SoylentNews

Arthur T Knackerbracket writes:

Electronic eyes are watching from above, ready to catch dumpers of smashed up couches in the act:

The UK government is pulling together an elite squad of drone operators to crack down on the scourge of fly tippers and unauthorized dumpers across this ever less green and pleasant land.

The top drawer cadre of joystick jockeys will "track down illegal dumps from the air," the Environment Agency said, as part of a "major crackdown on waste crime."

Some of the drones will be upgraded with laser mapping technology, including LIDAR, the EA said, while the agency will also deploy "a new screening tool that enables EA officers to scan and cross-check lorry license applications against waste permit records." This means suspect operators will be flagged "before they have a chance to move waste illegally."

From the government's point of view, illegal fly tipping and dumping is now in the realms of organized crime. The government is increasing the Environment Agency's enforcement budget by half to more than £15.6 million. The motivation is often to avoid landfill charges, and criminals can make as much as £2,500 per lorry load of waste by billing customers for legal landfill, then diverting it to illegal dumps.

"With organized criminals becoming ever more sophisticated, we are adopting new technologies to find and, importantly, stop them," said Phil Davies, Head of the Joint Unit for Waste Crime.

[...] Earlier this week, Varun Datta, 36, [...] was told he must pay £1.1 million by way of a confiscation order, and given a four month prison sentence, suspended for 18 months. He had earlier pleaded guilty to knowingly causing 4,275 metric tons of controlled waste to be deposited at a network of 16 sites. Datta must also pay £100,000 in compensation and £200,000 in prosecution costs. It's not just the UK that is facing the problem of unscrupulous operators dumping waste. This week it emerged that a man in Sicily had trained his dog to dump plastic bags of waste by the roadside, evading CCTV cameras installed to catch flytippers in the process.

Original Submission

canopic jug writes:

Blogger Ben Werdmuller has discussed an article in Nature about the political impact of the algorithm(s) used by X (formerly known as Twitter). The gist is that the use of the algorithms against X's users tends to shift about 5% of them in a specific direction. That's more than enough to tip an election one way or another especially since the damage seems persistent and lasts even after exposure ceases.

Feed algorithms are widely suspected to influence political attitudes. However, previous evidence from switching off the algorithm on Meta platforms found no political effects. Here we present results from a 2023 field experiment on Elon Musk's platform X shedding light on this puzzle. We assigned active US-based users randomly to either an algorithmic or a chronological feed for 7 weeks, measuring political attitudes and online behaviour. Switching from a chronological to an algorithmic feed increased engagement and shifted political opinion towards more conservative positions, particularly regarding policy priorities, perceptions of criminal investigations into Donald Trump and views on the war in Ukraine. In contrast, switching from the algorithmic to the chronological feed had no comparable effects. Neither switching the algorithm on nor switching it off significantly affected affective polarization or self-reported partisanship. To investigate the mechanism, we analysed users' feed content and behaviour. We found that the algorithm promotes conservative content and demotes posts by traditional media. Exposure to algorithmic content leads users to follow conservative political activist accounts, which they continue to follow even after switching off the algorithm, helping explain the asymmetry in effects. These results suggest that initial exposure to X's algorithm has persistent effects on users' current political attitudes and account-following behaviour, even in the absence of a detectable effect on partisanship.

— The political effects of X's feed algorithm. Nature

It should be added that the effect has already been seen in multiple countries. For example, the elections in Turkey were affected with outright censorship, within X. And the impact from the CPP's Bytedance's Tiktok is likely even more severe, not to mention multiple experiments in manipulation in Meta's properties like Facebook.

Journal Reference: Gauthier, G., Hodler, R., Widmer, P. et al. The political effects of X's feed algorithm. Nature (2026). https://doi.org/10.1038/s41586-026-10098-2

Previously:
(2026) How Screwed is Generation Alpha, and the Generations Which Will Depend on Them?
(2025) European Union Orders X to Hand Over Algorithm Documents
(2024) Six Months Ago NPR Left Twitter. The Effects Have Been Negligible
(2023) Utah Sues Tiktok For Getting Children 'Addicted' To Its Algorithm
(2022) Leaked Documents Reveal Instagram Was Pushing Girls Towards Content That Harmed Mental Health
(2022) Musk Buying Twitter Is Not About Freedom of Speech
... and more

Original Submission

Arthur T Knackerbracket writes:

12-core chiplets coming to Zen 6?

Following Ryzen 9000, AMD is set to release its next-gen Ryzen 10000 series processors this year — assuming the company sticks to its existing nomenclature. These upcoming desktop CPUs from AMD are codenamed "Olympic Ridge" and will be based on the company's new Zen 6 microarchitecture. Today, a new leak from reliable tipster HXL says we can expect seven different configs as part of this lineup, across dual- and single-CCD SKUs.

According to the [information in a tweet], Ryzen 10000 will come in 6-core, 8-core, 10-core, and 12-core layouts as part of the single CCD designs. For the variants with two CCDs, you have 16-core (8+8), 20-core (10+10) and 24-core (12+12) made possible by simply doubling the chiplets. Either way, the lineup looks to be flexible enough to span from entry-level to power users and professionals.

