved its Kiro agentic AI coding tool. Moltbook, the social network for AI agents recently acquired by Meta, had a security flaw that exposed user information thanks to an oversight in the vibe-coded platform.
This article originally appeared on Engadget at https://www.engadget.com/ai/a-meta-agentic-ai-sparked-a-security-incident-by-acting-without-permission-224013384.html?src=rss
Unseen Diplomacy 2 makes excellent use of roomscale play that’s let down by a finicky inventory system and a few unpolished details.
Unseen Diplomacy 2’s big party trick is “Environmental Redirection,” a roomscale design system that has players physically walking around their real-world space while the game subtly changes the environment around them. The result is that the player feels as if they’re moving through a massive continuous level without requiring a massive real-world play space. And it (mostly) works.
After setting up a roomscale boundary in-game, and once the game confirmed this boundary to be large enough to suit, I spent the next few days jumping into and out of underground bunkers, maximum-security mansions, and bases of insidious villains across the globe without ever leaving my home office.
The Facts
What is it?: A unique roomscale espionage game with fun gadgets and full-body immersion. Platforms: Meta Quest, Steam Release Date: March 16, 2026 Developer: Triangular Pixels Publisher: Triangular Pixels Price: $18.99
0:00
/1:56
Unseen Diplomacy 2 gameplay clips captured by UploadVR
Design and Mechanics
Unseen Diplomacy 2’s game field is presented as a honeycomb map, where players select individual missions involving espionage, infiltration, and other covert ops. In one stage you might be tasked with finding a stolen artifact from an ancient civilization, while another might have you hacking a terminal to disable a super-villain’s nano-bots.
The game’s overall objective is multifaceted; you’ll need to clear individual missions to acquire intel, then use that intel to uncover the location of the game’s final showdown. Reach that space before the countdown timer reaches “Zero Days Until Armageddon” and you earn your chance to save the world.
Traversing these stages demands constant physical movement, and this physicality is the core of Unseen Diplomacy 2’s gameplay. You’ll duck through crawlspaces, sidle along narrow ledges, squeeze into hiding spots, and scamper up and down ladders and ropes. At the same time, you’re sneaking past cameras and patrol bots, hacking security systems, and avoiding deadly traps such as spinning saw blades and swinging axes.
It’s worth noting that the game doesn’t strictly require roomscale play, nor excessive physical movements. There are plenty of accessibility and comfort options, such as thumbstick movement, snap-turning, button-based crouching, and even seated play. These alternative controls work reasonably well, but they also introduce a level of jank that makes the experience noticeably more tedious.
Your espionage is aided through a suite of gadgets and tools. Screwdrivers allow you to unscrew grates to access vents, or open panels to remove batteries from the security system. Wire cutters allow you to snip cables and shut down electrical systems. A blow dart helps you disable enemy patrol bots and cameras. And there’s plenty more.
Visually, Unseen Diplomacy 2 uses a bold, comic-inspired art style that will remind older gamers of the cel-shaded games of the early 2000s (XIII and Jet Set Radio). Thick, black outlines, flat colors, and exaggerated designs fit well, bringing vibrancy and life to the game’s heightened Cold-War world of secret agents and cartoonish spies.
Audio design is strong, with a soundtrack throwing back to that same Cold-War era. Horns swoon in spy motif, sneaky twangs tip-toe in your ears, while crunchy analog tech clicks, whirrs, and rattles.
The intriguing game design, interesting roomscale controls, and pleasing audio-visuals combine so that, at its best, Unseen Diplomacy 2 feels like a perfect spy fantasy. You’re not just playing a game, you’re physically sneaking, ducking and working your way through a stealthy adventure, gathering intel to save the world from certain annihilation. There’s a real thrill in that, especially early on, when each new mission feels like a fresh test of your spatial awareness, wit, and dexterity.
