System Admin Stuff

Flashback: The 2016 Chrome and Android‑apps story that predicted today’s Googlebooks moment

2026-05-23T00:57:06.684+05:30

A 2016 warning about Android apps on Chrome feels newly relevant as Googlebooks pushes Android PCs into the agentic era and challenges Microsoft again.

source https://www.windowscentral.com/microsoft/windows-10/flashback-the-2016-chrome-and-android-apps-story-that-predicted-todays-googlebooks-moment

Migrating from Ubuntu 16.04 to FreeBSD

2026-05-23T00:38:23.785+05:30

Bruno Croci’s blog had been running on Ubuntu 16.04 for a long time, well past the Linux distribution’s expiration date. As such, it was time to upgrade, but instead of opting for something standard like another Ubuntu release, he opted for FreeBSD instead.

This blog has been running on a Digital Ocean VPS for over ten years. A machine hosted in New York City, running Ubuntu 16.04 LTS. An LTS that hasn’t been in support for at least 5 years. It was about time to change it. After some considerations, I migrated to a Hetzner virtual machine that is way better than my old Ubuntu one, less than half the price of what I used to pay, and just across the country from me. Not only that, but I took the challenge to move my stack to FreeBSD. It’s a long text, but stay for a cool introduction of FreeBSD Jails with Bastille and some interesting site load benchmarks.
↫ Bruno Croci

I absolutely adore the recent surge in people (re)discovering the BSDs as a valid alternative to Linux in both the server and desktop space. In this particular case, it was FreeBSD’s Jails and ZFS support that won Corci over, and it’s easy to see why. While there are countless alternatives to Jails in the Linux world, ZFS is harder to come by as it can’t be part of the kernel due to licensing issues. With how powerful and capable ZFS is, it makes sense to want to use it on your server, and in that case, FreeBSD is probably a better choice than most Linux distributions.

There are countless reasons to choose one of the BSDs over a Linux distribution, and I’m glad we’re seeing an uptick.

source https://www.osnews.com/story/145056/migrating-from-ubuntu-16-04-to-freebsd/

Soaring solar and a surge in hydro push more coal off the US grid

2026-05-23T00:13:37.985+05:30

Last year, the first few months of data from the US grid suggested that fears of a data-center-driven surge in demand were becoming a reality. Demand had risen by about 3 percent, triggering a surge in coal, interrupting what had been a long downward trend. But over the course of the year, both trends slowed considerably.

A year later, all of that seems to be in the past, as the US has returned to its normal pattern: slow growth, with renewables pushing coal off the grid. The one oddity is that hydroelectric production has surged without a corresponding increase in capacity, likely due to unusually warm weather in the western US causing the snowpack to melt early. That may have consequences later in the year.

Pushing fossil fuels out

Overall demand in the US grew by only 1.5 percent in the first quarter of 2026 compared to the same period the year before. Often, changes in demand during this part of the year are driven by weather-related heating demand. But the US had an unusual combination set of weather conditions to start 2026, with the western half baking in unseasonal warm temperatures, while the eastern half suffered a deep freeze. So we'll probably need data from more of the year before we read too much into the small rise in demand we've seen so far.

Read full article

Comments

source https://arstechnica.com/science/2026/05/soaring-solar-and-a-surge-in-hydro-push-more-coal-off-the-us-grid/

Trump abruptly cancels EO signing event after top AI firm CEOs declined to go

2026-05-23T00:13:37.553+05:30

President Donald Trump abruptly canceled an event on Thursday just hours before he was scheduled to sign an executive order granting the government the power to test frontier AI models before their public release.

As The New York Times explained, Trump had been hoping that top executives from leading AI firms would attend the signing. He decided to pull the plug after learning that some CEOs couldn't make the event. That made Trump unhappy, even though he'd only given them 24 hours' notice. Other AI executives who quickly rearranged their schedules to go "were midair on their way to the Oval Office" when they found out that the trip was for nothing.

Reporting from Semafor indicated that OpenAI "supported" the signing. However, xAI founder Elon Musk and Meta CEO Mark Zuckerberg reportedly helped "derail" the executive order, supposedly urging Trump to "call it off." Additionally, Trump's former AI advisor David Sacks—whose special government employee designation expired in March, The Information noted—joined the push to delay the signing, Semafor reported.

Read full article

Comments

source https://arstechnica.com/tech-policy/2026/05/trump-canceled-ai-safety-testing-eo-after-snub-from-tech-ceos/

Deus ex machina: Half of US Christians trust AI's spiritual advice

2026-05-22T00:33:55.725+05:30

Who needs a minister when you have an LLM? America’s Christian population appears to have found God in precisely the place you’d expect a manifestation of the divine to be spotted in 2026: Amid AI chatbot responses. A survey of Americans published this week by Evangelical polling outfit Barna sought to discover what Christians thought about AI’s ability to serve as a spiritual mentor, and the split is surprisingly even: A full 48 percent of practicing US Christians told the organization that they trusted AI’s advice to aid their spiritual growth. Potentially more surprising than that, 34 percent said spiritual advice dispensed by an AI was just as trustworthy as what they'd get out of a flesh-and-blood pastor. That share rises, unsurprisingly, among younger Christians, with 39 percent of Gen Z respondents and 44 percent of Millennials agreeing that preachers and AI are at trust parity. Pastors themselves, it likely won’t surprise you to learn, are splitting sharply from their flocks on the matter of AI’s ability to fill their roles in the lives of congregates, with just 12 percent saying they agree that AI can help people grow spiritually. That said, there’s a pretty serious tension among American Christians when it comes to AI. At the same time half say it’s aiding their spiritual journeys, most also expressed concerns about negative effects of AI on spirituality. A full 83 percent of practicing US Christians believe AI is likely to misinterpret scripture, 73 percent are worried AI will cause loss of religious faith, and 72 percent believe that AI is beginning to act as a replacement for God and earthly spiritual leaders. “Christians say they trust AI with spiritual growth, and a meaningful share say its spiritual guidance is as trustworthy as a pastor’s—yet large majorities are simultaneously concerned about AI misinterpreting scripture, replacing God, or undermining the role of spiritual leaders, Barna VP of research Daniel Copeland said of the findings, which he called “confounding.” “That level of openness is higher than we might have expected,” Copeland added in the Barna’s report. Worshipping at the altar of Altman AI and religion have been butting heads for the past couple of years, with the Catholic Church particularly outspoken about the technology. The late Pope Francis called on world governments to establish global AI regulations in 2023, as well as calling on people to avoid turning to AI models for moral and ethical decisions. Vatican AI authority Friar Paolo Benanti later accused Silicon Valley elites of playing God with their creations, AI included, noting that “the focus will always be on using AI for profit,” which - according to the good book itself - isn’t compatible with Christianity. That hasn’t stopped some of God’s faithful from creating an AI Jesus and Christian AI platforms, and the new Pope, Leo XIV, has continued his crusade against the technology. "By simulating human voices and faces, wisdom and knowledge, consciousness and responsibility, empathy and friendship… artificial intelligence [could] not only interfere with information ecosystems, but also encroach upon the deepest level of communication, that of human relationships,” Leo said earlier this year. Leo further expressed worry that AI was turning people into “passive consumers of unthought thoughts,” and that’s not even to touch on the fact that AI has a tendency to make stuff up to appease its questioners, potentially leading the spiritually curious into full-blown episodes of psychosis aided by a digital yes-man in the guise of an authority. ®

source https://www.theregister.com/ai-ml/2026/05/21/deus-ex-machina-half-of-us-christians-trust-ais-spiritual-advice/5244371

Google accidentally exposed details of unfixed Chromium flaw

2026-05-22T00:24:33.678+05:30

Google has accidentally leaked details about an unfixed issue in Chromium that keeps JavaScript running in the background even when the browser is closed, allowing remote code execution on the device. [...]

source https://www.bleepingcomputer.com/news/security/google-accidentally-exposed-details-of-unfixed-chromium-flaw/

Strava Just Overhauled Its Strength Activity Experience, and It Looks Like a Major Upgrade

2026-05-22T00:15:17.888+05:30

Strava announced today a complete overhaul of its strength training experience. It's one of the bigger gym-based updates I've seen from the app—the new features include 14 partner integrations, a dedicated workout log, auto-populated muscle maps, and five new strength-specific shareables. According to Strava, the update is rolling out globally in the coming weeks, so you should be able to try all of these new goodies soon.

