SoylentNews

An Anonymous Coward writes:

This release also introduces new System Information and System Administration tools, as well as updates to existing tools:

Delayed a bit due to the Christmas and New Year's holidays, but Linux Mint 22.3 "Zena" is now available for download as the Linux Mint team has just started publishing the final ISO images on the official mirrors.

Based on the Ubuntu 24.04 LTS (Noble Numbat) operating system series and powered by Linux kernel 6.14, which should boost hardware support, Linux Mint 22.3 "Zena" ships with the usual editions featuring the Cinnamon 6.6, Xfce 4.18, and MATE 1.26 desktop environments.

As you can see, the biggest attraction of the Linux Mint 22.3 release is the latest Cinnamon 6.6 desktop, which introduces a redesigned application menu applet, modernized keyboard handling, virtual keyboard improvements, improved theme support, updated applets, and much more.

Linux Mint 22.3 also introduces new System Information and System Administration tools, per-app panel notification indicators, an always-on Night Light feature, the ability to pause snapshots in the Timeshift backup tool and file operations in the Nemo file manager, and support for XApp Symbolic Icons (XSI).
"Symbolic icons are simple, monochrome icons used throughout applications for buttons, menus, and status indicators. They scale cleanly at different sizes and remain clear in both light and dark interface themes," said the Linux Mint developers.

Among other changes, Linux Mint 22.3 updates the Warpinator app with IPv6 support and the ability to send text messages, updates the Hypnotix IPTV player app with the ability to hide the mouse cursor in full-screen and forward keys towards MPV,

It also updates Captain, the suite of tools created by the Linux Mint for modern package management, to handle the installation of multiple packages via apt:// URLs, as well as the Update Manager utility to show a warning icon in the system tray if a restart is recommended after applying updates.

Last but not least, Linux Mint 22.3 adds a new "Include All" button to the mintbackup tool to make it easier for users to add all hidden files and folders before making a backup. Of course, this release also includes all the package updates and security patches from the upstream Ubuntu 24.04 LTS repositories.

Original Submission

hubie writes:

MIT physicists propose that under certain conditions, a magnetic material's electrons could splinter into fractions of themselves to form quasiparticles known as "anyons":

In the past year, two separate experiments in two different materials captured the same confounding scenario: the coexistence of superconductivity and magnetism. Scientists had assumed that these two quantum states are mutually exclusive; the presence of one should inherently destroy the other.

Now, theoretical physicists at MIT have an explanation for how this Jekyll-and-Hyde duality could emerge. In a paper appearing today in the Proceedings of the National Academy of Sciences, the team proposes that under certain conditions, a magnetic material's electrons could splinter into fractions of themselves to form quasiparticles known as "anyons." In certain fractions, the quasiparticles should flow together without friction, similar to how regular electrons can pair up to flow in conventional superconductors.

If the team's scenario is correct, it would introduce an entirely new form of superconductivity — one that persists in the presence of magnetism and involves a supercurrent of exotic anyons rather than everyday electrons.

"Many more experiments are needed before one can declare victory," says study lead author Senthil Todadri, the William and Emma Rogers Professor of Physics at MIT. "But this theory is very promising and shows that there can be new ways in which the phenomenon of superconductivity can arise."

What's more, if the idea of superconducting anyons can be confirmed and controlled in other materials, it could provide a new way to design stable qubits — atomic-scale "bits" that interact quantum mechanically to process information and carry out complex computations far more efficiently than conventional computer bits.

[...] Anyons are entirely different from the two main types of particles that make up the universe: bosons and fermions. Bosons are the extroverted particle type, as they prefer to be together and travel in packs. The photon is the classic example of a boson. In contrast, fermions prefer to keep to themselves, and repel each other if they are too near. Electrons, protons, and neutrons are examples of fermions. Together, bosons and fermions are the two major kingdoms of particles that make up matter in the three-dimensional universe.

canopic jug writes:

Bruce Schneier, the cryptographer and privacy specialist, and J B Branch, an accountability advocate at Public Citizen, have written a post about AI and the Corporate Capture of Knowledge. They raise hard questions about what happened to Aaron Swartz in the context of the what is going on with artificial intelligence, copyright, and ultimately the control of knowledge:

As AI becomes a larger part of America's economy, one can see the writing on the wall. Judges will twist themselves into knots to justify an innovative technology premised on literally stealing the works of artists, poets, musicians, all of academia and the internet, and vast expanses of literature. But if Swartz's actions were criminal, it is worth asking: What standard are we now applying to AI companies?