Diplomatic Faux Pas
The longer you play, the more the cracks begin to show. Most egregious is a general lack of polish that manifests in ways that are hard to ignore. While the core tech of Unseen Diplomacy 2, its environmental redirection and procedurally generated mission spaces, is extremely cool and works as advertised, and while movement itself feels engaging and fun, several systems layered on top of that foundation don’t always work.
The inventory system is a constant source of frustration. Tools are awkwardly mounted to your chest, wrist, and shoulders, and retrieving or storing them often feels imprecise. In theory, this physicality enhances immersion. In practice, I dropped my screwdriver more often than I used it.
Simple actions can feel clumsy while others are outright broken. Wire cutters jitter in midair, while your main offensive tool, a blow dart gun, often fails to fire for no discernible reason. And these aren’t just occasional hiccups. The problems happen frequently enough to really disrupt the flow of play.
Halfway through my first mission, it was very clear that a more traditional inventory system would have been far more effective (especially considering the inventory tools are generally pretty fun to use, when they work).
Hacking terminals can also feel tedious. These mini-games play out like the classic game Pong or Breakout, where you must bounce a ball to eliminate blocks. It’s an interesting idea that falls a bit flat, since we have to physically interact with the buttons on the terminal to move our paddle, which can be finicky.
0:00
/1:20
Unseen Diplomacy 2 gameplay clips captured by UploadVR
And then there are the more unforgivable game-breaking issues. In the few days I spent playtesting, there were several instances where levels extended beyond my established roomscale boundary in ways that the system simply could not reconcile. In a few instances this left me unable to physically progress through the mission, soft-locking me so that I had to fully reset. This is especially frustrating in a game built around immersion. When the illusion breaks, it’s hard to justify climbing back in.
Mission design also can become overly familiar, with much of the experience boiling down to squeezing through tight spaces, searching drawers or panels, retrieving some object or another, and then moving on through the uncanny environment. The structure doesn’t evolve enough, so that what begins as a novel experience gradually begins to feel like a series of procedurally generated chores.
This impression wouldn’t be quite so unfortunate, if only the game’s other systems worked perfectly. If the inventory system was better, if the environmental interaction was more polished and direct, I’d be happy to spend far more time in the game’s neat environments, be they repetitive or not.
Comfort
Unseen Diplomacy 2 features full roomscale play, plus accessibility and comfort options such as thumbstick movement, snap-turning, button-based crouching, and seated play.
Unseen Diplomacy 2 – Final Verdict
All things considered, there’s something undeniably clever about Unseen Diplomacy 2. When it all comes together, when a mission is particularly interesting and the gadgets all work and the procedurally-generated levels don’t back us into a corner, it delivers a kind of immersion that few games can match.
But the delivery of those moments is inconsistent. The lack of polish in key systems makes it difficult to fully recommend in its current state. With a couple of patches, who knows? As it stands, Unseen Diplomacy 2 is a fun diversion, one that’s just a step or two away from being truly great.
Unseen Diplomacy 2 is available now on Meta Quest and Steam.
UploadVR uses a 5-Star rating system for our game reviews – you can read a breakdown of each star rating in our review guidelines.
If you’ve been using the Internet for any length of time, you’ve probably used a tool like Google Translate to convert webpages or snippets of text to and from languages ranging from Uzbek to Esperanto. But what if you want to translate into more esoteric “languages” like “LinkedIn Speak,”“Gen Z slang,” or “horny Margaret Thatcher”?
This week, many people across the Internet have been bemused to find that the AI-powered Kagi Translate can perform these and countless other unlikely “translation” tasks. And while the collective discovery highlights the playful, creative side of large language models, it also exposes the risks of letting users play with generalized LLM tools.
What is a “language,” really?
While you might know Kagi best as the paid competitor to Google’s ever-worsening search product, the company launched its Kagi Translate tool back in 2024, saying at the time that it was a “simply better” competitor to tools like Google Translate and DeepL. At launch, the company said Kagi Translate “uses a combination of LLMs, selecting and optimizing the best output for each task,” a fact that “can occasionally lead to quirks that we’re actively working to resolve.”