For anyone who has ever wished Strava treated their weight training with even a fraction of the seriousness it gives to runs and rides, the update certainly looks promising. Whether it fully delivers on its potential is a question that will require some testing to answer, given that most strength apps look better on paper than they actually work in practice.

Strava's workout log is getting a full refresh, including muscle maps

The core update is a completely refreshed workout log that lets you record sets, reps, and weight directly inside Strava. In theory, this could close a gap for all of the Strava runners out there who also lift, but lack a reliable way to keep that training data in the same place as everything else.

For me, the most eye-catching feature is that every logged strength workout will now automatically generate a visual muscle map highlighting which muscle groups were worked based on the exercises recorded. This could be a great way to understand your training balance, avoid overuse, and make sure you're not accidentally skipping the same muscles every week. Of course, a muscle map is only as good as the underlying exercise data, and a factor there could be how well the different partner apps feed into Strava. .

These partner apps now work with Strava

Strava is launching with 14 partner integrations, pulling strength data from apps and devices across the fitness ecosystem. The initial partners include: 24 Hour Fitness (coming this summer), Amazfit, Caliber, Coros, Fitbod, Garmin, Hevy, iFIT Personal Trainer, JEFIT, Liftoff, Motra, Remaker, Runna, and Whoop.

I'll be covering these integrations in more depth once I have a clearer picture of exactly what data each one sends across and how it surfaces inside Strava. But a few things stand out to me. The Hevy integration seems particularly promising. Hevy is a well-regarded dedicated strength logging app, and pulling detailed workout data from it directly into Strava could be a welcome consolidation for people already using it.

For Coros and Garmin users, the integration should mean strength sessions tracked on your watch will flow into Strava automatically, just like your runs do now. I'm curious to dig into exactly what that looks like in practice—how exercises are classified, whether sets and reps come through cleanly, and whether the muscle maps populate correctly from watch-captured data versus manually logged workouts.

The Whoop integration is also worth noting, as Whoop just added expanded strength training features of its own. Given Whoop's status as a launch partner, I imagine there are plenty of ways the two platforms complement each other—say, with Whoop's recovery and strain data sitting alongside Strava's activity log and social elements. And on that note, Strava is also adding five new strength-specific share formats, designed to give gym workouts the same social motivation juice that outdoor activities have always had on the platform.

This New Flipper Device Is Like a Pocket-Sized Linux PC

2026-05-22T00:15:17.120+05:30

Flipper Devices, the company behind the semi-infamous Flipper Zero "hacking" multi-tool, is developing a Linux-powered mini-PC with the goal of creating a "truly open hardware platform." The Flipper One is described as a pocket-sized ARM device for high-performance computing, IP networking, and on-device AI use. The device is in development, so there is no price or release date as yet.

The Flipper One is being built on a Rockchip RK3576 processor and the modular design means it can be expanded. "Flipper Zero taught us how much you can do with a tightly scoped, open product and a community that pushes it further than you can. Flipper One is what happens when we apply the same approach to a much bigger problem—building a fully open ARM Linux device that doesn't go obsolete the moment it ships," Pavel Zhovner, Co-founder and CEO of Flipper Devices, said in a press release.

In keeping with the Flipper "open everything" ethos, the Flipper One is a community developed project, and anyone who wants to can check in and/or help out through the Developer Portal.

What you will be able to do with a Flipper One

Network debugging: The Flipper One will feature high-speed connectivity, including Ethernet, wi-fi 6E, and optional 5G, allowing it to function as an advanced network debugging tool.

On-Device AI: The Flipper One will have local AI hardware acceleration, so it will be able to handle compute-heavy tasks without needing a cloud connection.

Wireless analysis: The device will capture and analyze wireless traffic and network signals in real time.

Who is the Flipper One for?

If you work in network administration, you probably already know why you need or don't need a Flipper One, but if you're just a casual tech enthusiast or tinkerer, it might provide a powerful sandbox. Here are some potential uses:

As a retro console: Presumably, it will have enough power to run Linux video game emulation software, so you could connect a Bluetooth controller, plug into a TV, and have a powerful portable gaming rig in your pocket.
As a home server: You could use this to run a network-wide ad-blocker, or use it as a portable media server that could be accessed by any device on your wifi.
As private AI: Since it can run artificial intelligence, you can experiment with LLMs in complete privacy.
To learn about the devices around you: We are surrounded by Bluetooth beacons, wifi probes, and radio signals we never think about. This will give you some insight into that invisible world.

Oh boy, it's real: the Kingdom Come: Deliverance 2 dev is making a new The Lord of the Rings RPG in Middle-earth — and a third KCD game is coming next year, too

2026-05-21T00:35:15.974+05:30

Kingdom Come: Deliverance 2 developer Warhorse Studios is officially working on a The Lord of the Rings RPG, and also a new KCD game that's coming soon.

Fedora: Microsoft is all aboard, but Deepin is dumped

2026-05-21T00:33:21.627+05:30

Microsoft has announced a new, Fedora-based Linux distro for Azure VMs, while Fedora has consigned the Deepin desktop to the bin. Fedora decided to remove a component maintained outside Red Hat. In the same week, another external company - granted, a slightly better-known one - decided to rebase one of its projects onto Fedora as its upstream distro. It’s the circle of life, or something. Fedora 💔Deepin Seven years after it added the Deepin Desktop Environment in Fedora 30, Tuesday's FESCo meeting decided to drop Deepin from the distro. The minutes say: AGREED: Retire all packages maintained by the deepinde-sig group The decision comes one year after the project called for a security review of the Deepin Desktop Environment, after openSUSE dropped the desktop following a negative security assessment. We reported on that decision at the time. SUSE asked Deepin for feedback, but didn’t get good enough answers – for which, some months later, the Chinese project issued an apology. Linux Deepin is very much still around: we most recently looked at version 25 in January, and, back in 2023, the project claimed it had passed three million installs of its paid Tongxin UOS desktop edition. It’s a very pretty Windows-like desktop environment, but it never made it to having its own Fedora spin – and it certainly won’t now. Microsoft ❤️ Fedora But as one door closes, another opens. Fedora is still winning new friends and allies, and mere days earlier, there was a surprise announcement at the Open Source Summit North America, which as we write is winding down. On Monday, Microsoft announced a new version of its in-house Linux distro, Azure Linux 4, along with a companion distro called Azure Container Linux. There have been products called Azure Linux for quite a while. It’s based on the much more minimal CBL-Mariner distro, which we tried in 2022. The Register reported on Azure Linux becoming generally available in 2023, and then on the release of version 3 in 2024. We also knew back then that the company was working on turning it into a more general-purpose server OS: we reported on it migrating LinkedIn to Azure Linux in place of CentOS Linux that same year. There isn’t very much information about Azure Linux 4 yet; the broader rollout will be at the Microsoft Build conference next month. For now, all you can do is fill in a form to register your interest. However, the announcement reveals that version 4 switches to Fedora as its upstream distro. It was already based on the RPM packaging tools, hinting at some Red Hat or SUSE heritage in there somewhere. There’s slightly more information about Azure Container Linux. This is a separate distro, an immutable host OS for running containers. The announcement says “Azure Container Linux is based on the Flatcar project.” Flatcar is the continuation of CoreOS Container Linux, which The Reg has covered since it released its first version in 2014. As Linux Weekly News reported that year, CoreOS was based on Google’s ChromeOS, but redesigned to host containers. Red Hat acquired CoreOS in 2018, and then two years later, discontinued Container Linux. It replaced the Google and Gentoo-based distro with a new one based on Red Hat’s own immutable tool chain, called Fedora CoreOS. German FOSS consultancy Kinvolk forked the CoreOS code and continued development under the name of Flatcar Container Linux. Kinvolk was acquired by Microsoft in 2021, but continued to work on Flatcar. Now it seems that, with the announcement of Azure Linux 4 as well as Azure Container Linux, Microsoft has two separate in-house distros: one based on Fedora, and one based on ChromeOS. For now, Flatcar is still trundling along the tracks just fine, but we suspect some future consolidation may be coming down the line. ®

source https://www.theregister.com/oses/2026/05/20/microsoft-rebases-azure-linux-on-fedora-as-fedora-drops-deepin/5243629