The question is not simply whether copyright law applies to AI. It is why the law appears to operate so differently depending on who is doing the extracting and for what purpose.

The stakes extend beyond copyright law or past injustices. They concern who controls the infrastructure of knowledge going forward and what that control means for democratic participation, accountability and public trust.

The questions they raise are important questions because the foundation of democracy is being able to make informed decisions through participation in civilized, well-rounded discussions. The prerequisite for that is knowledge.

Previously:
(2025) OpenAI Desperate to Avoid Explaining Why It Deleted Pirated Book Datasets
(2025) Meta Pirated and Seeded Porn for Years to Train AI, Lawsuit Says
(2025) Creating AI Based Entirely on Ethically-Sourced Data
(2025) Copyright Office Thinks AI Companies Sometimes Stole Content
(2024) OpenAI Whistleblower Found Dead in San Francisco Apartment
(2024) OpenAI Blamed NYT for Tech Problem Erasing Evidence of Copyright Abuse
(2024) AI Companies Are Finally Being Forced To Cough Up For Training Data
and more ...

Original Submission

Arthur T Knackerbracket writes:

EU-only ops, German subsidiaries, and a pinky promise your data won't end up in Uncle Sam's hands:

Amid continued trade and geopolitical volatility between Europe and the US, Amazon Web Services is making its European Sovereign Cloud generally available today and plans to expand so-called Local Zones.

Amazon says the cloud is "entirely located within the EU, and physically and logically separate from other AWS Regions." It will initially offer 90 services from compute to database, networking, security, storage, and AI.

It is "independently operated" by EU residents and "backed by strong technical controls, sovereign assurances, and legal protections designed to meet the needs of European governments and enterprises for sensitive data." Only authorized AWS staff running the European Sovereign Cloud will have access to a "replica of the source code needed to maintain" services.

The footprint of this cloud is being extended from the AWS Region in Germany across the EU to allay customers' concerns. Belgium, the Netherlands, and Portugal are set to kick off AWS Local Zones.

AWS says customers with more stringent requirements for data isolation or data residency can use its Dedicated Local Zones, AI Factories, or Outposts in the preferred locations they select, including on-prem.

For the uninitiated, AWS Local Zones are built to provide low-latency access to services in specific cities. This same capability is provided by AWS Dedicated Local Zones, but these are created for the exclusive use by one customer or community so by their nature are meant to offer additional security, governance and data residency features for sovereign workloads.

Customers will keep all metadata they create (roles, permissions, resource labels, and configurations) only in the EU, including sovereign Identity and Access Management (IAM), billing, and usage metering systems.

EU citizens "obligated to abide by European law" will run a new parent company and three local subsidiaries incorporated in Germany that manage the AWS European Sovereign Cloud. An advisory board was also set up, comprising three Amazon staff and two independent board members.

looorg writes:

Apparently there is a new app. If you don't use your phone enough or if there is no activity it informs your next of kin that you might be dead.

A new bleak-sounding app has taken China by storm.

Named Are You Dead? the concept is simple. You need to check in with it every two days – clicking a large button – to confirm that you are alive. If not, it will get in touch with your appointed emergency contact and inform them that you may be in trouble.

It was launched in May last year to not much fanfare but attention around it has exploded in recent weeks with many young people, who live alone in Chinese cities, downloading it in droves.

The app, which is listed internationally under the name Demumu, ranks in the top two in the US, Singapore and Hong Kong, and top four in Australia and Spain for paid utility apps - possibly driven by Chinese users living overseas.

Still if you are so lonely and have no friends or family. Who will it contact? Is the contact then "big-brother"?

https://www.cbsnews.com/news/are-you-dead-viral-chinese-app-changing-its-name/
https://www.euronews.com/next/2026/01/14/what-is-the-are-you-dead-app-and-why-is-it-so-popular-in-china-and-beyond
https://www.bbc.com/news/articles/c3381r5nnn6o

Original Submission

canopic jug writes:

The Absurd Pirate's Internet Blog asks, is gen alpha screwed?:

However, I do think there is a STARK contrast between a curated show from the 90s-00s and a show like Cocomelon that is designed to be like heroin for babies. I walked in on my MIL and daughter watching Cocomelon together one time, and it was jarring seeing how, for one, low effort the animation and songs are, and two, how stimulating this show is, between the incredibly saturated colors to the jump cuts every second. What I learned was that this show uses focus groups of children to make it so there is not a break in the concentration. If a kid shifts his eyes away from the screen, the scene gets edited to address that.