In a post on its community blog today, Meta announced the timeline for the shutdown of Horizon Worlds for VR users. The Horizon Worlds app and related Events will disappear from Quest headsets by March 31, and VR users will no longer be able to use the social hub at all after June 15, 2026. Horizon Worlds will continue, Meta says, but only for mobile users.
“We are separating the two platforms so each can grow with greater focus, and the Horizon Worlds platform will become a mobile-only experience,” the company explained in the post. “This separation will extend across our ecosystem, including our mobile app.”
Horizon Worlds launched in 2021 as a VR-only platform where users could interact in a virtual space, but it was beset by early tech and design limitations (most famously, that user’s Metaverse avatars didn’t have legs). Still, at the time, the company hoped the Metaverse would ultimately attract over a billion users, and Horizon Worlds was seen as an integral part of it. Obviously, that didn’t happen—at its peak, Horizon Worlds‘ monthly user count was only around 200,000.
What happens to users’ digital purchases and creations?
Long-time Horizon Worlds users are surely asking what will happen to their in-world digital items. The good news is that your Worlds-specific purchases and creations won’t instantly be wiped; Meta says your digital items or currency will remain tied to your account. The bad news is you’ll only be able to access them through the mobile app in “mobile-optimized” worlds, so if a creator hasn’t updated their world for mobile, your items there may effectively become inaccessible.
After June 15, you will no longer be able to build or edit worlds in VR. Meta is encouraging the use of their web-based tools, but the immersive building experience that defined the platform is officially ending.
Meta is shifting focus from VR to AI
The change is part of an overall strategy shift from Meta, which will be funneling more resources into AI and smart glasses. In January, Meta shuttered its AAA VR game development companies, stopped updating its first-party subscription-based fitness app Supernatural, and laid off 1,500 people from its VR-division Reality Labs.
Even given all that, though, Meta says it’s not pulling out of the VR game altogether. In a blog post on Feb. 19, Samantha Ryan, the vice president of content for Reality Labs, promised Meta was “doubling down on VR,” but is nevertheless moving away from first-party development to focus on hardware, supporting third-party developers, and adding features to the Quest itself. “It’s no secret that we’re still in the hardware game,” Ryan wrote. “We have a robust roadmap of future VR headsets that will be tailored to different audience segments as the market grows and matures.”
The kinda sad end of Horizon Worlds
I poked around Horizon Worlds for when I got my Quest 3 headset a few years ago. “Oh, I can decorate a little house or meet people” I thought; then I logged out and never went back—I have a real house I can decorate, and I use VR because I don’t like people. But a few months ago, when it became clear that Meta was pulling away from the VR space it created, I got curious, strapped on the face computer, dusted off the old avatar, and went on a Horizon Worlds safari. I’m glad I did.
Going to Horizon Worlds feels eerie. It’s like visiting someone else’s dreams—specifically Mark Zuckerberg’s dream and the sub-dreams of Worlds’ volunteer creators. “The defining quality of the metaverse will be a feeling of presence…Feeling truly present with another person is the ultimate dream of social technology. That is why we are focused on building this,” Mark Zuckerberg said at Connect 2021, and he believed it enough to spend billions (maybe as much $25 billion) on his dream word, where nothing ever gets dusty and everyone is an endlessly smiling cartoon ready to funnel actual money to The Overseer for the latest digital sneakers.
Then there are the thousands of creators who spent countless hours building more than 10,000 worlds you can visit—nightclubs, basketball stadiums, restaurants, etc., though they’re almost all empty. As flashy as it looks, no one is lined up at the digital nightclub’s velvet rope. Horizon Worlds is a gigantic dead mall, a capitalist cathedral with no congregation, the ultimate liminal space.