Sony's Signature WH-1000XM5 Headphones Are $150 Off Right Now

2026-05-21T00:15:35.303+05:30

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

The Sony WH-1000XM6 headphones are the most recent entry in Sony's 1000X series, which has been around since 2016. They're the best over-ear headphones in terms of audio quality currently on the market—but not everyone can (or needs to) drop nearly half a grand on headphones (though they're currently at their lowest price right now with an additional $30 gift card).

The WH-1000XM5 are a generation older, and were at the top of their class until they were supplanted by the XM6. On the plus side, by virtue of being older, they're currently discounted to $248 (down from the original $399.99). That's $5 off the lowest price they've ever reached, according to price-tracking tools, a good deal for cans that still sound incredible, even in 2026.

I have been a loyal customer of the 1000X line for many years; they're my go-to headphones for most activities. The WH-1000XM5 came out to an "outstanding" review from PCMag for their top-notch audio quality and exceptional audio when using the best-in-class active noise-canceling. The headphones are also well-designed to be comfortable for long listening sessions. The ear controls use tapping and swiping, which isn't my favorite, but it's what all modern headphones are moving toward. There's a companion app that lets you adjust your EQ settings to your liking, including what the swiping and tapping functions do.

These headphones also have a stereo 3.5mm connection, which is perfect for those who want to use a cable and not worry about battery life. And speaking of battery life, Sony says you can expect about 30 hours, but it will vary depending on your use of ANC. They are compatible with AAC, LDAC and SBC codecs and have multipoint connection, meaning you can pair them with more than one device at the same time.

Our Best Editor-Vetted Tech Deals Right Now

Apple AirPods 4 Active Noise Cancelling Wireless Earbuds — $148.99 (List Price $179.00)

Apple Watch Series 11 (GPS, 42mm, S/M Black Sport Band) — $329.00 (List Price $399.00)

Apple iPad 11" A16 128GB Wi-Fi Tablet (Silver, 2025) — $299.00 (List Price $349.00)

Fire TV Stick 4K Plus Streaming Player With Remote (2025 Model) — $29.99 (List Price $49.99)

Sonos Move 2 — $399.00 (List Price $499.00)

Sony WH1000XM6- Best Wireless Noise Canceling Headphones — $398.00 (List Price $459.99)

Ring 2nd Gen 2K Wired Video Doorbell (2026 Release) — $49.99 (List Price $79.99)

Deals are selected by our commerce team

Gemini Is Getting a UI Overhaul and Embracing Its Agentic Side

2026-05-20T00:15:25.378+05:30

Google's Gemini app is getting an overhaul, in terms of both design and function. At Google I/O 2026, the company announced Gemini is getting upgraded models that will lean more into agentic AI capabilities, which allow the assistant to send emails, add calendar events, and perform other actions on your behalf.

(Google does emphasize that all agentic capabilities are designed to be used under your supervision, which hopefully allays fears of the agent going rogue and, I don't know, draining your bank accounts.) Here's is everything Google announced about Gemini's app upgrades during this year's event.

Gemini has a new look

Google says it has revamped the design language for the Gemini app. The company calls it Neural Expressive, and it includes new animations, colors, typography, and refreshed haptic feedback. Another change: The Gemini Live conversational experience is now integrated into the Gemini app. This means you can easily switch from typing your questions to using your voice to talk to the AI.

There are also changes to the voice chat mode, allowing you to tap and talk to Gemini at your own pace. This prevents unwanted interruptions and doesn't force you to keep speaking at a certain speed if you need more time to frame your thoughts, according to Google. You'll soon also have an option to ask Gemini to speak in regional dialects, which can offer a more personalized experience to some people.

I think the best part of Neural Expressive design is Google's focus on making interacting with Gemini look like something more than a wall of text. The company claims Gemini will now show responses with images, interactive timelines, narrated videos, and dynamic graphics, instead of just huge blocks of text. This new design language is rolling out today and will be available on the web, Android, and iOS.

Turn Gemini into a personal AI agent with Spark

Gemini Spark marks Google's big play on agentic AI services. It transforms Gemini from a simple word generator into an assistant that can take actions for you. This feature works even if you've switched off your laptop or phone, and it can continue to work independently based on your requirements. Some examples Google shared during I/O 2026 include the ability to set recurring tasks or triggers, such as scanning your credit card bills to flag new or hidden subscription fees, go through your meeting notes to create a "polished" document in Google Docs with a summary of its findings, and drafting a companion email to kickstart a project. Google says you'll also be able to teach Gemini Spark to pick up on deadlines for school projects and share them with your spouse, which could allow you to keep a tab on your child's schoolwork without actively monitoring their inbox every day.

It's worth noting that Google's repeatedly emphasized the fact that Spark works on your commands. You'll have to choose which apps it can connect to, and it'll only take high-stakes actions with your explicit consent, whether that's spending money or sending an email.

Gemini Spark is rolling out to Google AI Ultra subscribers in the U.S. next week, but it will be available to "Trusted Testers" this week. Google also said that Gemini Spark will is coming to the Gemini app for macOS later this summer, allowing you to automate workflows involving local files on your computer.

Daily Brief creates a personalized morning digest for you

Gemini is getting a new morning digest feature called Daily Brief, which works only if you opt in to receive it. The service accesses your Gmail inbox and your calendar to give you an overview of what's ahead in your day. Google says this feature will prioritize your tasks based on your goals, and suggest the next steps you should take to tackle upcoming tasks. You'll have the ability to tune the AI service's feedback by giving it a quick thumbs up or down in the app. The Daily Brief feature is rolling out in the U.S. starting today, and it'll be available to Google AI Plus, Pro, and Ultra subscribers.

Turn ideas into cinematic videos with Omni

In a move to fill the void left by OpenAI's shuttering of Sora, Google is launching Gemini Omni, which allows you to combine text, images, and video inputs to generate videos. The company claims Omni makes it easy to edit videos using conversational inputs, such as swapping out backgrounds or applying cinematic zoom effects. You'll be able to upload footage from your camera roll and use Omni to apply built-in templates to create videos quickly without needing editing software expertise. There's also an option to create a custom AI avatar that looks and sounds just like you, and drop this avatar in your videos.

This feature does suggest a few ethical concerns, since it sounds like it would make it easy for a person with malicious intent to generate deepfakes. While it's nice to see that video editing is becoming more accessible, generating videos with real people's faces is in a grey area, and I hope that Google has strong guardrails in place to safeguard against abuse. Gemini Omni is rolling out starting today to Google AI Plus, Pro, and Ultra subscribers across the globe.

Google Just Announced a Bunch of Workspace Features at I/O 2026

2026-05-20T00:15:24.581+05:30

Google Workspace is getting a slew of AI-related updates, including a brand new app called Google Pics, which can generate and edit images for you. Many of these updates will be available to those who use Gmail, Google Docs, Google Drive, and other Google Workspace apps, the company announced during the Google I/O 2026 keynote. This includes conversational features that will allow you to control Gmail, Docs, and Keep; a new AI inbox that aims to help you stay on top of your email; and a personal AI agent called Gemini Spark that can integrate with Google Workspace apps to get things done for you.