Companies are literally designing everything for addiction these days. Trying to get you hooked on whatever they can profit off of as early in your development as possible.

The points raised there are discussed further by Andre Franca. He adds,

The author also mentions the "mental death" of parenting under modern life, and I totally get that. There are days when I'm so drained that a screen feels like a life raft, so the comparison of high-stimulant shows to "baby heroin" makes total sense to me. That crap is bad enough for an adult; for a child, it can be devastating. I've watched my oldest son's behavior shift in real-time depending on what he's consuming. When it's junk, he turns into different person - more reactive, less patient. It makes me realize that parenting today is largely about shielding them from a culture that wants to outsource their development to an algorithm.

What happens when a substantial portion of a whole generation achieves an age of majority with an nearly complete substitution of life experience for exposure to mindless digital heroin?

Previously:
(2025) Ban Social Media for Under 15s, Says French Report Warning of TikTok Dangers
(2025) Social Media Is Dead – Here's What Comes Next
(2015) Kids These Days: Six or Seven Nicknamed Generations

Original Submission

Arthur T Knackerbracket writes:

It's no surprise to see similar designs appearing in Ukraine's interceptor drone fleets:

It's rare that drone hobbyists can hang with the professionals, especially when it comes to setting records. But hobbyist drone developer and videographer Luke Bell and his dad have recaptured the world record for the world's fastest first-person drone, reaching 408 miles per hour [657 km/hr -Ed.].

The Bells have been competing in this niche game of record-breaking cat-and-drone for the past year. As TechSpot details, they set a record just shy of 300 miles per hour in June 2025, only for fellow YouTube content creator Samgo to hit 346mph with his Fastboy 2. The Bells responded with a new design of their own that upgraded cooling, aerodynamics, and power, managing 363mph, only for professional drone videographers Ben Biggs and Aiden Kelley to steal the crown with a new 389mph drone flight.

Now, the Bells are back, and this time they're smashing through the ceiling. Their latest design uses new motors, a new aerodynamic chassis, and trimmed rotors to achieve the latest milestone record: 408mph.

To really show off their achievements, they built a second version of the drone and fitted it with a 360-degree camera so it could capture incredible second-hand footage of the world's fastest FPV drone. It makes a fantastic whine noise as it flies past observers, almost too fast to see.

I don't typically follow consumer drone developments, but having kept an eye on Ukraine's own innovations in this space over the past few years, it's fascinating to see this style of drone in Ukraine's Wild Hornet Sting drone interceptor. Notable for their ability to catch up to some of Russia's jet-powered one-way-attack drones, you have to imagine that Ukrainian developers will be keen to see if they can make similar speed advances to improve the interception rate moving forward.

Original Submission

upstart writes:

Pluralistic: Sorry, eh (13 Jan 2026) – Pluralistic: Daily links from Cory Doctorow:

Like all the best Americans, I'm Canadian, and while I have lived abroad for most of this century, I still hew faithfully to our folkways, which is why I'd like to start this essay by apologizing.

I'm sorry.

I'm sorry! I'm a technology writer, which means I'm supposed to be encouraging you to throw hundreds of billions of dollars at the money-losingest technology in human history, AI. No one has ever lost as much money as the AI companies.

There is no way to operate one of Nvidia's big AI-optimized GPUs without losing money. The owners of these GPUs who have lost the least money are the ones who rushed into buying GPUs without ensuring they'd have electricity to power them, and have been forced to leave their GPUs to age in warehouses. The minute they plug in those GPUs, they'll start losing money, and the more they use them, the more money they'll lose.

I'm sorry. As a technology writer, I'm supposed to be telling you that this bet will some day pay off, because one day we will have shoveled so many words into the word-guessing program that it wakes up and learns how to actually do the jobs it is failing spectacularly at today. This is a proposition akin to the idea that if we keep breeding horses to run faster and faster, one of them will give birth to a locomotive. Humans possess intelligence, and machines do not. The difference between a human and a word-guessing program isn't how many words the human knows.

I'm sorry. I know that when we talk about "digital sovereignty," we're obliged to talk about how we can build more data-centres that we can fill up with money-losing chips from American silicon monopolists in the hopes of destroying as many jobs as possible while blowing through our clean energy goals and enshittifying as much of our potable water as possible.

I don't have any advice for how to do that. I'm sorry!