A vibe that strange is enough reason to visit, but there’s a genuine side to Worlds too. I eventually found a world where people conquered the unsettling nature of VR and created a real community. The Soapstone Comedy club is not huge, but it’s thriving, and it’s one of many spots where small groups have gathered in Horizon Worlds. There are conversation pits people use to meet up, nightly planned comedy shows, and a collection of friendly regulars to chat with. It grew from the ground up, too, just as Zuck predicted. Okay, the comedy is rarely funny. But the people are good.
Some Soapstoners are housebound and handicapped, and VR gives them opportunities the real world denies them. Some seem like weirdos who probably have trouble finding real life friends who can’t mute them at will. And some are regular people blowing off some steam after work. I’m sad for all of them—a comedy club on a phone screen just isn’t the same. So before the digital wrecking ball takes down the hang-out spots of all Horizon Worlds‘ remaining residents, you should stop in and say hi. There’s not much time left.
With today’s latest Stitch updates, Google is trying to make “vibe design” happen, reports The Verge’s Jay Peters. The AI-native design platform encourages users to describe goals, feelings, or inspiration in “natural language,” rather than starting with traditional blueprints.
In a blog post, Google Labs Product Manager Rustin Banks says that Stitch can turn those inputs into interactive prototypes, automatically map user flows, and support real-time iteration. It introduces voice capabilities that allow users to “speak directly to [the] canvas” for feedback or changes. Tools like DESIGN.md also help users create reusable design systems across various projects.
Because it’s all in the reflexes, this is the heavily requested Screen Junkies honest trailer for Big Trouble In Little China. It’s over six minutes long though, which is 1/20th of the entire movie. Not that I’m complaining, I could quote Big Trouble In Little China all day. And do most days. My girlfriend doesn’t always like it, but *shrug* sooner or later I rub everybody the wrong way. “Nothing or double?” The check is in the mail.
The European Union may soon ban nudify apps after Elon Musk’s chatbot Grok emerged as a prime example of the dangers of an AI platform failing to block outputs that sexualized images of real people, including children.
In a joint press release, the European Parliament’s Internal Market and Civil Liberties committees confirmed that lawmakers voted 101–9 (with 8 abstentions) to simplify the Artificial Intelligence Act and “propose bans on AI ‘nudifier’ systems.”
The vote came after the European Commission concluded earlier this year that the AI Act does not prohibit “AI systems that generate child sexual abuse material (CSAM) or sexually explicit deepfake nudes.” At that time, the Commission signaled that Parliament members were already proposing ways to amend the law to strengthen protections against such harmful content.
In the US, the economics of coal power generation are marginal at best, and a large number of coal plants have shut down as cheaper renewables and natural gas have surged. The Trump administration has used a number of methods to swim against this economic tide, the simplest of which has been to order plants scheduled for closure to remain operational.
The Department of Energy has used the Federal Power Act and a Trump executive order declaring an energy emergency to block the closure of coal plants nationwide. The orders requiring plants to stay open have been accompanied by a steady stream of triumphal press releases, suggesting that the Department of Energy was taking the step solely to ensure grid reliability.
The latest of these releases, issued on Monday, pertains to a plant in Centralia, Washington, that was scheduled to close last year to be converted into national gas generation. A Department of Energy emergency order had kept it operational over the winter, but that order was set to expire yesterday. With yesterday’s new order, the plant will remain operational through mid-June. According to the press release, the action was taken “to ensure Americans in the Northwestern region of the United States have access to affordable, reliable, and secure electricity.”
Longtime Slashdot reader UnknowingFool writes: Users of Samsung PCs are reporting the inability to access the C: drive after the Windows 11 February update. The bug seems to be in connection with the Samsung Galaxy Connect app, which allows Samsung phones and tablets to connect to Windows machines. [A previous stable version of the app has been re-released to prevent this problem from spreading.] This parody explains the situation with humor. The issue stems from update KB5077181 and is impacting Samsung PCs running Windows 11 25H2 or 24H2. Microsoft and Samsung have confirmed the issue and published a workaround, but as PCWorld notes, it will take some time. The workaround “requires removing the Samsung application, then asking Windows to repair the drive permissions and assigning a new owner, then restoring the Windows default permissions, including patching in some custom code that Microsoft wrote.”