Here are all the major Google Workspace updates unveiled during Google I/O 2026. These features are rolling out for Google AI subscribers and paying Google Workspace users. This means that you won't be able to try much if you're on the free tier of Google services.

Create and edit images with AI in Google Pics

Google Pics is a new app built with the company's latest Nano Banana AI image generation model, and allows you to create and edit photos using AI. Google claims this tool makes useful image editing tools more accessible, highlighting examples such as modifying and translating text, as well as transforming specific elements in an image (e.g., changing a sweater's color or turning a dog into a cat). You can also generate an image with a text prompt, but the real highlight seems to be granular editing.

Like every other AI-related image generation tool, Google Pics also raises ethical questions about the nature of an image. It's not hard to imagine how tools like these could be misused to manipulate and deceive. One can only hope that Google has guardrails in place to clearly identify images edited with Google Pics, and to prevent misuse. This app is available to "Trusted Testers" today, and will be rolling out globally to Google AI Pro and Ultra subscribers (and in preview to Google Workspace business users) this summer.

Enhanced conversational capabilities in Gmail, Docs, and Keep

Google is adding new voice features to Gmail, Google Docs, and Google Keep. The email service is getting a feature called Gmail Live, which will search your inbox to answer your questions. For instance, you could ask something like, "What's going on at my kid's school this week?" The service will then search your inbox and find all the emails from your child's school. The idea is to ask questions with natural language, which is what makes it "conversational."

The next new feature is Docs Live, which acts as your partner and co-writer, according to Google. You can talk to Docs Live, and it will help you brainstorm ideas, organize your thoughts, and structure your document. Google says the feature can even pull information from Gmail, Google Drive, Google Chat, and the web with your permission. You can think of Docs Live as an outlining tool that can gather a bunch of information quickly, making the writing process quicker for those who struggle to get started with a blank document.

For Google Keep users, the new conversational features can turn your "brain dump" into organized notes and lists. You can say everything you want noted down, and the AI assistant will do the sorting and organizing for you. Like Google Pics, all of these conversational features will be rolling out to Google AI Pro and Ultra subscribers (and in preview to Google Workspace business users) this summer.

New features in Gmail's AI Inbox

Gmail's AI Inbox, which offers AI-powered summaries, suggestions, and proofreading, is getting a few new tricks too. During I/O 2026, Google announced that AI Inbox will get personalized draft replies, instant file access, and streamlined task management. AI Inbox will be able to generate a contextual draft for you when it detects that an email requires an urgent response, the company claims. With instant file access, AI Inbox can find and display a link to a document from Google Docs, Sheets, or Slides when you need it, which could make it easier to find the right document in email threads. The feature also lets you mark all emails in a certain topic as read. These features will be available to Google AI Plus and Pro subscribers in the U.S. starting today.

A new personal AI agent: Gemini Spark

The final Workspace-related update today is a new AI agent called Gemini Spark. The company says Spark transforms Gemini from an assistant that answers your questions to one that gets things done with your permission. Once you enable this feature, Google says Gemini Spark can perform tasks such as sending emails on your behalf, and adding events to your calendar. The company adds that it will always ask your permission before performing these actions. Gemini Spark will be rolling out to Google Workspace business users in the coming week, and it'll be available through the Gemini app.

Google announces agent-optimized Gemini 3.5.Flash and a do-anything model called Omni

2026-05-20T00:13:28.597+05:30

At last year's I/O event, Google was still talking about the 2.5 branch of Gemini, and what a difference a year makes. We've gone through the 3.0 and 3.1 families since then, and now it's on to version 3.5. Gemini 3.5 Flash is rolling out across a wide range of Google products starting today, and Google again claims this model is even better than its last-gen Pro model.

That has been a trend with Google's tick-tock model updates over the past year, but the team says this release is special. Gemini 3.5 Flash allegedly offers frontier-level intelligence while also being efficient enough that it may finally make complex agentic tasks worth doing at scale. Tulsee Doshi, senior director of product management for Gemini, explains that the innovations of Gemini 3.5 Flash are woven through multiple Google products, and this is just the start.

Credit: Google

Read full article

Comments

source https://arstechnica.com/google/2026/05/google-announces-agent-optimized-gemini-3-5-flash-and-a-do-anything-model-called-omni/

Diablo 4 has fixed its controversial shop problem by giving me a violent knife-wielding crab

2026-05-19T00:57:12.605+05:30

Paid skins prove you have cash, but Diablo 4’s hidden grinds prove dedication beats a wallet every time.

source https://www.windowscentral.com/gaming/diablo-4-has-fixed-its-controversial-shop-problem-by-giving-me-a-violent-knife-wielding-crab

Yes, you can serve a website from a $1 microcontroller

2026-05-19T00:33:50.503+05:30

Web hosting bills getting too expensive? Maybe you ought to consider serving your site from a one-dollar 8-bit microcontroller. Okay, you won’t exactly be serving up a high-performance, graphic-rich website using this project from a European developer and blogger who goes by Maurycyz. The setup is limited to one URL, but hey, it actually works, provided an influx of visitors hasn’t killed the site yet. The bargain-basement chip that serves as the central component of this project is the AVR64DD32, which currently retails from DigiKey for $1.30. It has a single 8-bit AVR core with a blistering 24 MHz max clock speed, 8 KB of static RAM, 64 KB of flash memory, and 256 bytes of EEPROM non-volatile memory for storing a very limited amount of data. Serving a web page from such a limited chip is a task, to say the least, and Maurycyz had to do a lot of legwork to get the thing working. The I/O pins on the AVR max out at 12 MHz, which Maurycyz explained meant that it wouldn’t be possible to use Ethernet for the project, as the data flow from even the aged baseline Ethernet connection of 10BASE-T is too fast for the chip to handle. “10BASE-T still runs at 10 megabits/second,” Maurycyz wrote. “Worse, it uses Manchester encoding: a zero is sent as ‘10’ and a one as ‘01,’ so 10 megabits of data is actually 20 megabits at the wire.” “The proper solution is to buy a dedicated Ethernet chip from DigiKey, but then I'd be waiting weeks to finish this project,” Maurycyz noted. Instead of waiting, he decided to take a different approach by turning to Serial Line Internet Protocol (SLIP), just like the guy who turned a discarded vape into a web server last year. For those unfamiliar with SLIP, it’s a 38-year-old protocol designed to encapsulate IP traffic for transmission over serial lines, and it was widely used to make internet connections in the olden days. SLIP is still supported in modern Linux builds due to its compact size and the fact that it’s often used to connect microcontrollers to the internet. Now, giving the AVR an internet connection didn’t solve the harder problem of actually serving a web page to visitors. Maurycyz said the chip could generate response packets by swapping the source and destination addresses on incoming traffic and resetting the packet’s TTL value, but implementing TCP still took several days of work. HTTP handling was simplified by returning a hardcoded response for every request, which works as long as the site only serves a single URL. Here’s that limitation we were talking about: “This works fine as long as there's only a single URL on the site,” Maurycyz said. Sorry for those wanting to host more pages from that $1 microcontroller. Lastly, Maurycyz said he had to figure out how to get requests from the internet to the microcontroller without spending money on a publicly routed IP address. That was resolved by using WireGuard to connect the microcontroller located at his home to a public-facing machine at a Helsinki datacenter, which then proxied requests to the microcontroller using a local address block. “This means that visitors aren't directly connecting to the MCU's TCP/IP stack... but hey, it's the same setup that the Vape Server uses and no one complained,” Maurycyz said. And all without having to buy a vape or root through dumpsters to find an old one. Source code and a pre-built binary are available from Maurycyz’s site for those who want to try this experiment for themselves. We reached out to Maurycyz for more info on the project, but didn’t hear back. ®

source https://www.theregister.com/offbeat/2026/05/18/yes-you-can-serve-a-website-from-a-1-microcontroller/5242096

5 Steps to Managing Shadow AI Tools Without Slowing Down Employees

2026-05-19T00:24:25.988+05:30

Many employees already use shadow AI tools at work without security review. Adaptive Security breaks down how teams can build practical AI governance without adding friction for employees. [...]

source https://www.bleepingcomputer.com/news/security/5-steps-to-managing-shadow-ai-tools-without-slowing-down-employees/

You Can Get a Free Band or a $35 Credit When You Preorder the New Fitbit Air

2026-05-19T00:15:19.105+05:30

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

The new Google Fitbit Air is almost available to the public, and our Senior Health Editor, Beth Skwarecki, is already impressed with it. It costs $99.99, but if you preorder by May 25, you can choose either a $35 Google Store credit after your order ships if you get it through Google;, or a free Active Band if you preorder through Amazon. The watch also comes with three months of Google Health Premium regardless of where you preorder from, which unlocks a wide range of Gemini-powered AI coaching features. Compared to Whoop, this is an especially appealing offer since Whoop’s premium plans start at $199 a year, depending on the region.