As Canada contemplates our response to the collapse of the American empire and its alliances with the world, the cornerstone of our current strategy is sacrificing our dollars, water and energy in order to become more dependent on America, in a weird and improbable bet that we will figure out how to make millions of Canadians unemployed. I'm sorry, that just doesn't sound like a great idea to me.

If I can beg your indulgence, I'd like to propose an alternative.

An Anonymous Coward writes:

Study of 20 years of kernel history finds bugs hide for 2+ years on average, some for decades:

It was not too long ago we talked about the first Rust CVE in the Linux kernel, which caused system crashes. That same day, 159 other CVEs were issued for C code. While that shows progress with Rust, it also highlights something more concerning; the kernel has bugs that hide for years before anyone finds them.

A research blog published on Pebblebed demonstrates how bugs often stay hidden for years before they are discovered and fixed.

Jenny Guanni Qu, a researcher at Pebblebed, analyzed 125,183 bugs from 20 years of Linux kernel development history (on Git). The findings show that the average bug takes 2.1 years to find. The longest-lived bug, a buffer overflow in networking code, went unnoticed for 20.7 years!

The research was carried out by relying on the Fixes: tag that is used in kernel development. Basically, when a commit fixes a bug, it includes a tag pointing to the commit that introduced the bug.

Jenny wrote a tool that extracted these tags from the kernel's git history going back to 2005. The tool finds all fixing commits, extracts the referenced commit hash, pulls dates from both commits, and calculates the time frame.

As for the dataset, it includes over 125k records from Linux 6.19-rc3, covering bugs from April 2005 to January 2026. Out of these, 119,449 were unique fixing commits from 9,159 different authors, and only 158 bugs had CVE IDs assigned.

Plus, she found out that different parts of the kernel show significant variation in how long bugs remain hidden. CAN bus drivers have the longest average at 4.2 years, followed by SCTP networking at 4.0 years. GPU bugs get caught fastest at 1.4 years, and BPF bugs are found within 1.1 years.

The research also found that incomplete fixes are common. Someone notices undefined behavior and ships a fix, but the fix does not fully address the problem. In one case, a 2024 fix for netfilter set field validation was incomplete, and a security researcher found a bypass a year later.

Jenny's research goes much deeper than what I covered here. She has also developed an AI model called VulnBERT that predicts whether a commit introduces a vulnerability. The detailed blog post linked above includes elaborate technical explanations on that; it is a must-read!

Original Submission

jelizondo writes:

phys.org published an interesting article about a new hypothesis regarding the existence of worm holes:

Wormholes are often imagined as tunnels through space or time—shortcuts across the universe. But this image rests on a misunderstanding of work by physicists Albert Einstein and Nathan Rosen.

In 1935, while studying the behavior of particles in regions of extreme gravity, Einstein and Rosen introduced what they called a "bridge": a mathematical link between two perfectly symmetrical copies of spacetime. It was not intended as a passage for travel, but as a way to maintain consistency between gravity and quantum physics. Only later did Einstein–Rosen bridges become associated with wormholes, despite having little to do with the original idea.

But in new research published in Classical and Quantum Gravity, my colleagues and I show that the original Einstein–Rosen bridge points to something far stranger—and more fundamental—than a wormhole.

The puzzle Einstein and Rosen were addressing was never about space travel, but about how quantum fields behave in curved spacetime. Interpreted this way, the Einstein–Rosen bridge acts as a mirror in spacetime: a connection between two microscopic arrows of time.

Quantum mechanics governs nature at the smallest scales such as particles, while Einstein's theory of general relativity applies to gravity and spacetime. Reconciling the two remains one of physics' deepest challenges. And excitingly, our reinterpretation may offer a path to doing this.

The "wormhole" interpretation emerged decades after Einstein and Rosen's work, when physicists speculated about crossing from one side of spacetime to the other, most notably in the late-1980s research.

But those same analyses also made clear how speculative the idea was: within general relativity, such a journey is forbidden. The bridge pinches off faster than light could traverse it, rendering it non-traversable. Einstein–Rosen bridges are therefore unstable and unobservable—mathematical structures, not portals.

Yet there is no observational evidence for macroscopic wormholes, nor any compelling theoretical reason to expect them within Einstein's theory. While speculative extensions of physics—such as exotic forms of matter or modifications of general relativity—have been proposed to support such structures, they remain untested and highly conjectural.