Strong forces tempting humans out of the AI loop, and reducing the experience needed to supervise and reviewQCon London AI is in a dangerous state where it is too useful not to use, but where by using it, developers are giving up the experience they need to review what it does, said a speaker at QCon London, a vendor-neutral developer conference underway this week.…
Neanderthals may have used birch tar as more than just glue; it could have helped them ward off infection and even insect bites.
People from several modern Indigenous cultures, including the Mi’kmaq of eastern Canada, use tar from birch bark to treat skin infections and keep wounds from festering. We know from several archaeological sites that Neanderthals also knew how to extract birch tar and that they used it as an adhesive to haft weapons. A recent study tested distilled birch tar against the bacteria S. aureleus and E. coli and found that Neanderthals could easily have used the same material as medicine for their frequent injuries.
This is the simplest step-by-step tutorial for making birch tar: find a tree, set some bark on fire, get messy hands.
Credit:
Tjaark Siemssen, CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)
Medicine can be messy
What we call “birch tar” in English has a lot of other names in multiple Indigenous languages, and it can range from an oily fluid to a brittle, almost solid tarry resin, depending on how long you heat it in the open air after extracting it from the bark. The Mi’kmaq of eastern Canada prefer the more fluid version, which they call maskwio’mi, for wound dressings and skin ointment.
If there’s one thing PC gamers universally despise, it’s shader compilation stutters, but it’s also a drag to launch a brand-new game, ready to play, only to be greeted by a “Compiling Shaders” progress bar slowly crawling across the screen. The end of the shader struggle is finally here, and it’s being spearheaded by a massive industry push
The GlassWorm malware made news when it pivoted from exclusively targeting Windows users to also targeting Mac OS users in January, and in the time since, the malware campaign has spread across at least 413 code repos on npm, VSCode, OpenVSX, and even GitHub. Evidence points to this all being tied to a single threat actor since they’re all
Google and cybersecurity companies Lookout and iVerify have detailed a new hacking technique that potentially puts a significant portion of iPhone users in danger, just by visiting the wrong web page. The hack is called “DarkSword” and since it specifically targets several different versions of iOS 18, it could affect “close to a quarter of iPhones,” Wired writes.
DarkSword is a “fileless” hack that leverages a collection of exploits to access sensitive data when an iPhone visits an infected website. Rather than install spyware that hangs around on a user’s phone after messages and other private information are stolen, fileless hacks like DarkSword take control of “the legitimate processes in an iPhone’s operating system to steal data,” according to Wired. Even more troubling, DarkSword deletes any evidence it was running on an iPhone after it finishes stealing your information.
The hack starts as soon as an iOS device encounters an “malicious iframe embedded in a web page,” after which it works its way through your iPhone, gathering sensitive information like passwords before deleting itself. DarkSword can abscond with things like messages and iCloud content, but it’s also specifically designed to access crypto currency wallets, Lookout says, which could indicate who was using DarkSword before it became widely available.
DarkSword has reportedly been used in Ukraine, Saudi Arabia, Malaysia, Turkey and Russia, and its origins could be tied to a different hacking toolkit called Coruna that TechCrunch reports may have been created for the US government by a company called Trenchant. Regardless of where DarkSword came from, the tool didn’t become widely available until its Russian users left DarkSword’s source code on a website for anyone to access, “complete with explanatory comments in English that describe each component and include the ‘DarkSword’ name for the tool,” Wired writes.
Apple patched the exploits that DarkSword and Coruna used in recent updates to iOS 26, the yearly software release from 2025 that followed iOS 18. The problem is that not everyone is using Apple’s latest update. DarkSword targets iOS 18 releases between iOS 18.4 and iOS 18.6.2, and according to Apple’s latest iOS usage stats for developers, around 24 percent of iOS devices are still on iOS 18. Without more detail, it’s hard to know how many people that leaves exposed, but as a rule of thumb, if your iOS device can update to a newer software release, you should do so as soon as possible to stay secure.