Comes with $35 Google Store credit

Google Fitbit Air

$99.99 at Google

With free Active Band

Google Fitbit Air - Lavender with Large Fog Active Band Bundle

at Amazon

The Google Fitbit Air is essentially a Whoop alternative with its minimalist, screenless design, built to be worn discreetly while quietly tracking metrics like health, fitness, sleep, and recovery. The screenless design is also a perk if you’re not a fan of bulky wearables and want something you can wear during workouts, at work, or while sleeping without it being a distraction.

That said, it won’t automatically reduce screentime. Although having no display is the point, it means you’ll need your phone to check stats and progress. It doesn’t come with true smartwatch features like accessing calls, texts, apps, or maps on a watch face. Additionally, many of the best features rely on a paid subscription to Google Health Premium. Still, the three free months are a nice bonus, and with optional plans starting at $99 per year, it’s half the cost of a yearly mandatory Whoop subscription. (It’s worth noting that the Google credit comes after your order ships, so you’re still required to pay the $99.99 upfront.)

Still, this is a great choice for people who want screen-free tracking like that offered by Whoop but at a more budget-friendly price point (both upfront and over the long term with the subscription), and for those who want to track their daily activity and get insights without wearing a full smartwatch. However, if you want features like a built-in screen, music controls, and notifications available on your wrist, you’re better off with a more traditional smartwatch.

Our Best Editor-Vetted Tech Deals Right Now

Apple AirPods Pro 3 Noise Cancelling Heart Rate Wireless Earbuds — $229.00 (List Price $249.00)

Apple Watch Series 11 [GPS 46mm] Smartwatch with Jet Black Aluminum Case with Black Sport Band - M/L. Sleep Score, Fitness Tracker, Health Monitoring, Always-On Display, Water Resistant — $329.00 (List Price $429.00)

Apple iPad 11" A16 128GB Wi-Fi Tablet (Silver, 2025) — $299.00 (List Price $349.00)

Fire TV Stick 4K Plus Streaming Player With Remote (2025 Model) — $29.99 (List Price $49.99)

Deals are selected by our commerce team

Samsung’s weather app sparks storm of controversy by handing territory to North Korea

2026-05-18T08:33:38.439+05:30

Samsung found itself facing down controversy in South Korea last week, when the weather app pre-installed on many of its devices incorrectly labelled an island territory named Dokdo as part of North Korea. Dokdo is a group of volcanic islets that is the subject of a territorial dispute between South Korea, North Korea, and Japan. Netizens were therefore outraged by a champion of South Korean industry handing the islands to foes in North Korea. Mislabelling the map was therefore sufficiently controversial that Samsung quickly pushed an update to fix the error – and blamed data from The Weather Channel as the source of the mistake. While we’re talking about islands … The Federated States of Micronesia, the Republic of Kiribati, and the Republic of Nauru last week connected to the world over a submarine cable for the first time. The three Pacific island nations hooked up to the East Micronesia Cable System, which NEC Corporation built and last week handed over to telecoms companies in the three nations. The cable can carry 100Gbps to each country where it lands, and has capacity to reach 10 Tbps. The three nations are collectively home to around 100,000 people. The governments of Australia, Japan, and the USA funded construction of the cable as part of ongoing diplomatic efforts to woo Pacific nations at a time China is also active in the region. Bitdefender spots alleged Chinese attack on Azerbaijan Antivirus vendor Bitdefender last week published what it says is evidence of a China-backed “multi-wave intrusion targeting an Azerbaijani oil and gas company from late December 2025 through late February 2026.” Bitdefender linked the attacks to the resurgent FamousSparrow crew, which apparently deployed “an evolved DLL sideloading technique” to drop the Deed RAT and Terndoor backdoors. The company’s researchers think attackers targeted a vulnerable Microsoft Exchange server. Senior security researcher Victor Vrabie suggested the attack is a sign that Russia and China are both trying to gain a foothold in Azerbaijan’s energy infrastructure, to gain leverage over energy supplies to Europe. The central Asian nation is a major oil and gas producer, and exports much of its output through pipelines that reach Turkey and Georgia. The importance of those routes has grown thanks to slowing gas exports from the Middle East. China seemingly shuns Nvidia to focus on its own alternatives The United States has allowed several Chinese companies to acquire Nvidia’s H200 accelerators, but Beijing won’t let local buyers do the deed. That’s the washup of President Trump’s visit to China last week, during which US authorities reportedly issued licenses allowing Nvidia to sell its wares to Chinese tech companies including Alibaba, Tencent, ByteDance and JD.com. But President Trump later remarked that China has not allowed its tech companies to buy H200s “because they chose not to, they want to develop their own.” That’s perhaps the strongest signal yet that China has decided to decouple from foreign tech stacks. Bottom drops out of India’s smartphone market Analyst firm IDC last week reported that smartphone shipments in India slumped by 4.1 percent in the first quarter of 2026. IDC said that subdued result would have been worse had brands not decided to front-load channel inventory before the cost of smartphone rise due to the soaring cost of memory. The firm said the result “signals a structural turning point for one of the world’s largest smartphone markets” because device-makers who sell entry-level devices “face shrinking margins and reduced market viability as memory costs continue to rise.” When consumers who buy sub-US$100 phones do upgrade, they “are being pushed upmarket by necessity rather than aspiration” – meaning demand is muted and will likely remain so for quite some time. ®

source https://www.theregister.com/personal-tech/2026/05/18/samsungs-weather-app-sparks-storm-of-controversy-by-handing-territory-to-north-korea/5241656

Linus Torvalds says AI-powered bug hunters have made Linux security mailing list ‘almost entirely unmanageable’

2026-05-18T06:33:30.489+05:30

Linux kernel boss Linus Torvalds has declared the project’s security mailing list has become “almost entirely unmanageable” due to multiple researchers using AI to find bugs and then filling the list with duplicate reports. Torvalds used his weekly state of the kernel post to deliver release candidate four for Linux 7.1 and report “fairly normal” progress towards a full release. He then pointed kernelistas to the project’s documentation, which he wrote “might be worth highlighting” as “the continued flood of AI reports has basically made the security list almost entirely unmanageable, with enormous duplication due to different people finding the same things with the same tools.” “People spend all their time just forwarding things to the right people or saying ‘that was already fixed a week/month ago’ and pointing to the public discussion,” Torvalds complained. The Penguin Emperor believes that kind of chatter is “all entirely pointless churn” and isn’t productive because “AI detected bugs are pretty much by definition not secret, and treating them on some private list is a waste of time for everybody involved – and only makes that duplication worse because the reporters can't even see each other's reports.” He then offered an opinion on how best to use AI to improve software security. “AI tools are great, but only if they actually help, rather than cause unnecessary pain and pointless make-believe work,” he wrote. “Feel free to use them, but use them in a way that is productive and makes for a better experience.” “The documentation may be a bit less blunt than I am,” he added, “but that's the core gist of it.” “So just to make it really clear: If you found a bug using AI tools, the chances are somebody else found it too. If you actually want to add value, read the documentation, create a patch too, and add some real value on *top* of what the AI did. Don't be the drive-by ‘send a random report with no real understanding’ kind of person. OK?” Torvalds' remarks contrast with recent comments from fellow kernel maintainer Greg Kroah-Hartman, who recently told The Register that AI has become an increasingly useful tool for the FOSS community. ®

source https://www.theregister.com/security/2026/05/18/linus-torvalds-says-ai-powered-bug-hunters-have-made-linux-security-mailing-list-almost-entirely-unmanageable/5241633