Our recent work revisits the Einstein–Rosen bridge puzzle using a modern quantum interpretation of time, building on ideas developed by Sravan Kumar and João Marto.

Arthur T Knackerbracket writes:

https://www.techradar.com/ai-platforms-assistants/society-cannot-function-if-no-one-is-accountable-for-ai-jaron-lanier-the-godfather-of-virtual-reality-discusses-how-far-our-empathy-should-extend-to-ai-in-episode-two-of-new-podcast-the-ten-reckonings

'Society cannot function if no one is accountable for AI' — Jaron Lanier, the godfather of virtual reality

Whether we like it or not, we can't ignore AI. What started as a fun, gimmicky chatbot on our desktops, albeit one that could talk a bit like a human, is already taking jobs, accessing medical records, and reshaping workplaces. We are rapidly approaching the point where the practical realities of building and governing advanced AI systems must be confronted.

As the recent furor over indecent Grok-generated images on X, and the use of Meta AI smart glasses to record women without their permission for social media clicks has shown, the guardrails meant to help society cope with the deluge of AI devices and new technologies seem seriously lacking.

Even before the latest controversies around AI-generated images, one of the biggest shocks to me was the way some AI companies decided it was perfectly acceptable to train their models on copyrighted material from authors and artists without permission – and the fact that, despite a few lingering lawsuits, they appear to have faced few consequences so far.

All of this makes me wonder whether we're really ready for a world in which AI runs everything with zero accountability. Two people who have been grappling with similar questions are technologist Jaron Lanier and Dr Ben Goertzel, CEO of SingularityNET and founder of the ASI Alliance, in the next upcoming episode of The Ten Reckonings podcast. [5:17 --JE]

This new episode forms part of a series where these issues are explored in depth. According to Goertzel, "The ASI Alliance's purpose is not to present a unified position, but to create space for the world's leading thinkers to openly debate and, in doing so, help society reckon with the profound choices ahead."

Arthur T Knackerbracket writes:

Why flies matter, their vital role in ecosystems, and surprising diversity

Summer is the season for flies, which belong to an insect group that scientists call Diptera. Many fly species are more active when the weather is warmer, and populations can boom thanks to the interplay of winter rains, warmer weather and abundant food sources.

Mosquitoes biting and bush flies buzzing around may be irritating, but there are thousands of fly species worth appreciating. CSIRO's entomology expert, Dr. Keith Bayless, a research scientist at CSIRO's Australian National Insect Collection, answers common questions about why flies matter—from their role in ecosystems to surprising facts about their diversity and importance.

People often hear about orchid hunters who are obsessed with searching for rare orchids. My work is similar, but with rare flies. I study and track species that are rarely seen in the wild.

One of the rarest is a species named Clisa australis. The species was named in the 1960s and then not recorded again for decades. It was originally found in caves where its larvae fed on bat poo. To locate it, I searched in similar habitats that were easier to access—such as pit toilets in national parks.

That search became the basis for a new children's book, "The Very Stinky Fly Hunt," by CSIRO science communicator Andrea Wild, which introduces five- to nine-year-olds to the science of fly diversity and the thrill of discovery.

A major goal in the hunt for Clisa australis is to learn more about its close relatives—the fruit flies (Tephritidae). Of the nearly 5,000 species of fruit flies, fewer than 100 are significant agricultural pests.

Clisa australis has similar egg-laying structures to fruit flies, but it doesn't feed on plants. Understanding these differences helps answer a bigger question: what makes certain species cause problems while most others are benign or beneficial?

To explore that, I'm building genetic trees of flies to map their relationships and gain a clearer picture of where flies come from and how they evolve.

The first location where Clisa australis was rediscovered was remnant rainforest south of Sydney. A single specimen flew into my malaise trap—a tent-like trap that channels flying insects into a bottle of liquid preservative designed to protect delicate features like hair and wings.

fliptop writes:

CISA advisory warns that unauthenticated Bluetooth access in WHILL devices allows for unauthorized movement:

On December 30, the US cybersecurity agency CISA published an advisory to inform the public about a serious vulnerability discovered by researchers in electric wheelchairs made by WHILL, a Japan-based company whose personal electric mobility devices are sold around the world.

According to CISA's advisory, WHILL Model C2 and Model F electric wheelchairs are affected by a missing authentication vulnerability. The issue is tracked as CVE-2025-14346 and it has been assigned a critical severity rating.