This article originally appeared on Engadget at https://www.engadget.com/cybersecurity/a-new-iphone-hacking-tool-puts-anyone-still-on-ios-18-at-risk-203745949.html?src=rss
We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.
With Amazon’s annual Big Spring Sale right around the corner, we’re starting to see prices drop on Amazon devices. That includes the newest model of theEcho Spot, which combines smart home and speaker perks with an alarm clock. Right now, there’s a 38% discount on this device, bringing it to $49.99 (down from $79.99).
The Echo Spot is a reasonably compact smart alarm clock. In fact, the screen is slightly smaller than you’d expect. Although it looks like it takes up half of the device, the screen itself is only a 2.83-inch square portion of the entire half-circle panel. Despite its size, the display does show a number of different data points, including the time, weather, and calendar events. It’ll show you the name of any songs that are currently playing, and you can set it to transition to your favorite music after the alarm goes off.
Like Amazon’s other Echo devices, this is compatible with Alexa for hands-free use. If the rest of your home is part of the Alexa ecosystem, you can use the Echo to make calls or set up custom routines like turning off all the smart lights in your home before you go to bed. As PCMag notes in its review, the speaker quality is surprisingly loud and dynamic for a smart alarm clock, and while you might prefer a larger speaker, this gets the job done. PCMag also mentions that when streaming over wifi, the audio quality is better than when connecting your phone via Bluetooth). Importantly, the Echo Spot lacks a camera and cannot play video.
The Echo Spot is a simple, easy-to-use smart clock that also doubles as a casual speaker with basic smart home features (not to mention Alexa compatibility). If that’s what you’re looking for, the Amazon Echo Spot is a worthwhile buy, especially at its current $49.99 price point.
Our Best Editor-Vetted Amazon Big Spring Sale Deals Right Now
The White House has been promising a set of national rules to guide artificial intelligence since late last year, and today Sen. Marsha Blackburn (R-Tenn.) fired the first volley. The senator shared a discussion draft for codifying the executive order signed by President Donald Trump in December calling for an AI bill. Her stated goal is a policy that “protects children, creators, conservatives and communities from harm.”
Blackburn has called for tougher policies for AI safety, and one of the core messages in this discussion draft is that it “places a duty of care on AI developers in the design, development and operation of AI platforms to prevent and mitigate foreseeable harm to users.” It also draws a line on the many copyright infringement questions raised by creative industries: “an AI model’s unauthorized reproduction, copying, or processing of copyrighted works for the purpose of training, fine-tuning, developing, or creating AI does not constitute fair use under the Copyright Act.”
Some of the other notable provisions are:
Requires covered online platforms, including social media platforms, to implement tools and safeguards to protect users under the age of 17 against online harms.
Protects the voice and visual likenesses of individuals and creators from the proliferation of digital replicas without their consent.
Sets new federal transparency guidelines for marking, authenticating and detecting AI-generated content.
Requires certain companies and federal agencies to issue reports on AI-related job effects, including layoffs and job displacement to the U.S. Department of Labor (DOL) on a quarterly basis.
It includes ending Section 230, marking the latest attempt to retire a law that has been questioned as a possible loophole for AI companies to escape liability when their tools cause harm. While AI critics might see positive signs here, remember that this is just the initial version of the framework. Lawmakers will likely spend a lot of time negotiating over the eventual result, which may be notably de-fanged from its current state. It could wind up with a lot more requirements echoing this Republican complaint: “Combats the consistent pattern of bias against conservative figures demonstrated by AI systems by requiring third-party audits to prevent discrimination based on political affiliation.” Despite the claims of suppression and censorship, we’ve consistently seen this conservative argument to be false — or at the very least misleading.
This article originally appeared on Engadget at https://www.engadget.com/ai/senator-blackburn-introduces-the-first-draft-of-a-federal-ai-bill-202509852.html?src=rss