Surprise AI bills leave AWS and Google Cloud users aghast

2026-05-18T04:33:48.991+05:30

KETTLE Hopefully you haven't had reason to notice yet, but there's a rising problem with AI services on Google Cloud, AWS, and other platforms sticking their customers with bills in the tens of thousands of dollars. This week's episode of the Kettle focuses on two such stories that The Register published this week, one concerning Google and another involving AWS. In both cases, cloud customers using AI incurred massive bills without any prior notification from their provider and not a lot of help to resolve the matter with any sense of urgency. Tune in to this week's episode to hear host Brandon Vigliarolo chat with O'Ryan Johnson and Richard Speed about their stories, what's causing these massive bills, and how you can avoid a similar situation at your own organization. You can listen to The Kettle here, as well as on Spotify and Apple Music, or read the transcript of the latest episode below. It's been lightly edited for clarity. Brandon (00:01) Hello everyone and welcome back to another episode of The Register's Kettle podcast. I'm Reg reporter Brandon Vigliarolo, and this week I'm joined by my colleagues Richard Speed and Kettle newcomer O'Ryan Johnson to talk about a recent spike in cloud AI API abuse that's sticking customers with some massive charges. We're talking tens of thousands of dollars that Google is seeming to...try hard not to refund. Guys, thanks for coming on. O'Ryan Johnson (00:29) Great to be here. Brandon (00:30) And O' Ryan, welcome again to your first Kettle episode. Glad to have you here. So in this case, this one is primarily based on an exclusive you published this week about compromised Google Cloud API keys. And from what I read, it seems like cyber criminals are using those keys to run all the AI inference they want on most expensive models that Google has without paying a dime. So walk me through what exactly this story's about. O'Ryan Johnson (00:33) So there were a couple parts of this. One is the API abuse. But then there was this policy by Google that kind of threw gasoline on the fire. So if you're a developer and you've created an API key for your projects, if your project uses Maps, you'll create an API key. And for years, the advice from Google was put that API key on the front end of that, make it public so that when users are using your site, it links back to your project. The problem was a couple years ago, they allowed those API keys, if they were configured correctly, to also access Gemini. And a lot of folks who were early adopters of AI went in and said, okay, I want to use Gemini with my project. And not really connecting the dots that their API key on the front end that was publicly available would now also allow anybody to inference Google's Gemini platform. And it wasn't a big deal, I think, for a lot of years because I don't think the platform was really that amazing. Brandon (02:01) Yeah, because you said this is a three year old change, right? O'Ryan Johnson (02:22) But recently...Nano Banana and the Veo 3 models came out. And that's when I think we started to see a lot of this. This great security company named Truffle wrote something about this in February saying, look, be careful because if you've put your API key out according to Google's instructions, and if you've also been working with Gemini models, there's a chance that you may have inadvertently opened up your API key to anybody to be able to inference [Veo] and NanoBanana to their heart's content. Brandon (02:40) And specifically a Maps API key, right? Okay, O'Ryan Johnson (02:51) Correct. Which again was, was Google had told everybody for quite a while was safe. And so, what happened kind of inevitably is folks were bad actors were in fact using that for for those purposes So you'd have these you know sort of like horror stories of waking up in the morning and seeing your Google account Which you maybe you never spent more than fifty dollars a month, all of a sudden you have a $3,000 bill, $5,000 bill. I talked to a guy who got a notification from his credit card company that "Hey, we're basically we're shutting off your account because you spent too much," and he's like "What the hell is going on?" And as he's in there trying to figure it out he sees the bill keeps going up.... Brandon (03:26) ...I think you mentioned basically that where this is, how you figure this out, is kind of buried, right? It's hard to find, right? So as he's looking, trying to frantically figure out what's happening, more charges are being added. I couldn't imagine waking up in the morning to that kind of scenario. O'Ryan Johnson (03:36) It's a rough, rough way to start the day. It's really tough. Brandon (03:50) So that's the first part, right? So what's the second part then? O'Ryan Johnson (03:52) So, right, that's the first part. The second part is that, you know, this happened to people who had spending caps in place. And Google has only recently put spending caps in place, but they're really loose caps. I talked to a developer in Australia who said, "Look, I put a $250 spending cap in place. And when I woke up, I had a $10,000 bill." ...And he said, "When I was going to going through afterwards, I looked and I said my spending tier was at the $100,000 limit. And I said, how does it happen?" Well, if you look like Google was actually very upfront about this. In March, they put out a blog and said, "Hey, we're going to help you out. If you've only got a $250 spending cap, if you spent $1,000 in the lifetime of your account and you've been a Google member for 30 days or more, a Google Cloud developer for 30 days or more, you can spend $100,000." Brandon (04:47) And there's no notification to the user accounts that this is being done? O'Ryan Johnson (04:50) Except for the emails that say this is how much you owe us, which is all after the fact. Brandon (04:57) And if you're less than 30 days, right, it's moving to tier two is, I think, what's the cap on that? O'Ryan Johnson (05:01) Two thousand dollars. Brandon (05:04) But even then, it's spend a hundred bucks in the lifetime of your account? O'Ryan Johnson (05:06) A hundred dollars and be three days old, and Google will give you a $2,000 cap to spend. Those are the most generous terms – you guys have been around IT for years, what distributor would ever give you terms like that like if you went to TD Synnex or if you went to Ingram Micro and said, "Hey, I'm 30 days old and I've spent $1,000. I would like $100,000 in credit with you." They would laugh . Brandon (05:13) They would laugh you out of the office and then maybe close your account. O'Ryan Johnson (05:37) Yeah, even the best distributor in the world is not going to give you those terms, but Google's opened that up. And then of course the problem is trying to get that money ... trying to get your account restored. like in two of the cases, the money had already been spent. So the credit cards, one was $17,000, one was $10,000. The money was already out of their account and they have this project. If they charge it back, they're afraid that Google's going to shut down their project and delete it. If they stick with the bill, then they're stuck with this debt that is obviously outside the bounds of any budget that they had set for their Google Cloud project. Brandon (06:12) Yeah, for a small developer that can be devastating. O'Ryan Johnson (06:15) Right. Right. So Google, though, we do have an update coming today. Google has refunded the two people we talked about and looked in their account. It looks like they're kind of going after this with more accounts too, based on what I've talked to with Google, they're going to look at a lot more of these issues...This didn't come to me in a vacuum. I mean, this this was on these posts have been kind of flooding Reddit. If you go to the Google Cloud subreddit there, you you pretty much don't go, there's two or three a day that are popping up saying, "Hey, my gosh, I've got $10,000. I got $7,000 in bills. Like I only ever spent, you know, $50 with these folks. How am I getting these bills?" Brandon (06:57) Right, so it's kind of a two-part story here. The automatic tier upgrades are obviously a problem, but are all these cases that you're seeing, are they tied back to the Truffle notice? I mean, these are all Maps API keys? O'Ryan Johnson (07:02) Not all of them. Some people say like, "Look, I never put my API key out publicly." And I talked to a guy yesterday who said, "Look, my API key has been hidden from everybody. I think I got brute forced." ....I don't possibility or the probability of being able to brute force an API key, they're huge, long chains of numbers and texts. Probably not impossible...But this guy, his bill was $127,000, which is just a huge, huge amount. Brandon (07:40) God that is so that is so much money that's ridiculous. Ten thousand dollars is bad enough add another zero to that and oh my God. O'Ryan Johnson (07:51) That's rough. Fortunately, he caught it before...That bill only exists with Google. Fortunately, the good side is, it's not in his credit card. So he doesn't have to try to pay that back. The bad news is, his Google project is looking at a possible deletion if he can't convince Google that, "Look, this wasn't me, this was really somebody else who brute-forced my API." Brandon (08:15) I'm guessing proving that is pretty difficult. O'Ryan Johnson (08:17) Well it's difficult, what makes it difficult is he no longer has access to the logs because he hasn't paid the account, so now he has to rely on somebody at Google to go through those logs and make his case for him. Brandon (08:34) When there's $100,000 on the line. O'Ryan Johnson (08:36) When there's 127 on the line. That's a gamble. That's a gamble. Brandon (08:43) So this is bad enough, but as I understand, Richard, Google's not the only company being a bit shifty with their AI billing. You wrote a story this week about an AWS customer who was billed $30,000 despite supposedly having a setting enabled to prevent this. So what's this all about? Richard Speed (08:50) It's kind of almost a cautionary tale in some ways. Again, we've talked about Google, there's also, this is AWS. And this is a user who was using AWS Bedrock. He wanted to take Claude Opus out for a spin, try it out. He had some startup credits fired by Activate. All great. Now he was using a tool called the AWS Cost Anomaly Detection Tool. What that does, that actually sends you alerts if you're doing some odd things and your account is incurring additional costs, and as well as using AI machine learning, you can also set some custom thresholds... "If I spend more than this then stop or shout at me or whatever Brandon (09:39) Yeah, cut me off. Yeah. Richard Speed (09:45) So he thought, "Great, I've got that, what could possibly go wrong?" And so he began to use his AWS Bedrock and no alerts were fired, all was good until about a month after he began using it he got a bill for $30,000 or $38,000 through where he was expecting hundreds. And the reason being was that AWS Bedrock apparently bills through AWS Marketplace, and that is not compatible with the cost anomaly detection. Brandon (10:06) So Marketplace is where you can pick up third party integrations for AWS, right? Richard Speed (10:17) Right, and that's where AWS Bedrock was being billed, was basically invoiced through. And to be completely fair to AWS, that is documented. It is in the documentation, "This will happen." So, hence the cautionary tale aspect. But again, I've had a few people say, actually it's pretty unintuitive, this. You kind of would assume it's being caught and it wasn't caught. And so this is gone through. Now, unfortunately, at the moment, I don't think there is the happy ending about a refund. If and when I get more information, I