CISA said the WHILL wheelchairs did not enforce authentication for Bluetooth connections, allowing an attacker who is in Bluetooth range of the targeted device to pair with it. The attacker could then control the wheelchair's movements, override speed restrictions, and manipulate configuration profiles, all without requiring credentials or user interaction.

The flaw was discovered by a team from QED Secure Solutions, a research-driven cybersecurity firm that helps private and government organizations secure operational technology (OT) and other critical systems.

[...] To demonstrate a high-impact theoretical scenario, the team developed an exploit designed to automatically compromise any WHILL wheelchair within proximity. SecurityWeek reviewed a video demonstration of this exploit, which showed a wheelchair being remotely driven off a flight of stairs at high speed.

Originally spotted on Schneier on Security.

Original Submission

jelizondo writes:

Ars Technica published an interesting article about a new AI assistant that provides strong assurances that user data is unreadable even to the platform operator,

Moxie Marlinspike—the pseudonym of an engineer who set a new standard for private messaging with the creation of the Signal Messenger—is now aiming to revolutionize AI chatbots in a similar way.

His latest brainchild is Confer, an open source AI assistant that provides strong assurances that user data is unreadable to the platform operator, hackers, law enforcement, or any other party other than account holders. The service—including its large language models and back-end components—runs entirely on open source software that users can cryptographically verify is in place.

Data and conversations originating from users and the resulting responses from the LLMs are encrypted in a trusted execution environment (TEE) that prevents even server administrators from peeking at or tampering with them. Conversations are stored by Confer in the same encrypted form, which uses a key that remains securely on users' devices.

Like Signal, the under-the-hood workings of Confer are elegant in their design and simplicity. Signal was the first end-user privacy tool that made using it a snap. Prior to that, using PGP email or other options to establish encrypted channels between two users was a cumbersome process that was easy to botch. Signal broke that mold. Key management was no longer a task users had to worry about. Signal was designed to prevent even the platform operators from peering into messages or identifying users' real-world identities.

All major platforms are required to turn over user data to law enforcement or private parties in a lawsuit when either provides a valid subpoena. Even when users opt out of having their data stored long term, parties to a lawsuit can compel the platform to store it, as the world learned last May when a court ordered OpenAI to preserve all ChatGPT users' logs—including deleted chats and sensitive chats logged through its API business offering. Sam Altman, CEO of OpenAI, has said such rulings mean even psychotherapy sessions on the platform may not stay private. Another carve out to opting out: AI platforms like Google Gemini may have humans read chats.

"AI models are inherent data collectors," Em [she keeps her last name off the Internet] told Ars. "They rely on large data collection for training, improvements, operations, and customizations. More often than not, this data is collected without clear and informed consent (from unknowing training subjects or from platform users), and is sent to and accessed by a private company with many incentives to share and monetize this data."

Arthur T Knackerbracket writes:

https://www.wired.com/story/x-didnt-fix-groks-undressing-problem-it-just-makes-people-pay-for-it/

After creating thousands of "undressing" pictures of women and sexualized imagery of apparent minors, Elon Musk's X has apparently limited who can generate images with Grok. However, despite the changes, the chatbot is still being used to create "undressing" sexualized images on the platform.

On Friday morning, the Grok account on X started responding to some users' requests with a message saying that image generation and editing are "currently limited to paying subscribers." The message also includes a link pushing people towards the social media platform's $395 annual subscription tier. In one test of the system requesting Grok create an image of a tree, the system returned the same message.

The apparent change comes after days of growing outrage against and scrutiny of Musk's X and xAI, the company behind the Grok chatbot. The companies face an increasing number of investigations from regulators around the world over the creation of nonconsensual explicit imagery and alleged sexual images of children. British prime minister Keir Starmer has not ruled out banning X in the country and said the actions have been "unlawful."

Neither X nor xAI, the Musk-owned company behind Grok, has confirmed that it has made image generation and editing a paid-only feature. An X spokesperson acknowledged WIRED's inquiry but did not provide comment ahead of publication. X has previously said it takes "action against illegal content on X," including instances of child sexual abuse material. While Apple and Google have previously banned apps with similar "nudify" features, X and Grok remain available in their respective app stores. xAI did not immediately respond to WIRED's request for comment.

For more than a week, users on X have been asking the chatbot to edit images of women to remove their clothes—often asking for the image to contain a "string" or "transparent" bikini. While a public feed of images created by Grok contained far fewer results of these "undressing" images on Friday, it still created sexualized images when prompted to by X users with paid for "verified" accounts.