willupdate. But the cautionary tale aspect is, I've heard from somebody else who said, yeah, similar sorts of things can happen. So I tend to go through directly through the AI provider. In this case, it's Anthropic. And there again, you can put limits in place. And those limits did save this particular person from a $50,000 mistake. And he only ended up paying $50 because he'd accidentally turned on a thing which enabled a lot more invoicing to happen, and of course it was stopped before it got out control. Brandon (11:25) I'm assuming a lot of customers, with the way they have their architectures and their infrastructure set up and their various providers, I mean, is it going to be simple for a lot of businesses to say, I'm going to skip AWS and go straight to the AI company itself? I mean, that seems like it might work in some cases, right? But a lot of people are going to be trying to integrate these. And so they're going to have to go through things. So does Cost Anomaly Detection function only with first-party Amazon products then basically? Anything in the Marketplace that you're pulling from a third-party provider doesn't get included in this? Richard Speed (11:59) Yeah, I believe so. Yeah, it's just through AWS services except for Marketplace stuff. But there are other checks and things in place in AWS. It's just in this instance, the expectation was if I'm using Cost Anomaly Detection, it should stop me running up a massive invoice or running up a massive bill using AWS Bedrock. In this case, it didn't. It was completely silent as the thousands and thousands and thousands began to rack up on the account. Brandon (12:05) And even, I think you wrote, even when his credits ran out. Like, he ran out of credits and switched to cash billing and there was no notice. Richard Speed (12:29) Exactly. It suddenly went from from credits to cash billing again with no notification or warning. It just happened. And so again, his account began to incur these charges. And so he didn't realize until the invoice came through. "Oh, my goodness me. How terrifying is this?" As as Ryan said, it's quite a shock when when you're used to a small amount per month and then suddenly a massive invoice comes through. O'Ryan Johnson (12:53) One thing that is kind of universal across this that one of these users pointed out, is that the most frustrating part is that they have the information. They can see what you're doing in your account and they don't stop it. All this information that we're talking about, whether it's your usage, whether it's your billing, all that stuff is within the four walls of, whether it's Google or AWS and they, whether it's intentionally or unintentionally – we live in this era where everybody talks about immaculate orchestration across all their environments, right? Like, I mean, if you're in SaaS, that's all you hear is about how amazing and perfect their SaaS products are. And we just don't see that in practice. You don't see that orchestration, and you certainly don't see it if it can ever give the user an advantage, or if it can ever give the user the ability to control how much they spend. Like if a user could shut off – if there was a notification that came in and said, "Hey, did you know that you're on Veo right now and you're generating videos? Would you like to shut that off?" Think about your credit card company. If I go one county over and I spend $10 at a Target, I'll get an alert from my card company. "Hey, are you sure?" Are you telling me, Google and AWS, that you can't do that? Like, don't give me that. I mean, this reminds me like when the banks in the US had overdraft fees, they used to – they could see how much money you had in your account. They would gladly let you spend much more than that so that they could fine you for every transaction. And so it was very similar. You'd open up your bank account and see like, I'm $800 in debt. So that was eventually determined to be, hey, that's an aggressive, that's not a good policy. We shouldn't allow people to do that. And I just wonder if, I wonder if there's gonna be some sort of trade regulation that kicks in on this. Brandon (14:26) I mean, it almost feels like there has to be. What we have in these two stories this week is multiple cloud platforms making their AI billing usage or usage billing so convoluted that a non-trivial number of customers are seeing their bill skyrocket, whether both due to cybercrime or simply the fact that Cost Anomaly Detection on AWS isn't very well-defined on the Marketplace, right? You're seeing multiple companies this is happening to, right? Again, O'Ryan, you kind of went right to the, the conspiracy theory, but that's where my mind goes too, this seems really convenient. Google's move in March. All these kinds of things are very well timed to ensure that companies that are adopting AI are being left with this ambiguous billing situation. Richard Speed (15:35) I mean, if only there was a tool that could spot strange patterns in data and frames. I mean, what would that look like? [Laughter.] Brandon (15:43) Yeah, I don't know. ⁓ There's no way, there's no way that ⁓ Google and AWS don't see this usage or can't monitor it. Can't pop a large language model on there to keep an eye out for ⁓ unusual billing and notify people. Like you said, if you never use [Veo] or never use NanoBanana and all of a sudden your account's racking up thousands of dollars of charges on it, Google should probably say, "Hey, is this you?" Right? Like, you know, that would be, I would hope that would happen. Right? You know, it's like you said, right? Your bank, Target will know, or your credit card company will notify if you spend things a county over. Right? If I try to log into a video game online from a different IP address, it locks me out and makes me me approve it. Right? Like this is not a complicated technology here. O'Ryan Johnson (16:32) No, think about the user agreements that we have like with all of our subscriptions like you know like Netflix. If my kid tries to log into my Netflix from where they live, they can't, and I get these notifications from Netflix, "Hey do you want to add somebody on your account?" Like don't tell me that you can't do that, Google. And Google actually says that they hat between the usage and the spend, they're better than AWS when it comes to being able to spot this. But it's like, it's still something like 28 days to be able to reconcile usage with spend. And that just does not make any amount of sense. Brandon (17:16) It takes Google 28 days? O'Ryan Johnson (17:18) They're pushing people into these products. They're pushing, they want you to use these products. They want developers to, they want to be able to say, we have X number of developers who are using this. We have X number of spend. All of those hijacked API keys are inevitably helping marketing for Gemini. Just through sheer usage numbers, through sheer revenue and dollar spend, that drives a narrative that they can then, you on the quarterly earnings call say, "Hey, look at all these people using our product. Look at all the spend on [Veo]. Look at all the spend on Banana." come on, you guys, you got to make it fair for the rest of us, man. Brandon (17:59) I'm just gonna toss it allegedly in there before Google comes after us, right? You know. We don't know for sure that this is what they're planning, but it sure seems, the ducks do line up, right? So guys, are you familiar? Do you know, are any other cloud platforms...are there similar issues on Azure, on other platforms? Have you heard anything? Or does this seem to be mainly confined right now to Google and AWS? Richard Speed (18:11) There have been some issues on Azure. I read a piece, oh crikey, several weeks, maybe even months ago now, regarding a similar thing to what's happened with AWS with a user who had, he hadn't realized that his startup credits didn't count towards AI usage. And then he found himself hit with a massive invoice because again, Microsoft just quietly said, "Yeah, sure. You want that service? No problem. Here you go. Use it." And so he used it and then the huge invoice came through. I think... I think it's important to point out that these companies, they're not doing anything wrong legally. Ethically, I'm with O'Ryan, they should be warning you to say, "Hey, you know, you're spending way more now than you ever used to before. These services that you've never used before, are you sure you want to be doing that? Are you sure about that?" Brandon (18:51) I was talking to my wife about Google before we started the podcast, right? Because when we were talking about the topic for this week, and I think Matt, our editor in chief said, "AI overage charges." I was like, "What? This is going to be a boring episode." And then I got to actually reading these stories and I'm like, "Oh my God, this is really interesting." My wife's like, "Surely this is illegal." I'm like, "I don't know, if it's in the terms of service, right? You know? Yeah." O'Ryan Johnson (19:23) It's like the South Park episode. Richard Speed (xx:xx) I think another aspect of this is there's a perception that AI services are inexpensive and you won't run up these massive costs. One thing I've come across a few times are companies saying, "Hey, we can increase the productivity of our staff enormously because we can roll out these AI tools and our employees can use them and they'll be massively more productive and it'll be great." They're forgetting that of course there is a cost to that. And I think what we're seeing here are people hitting these costs. So I think that the message has got to be, you need to be – I mean, until these companies actually put in warnings to say, know, perhaps make it very clear how much this stuff is really costing, I think you need to be aware that this isn't a free service, you know, it's going to be paid for somehow. Brandon (xx:xx) I guess that's kind of the big warning to businesses, right? Or AI users, anyone who's using AI in the cloud in general, It's like these things are not free. Yeah, sure, you can use ChatGPT for free if you're, you know, some random person logging into the website. But if you want to go enterprise with this or use it in any kind of business capacity, it's going to cost you money and potentially a lot of it. So Richard, you said that it looks like the AWS user might be a little bit hosed on getting a refund. Do you know is Amazon – did you talk to Amazon for the story? Do they have any intention to change the marketplace versus non-marketplace CAD policy? Richard Speed (xx:xx) They did respond, and at the moment there's no plans to change it. O'Ryan Johnson (xx:xx) Google is also, they're sticking by their automatic tier upgrades. They like the flexibility that it gives to developers. Flexibility, of course, meaning that developers can spend a lot more than they initially wanted to, or agreed to. Brandon (xx:xx) It's a very one-sided flexibility, really, when you think about it. O'Ryan Johnson (xx:xx) In fairness, we are kind of helping at least notify people that this could happen. This is something that is really happening to people and their bills really do become five-figure, in some cases six-figure bills at the end of the month through no intention of their own. Brandon (xx:xx) Yeah, so I guess basically the big, yeah, like we said, the big takeaway for business AI customers is to just really watch that billing, be sure that whatever system you have in place to prevent overages is actually doing its job, and hide those API keys. Well, like we said, guess this is just a cautionary tale, you know, to watch that billing. So if this keeps happening, we are definitely going to be talking about it and writing about it again. And we hope that you will tune in on a future episode of The Kettle to find out more. ®

source https://www.theregister.com/ai-ml/2026/05/18/surprise-ai-bills-leave-aws-and-google-cloud-users-aghast/5241348

New Windows 'MiniPlasma' zero-day exploit gives SYSTEM access, PoC released

2026-05-18T04:24:12.638+05:30

A cybersecurity researcher has released a proof-of-concept exploit for a Windows privilege escalation zero-day dubbed "MiniPlasma" that lets attackers gain SYSTEM privileges on fully patched Windows systems. [...]

source https://www.bleepingcomputer.com/news/microsoft/new-windows-miniplasma-zero-day-exploit-gives-system-access-poc-released/

Wanted: Digital chief for England's schools. Must enjoy data, AI, and concrete problems

2026-05-17T14:33:29.935+05:30

England's Department for Education is advertising a role paying up to £200,000 a year to lead a new digital and infrastructure group overseeing school buildings and maintenance, as well as technology and data. Its Director General, Digital and Infrastructure, will lead the technology function of around 1,800 staff, develop a new strategy covering digital services, data, and artificial intelligence, and lead work on a unique identifier for children and other learners in England. Scotland, Wales, and Northern Ireland run education services on a devolved basis. The successful candidate will also implement a new strategy for "the education estate" of schools, colleges, nurseries, and children's homes. The job ad warns the function "carries some of the highest levels of risk and accountability in the department - including life-and-death decisions on safety," citing ongoing work to remove unsafe reinforced autoclaved aerated concrete (RAAC) from schools. "I am looking for a leader who is motivated by impact - someone who is able to combine their digital and data expertise with their drive to improve outcomes for children and young people," writes the department’s permanent secretary, Susan Acland-Hood, in a briefing document with the advert. "Whilst you do not need to be an expert on education policy, you need to be curious and committed to rapidly building your understanding of the latest evidence, system, and policy landscape." The department is willing to base the job in Bristol, Cambridge, Coventry, Darlington, London, Manchester, Nottingham, or Sheffield, although those who do not work in the capital will need to go there frequently. Applications close on June 1. Several other departments have recently advertised digital director-general posts, the civil service job category just below permanent secretary (equivalent to chief executive). In January, England's Department of Health and Social Care advertised the role of director general for technology, digital and data with a salary of up to £285,000 a year. In February, the Ministry of Defence offered £270,000 to £300,000 for its chief digital and information officer job. And in April, the Department for Science, Innovation and Technology advertised for three directors-general, one paid £174,000 and the other two paying between £200,000 and £260,000 annually. ®

source https://www.theregister.com/public-sector/2026/05/17/wanted-digital-chief-for-englands-schools-must-enjoy-data-ai-and-concrete-problems/5240242

Microsoft rejects critical Azure vulnerability report, no CVE issued

2026-05-17T03:24:41.898+05:30

A security researcher claims Microsoft quietly fixed an Azure Backup for AKS vulnerability after rejecting his report, and without issuing a CVE. Microsoft disputes the claim, telling BleepingComputer the behavior was expected and that "no product changes were made," despite the researcher documenting a silent fix. [...]

source https://www.bleepingcomputer.com/news/security/microsoft-rejects-critical-azure-vulnerability-report-no-cve-issued/

I reckon Asha Sharma wants to give Xbox its exclusive games back — but these PlayStation comments reveal why Microsoft probably won't let her

2026-05-17T02:35:19.045+05:30

It's the year of GTA6. The game has an exclusive PlayStation marketing deal. Call of Duty is also not launching on PS4 this year, giving holdouts a reason to upgrade ... so why is Sony warning investors of decreased sales? And why is this relevant to Xbox exclusive games? Let's talk.