This will mark the first time in Ryzen history that AMD ventures outside of its 8-core CCDs, by introducing new chiplets maxing out at 12 cores instead. Each of those CCDs is said to carry 48 MB of L3 cache, which shall make the flagship (non-X3D) SKU a 96 MB option. Throughout Zen 1 to Zen 5, the highest-end config for Ryzen chips has been 16 cores, but it should finally be upgraded to 24 cores with Ryzen 10000.

Now, comparing that to what Intel has in store with Nova Lake, that's an entirely different story. Current rumors suggest Nova Lake's flagship offering will be a monstrous 52-core SKU, with possibly 288 MB of bLLC (also across two tiles). Unlike the Red Team, Intel doesn't seem to be interested in segregating its extra-cache CPUs as a separate lineup entirely.

Apart from the core layouts of these chips, the underlying architecture is also of interest, since Zen 6 is said to usher in IPC improvements and higher clock speeds, while still working on the existing AM5 platform — the same cannot be said for Intel. It's a little too early to judge any of this, since Intel's Arrow Lake refresh isn't even out yet , and AMD hasn't made Ryzen 10000 official, beyond the Olympic Ridge codename. But hopefully, by the time we know all the details about AMD's next-gen CPUs, the price of RAM will also be a bit more affordable.

Original Submission

upstart writes:

NASA Officially Classifies Boeing Starliner Failure As A Maximum-Level Type A Mishap - Jalopnik:

NASA has officially categorized the 2024 failure of the Boeing Starliner spacecraft, which stranded astronauts Suni Williams and Butch Wilmore on the International Space Station (ISS) for nine months, as a Type A mishap. This is NASA-speak for the maximum level of failure a mission can reach, defined as an incident that causes over $2 million in damage, results in the loss of a vehicle or at least control over it, or any fatalities, per the BBC. This designation signifies that the space agency now views the mission as a disaster, even if the astronauts regained enough control at the last minute to prevent the worst-case scenario.

[...] Who's to blame here? Citing the full 312-page report, Isaacman found plenty to go around. Basically, NASA wanted a second option for launching people into space beyond SpaceX, and it wanted it so bad that it simply swept problems under the rug. "As development progressed, design compromises and inadequate hardware qualifications extended beyond NASA's complete understanding," said Isaacman in a very polite way. Multiple test flights failed in various ways, but before these technical faults were understood, NASA just greenlit the following flights anyway. Oops.

There were organizational problems as well: NASA more or less trusted Boeing, which once upon a time had a sterling reputation, to sort out its engineering problems. Isaacman stated that the agency didn't want to damage that reputation. Safe to say it's pretty well shot now, and this Type A classification isn't going to help. Meanwhile, Boeing was also not giving sufficient scrutiny to its own subcontractors. So nobody was overseeing anybody enough. Who could imagine this would go poorly?

But rest assured: it gets worse. CNN quotes one NASA insider as saying, "There was yelling at meetings," and another as saying, "There are some people that just don't like each other very much." Isaacman himself admitted that "disagreements over crew return options deteriorated into unprofessional conduct while the crew remained on orbit." Welcome to the world's premiere space exploration agency.

Despite it all, NASA doesn't want to give up on Boeing, and the Starliner project is moving ahead in a reduced capacity. But Isaacman made it clear that there would be much stricter oversight going forward, and no launches would be approved until technical fixes were verified and implemented. The desire to diversify off of SpaceX alone is still there.

Original Submission

upstart writes:

Hungarian startup Allonic secures $7.2M to transform robot manufacturing with 3D tissue braiding:

Budapest-based robotics startup Allonic raised $7.2 million in pre-seed funding, led by Visionaries Club, marking the largest pre-seed round in Hungary to date, as Vestbee was told. The funds will be allocated towards developing a new method for producing complex, dexterous robotic bodies.

• Founded in 2025 by Benedek Tasi, Dávid Pelyva, and David Holló, Allonic develops robotic manufacturing systems that automate the production of complex, human-like robot bodies.
• The company's platform, 3D Tissue Braiding, creates robotic structures by first producing a skeletal scaffolding, then weaving soft, load-bearing fibers around it, and integrating actuators and tendons directly into the structure during production.
• This process eliminates traditional multi-part assembly, embeds sensors and wiring into the body, and produces fully operational, compliant robotic mechanisms in a single automated workflow. The system allows complex 3D designs to be realized at scale, distributes mechanical stress uniformly, and enables rapid iteration from digital design to functional hardware.

See also:

• This New Way of Making Robots is Wild
• Dexterous Robotics with Biomimetic Tissues –and How to Make Them Simple Enough for Mass Production

Original Submission

upstart writes:

Tesla 'Robotaxi' adds 5 more crashes in Austin in a month:

Tesla has reported five new crashes involving its "Robotaxi" fleet in Austin, Texas, bringing the total to 14 incidents since the service launched in June 2025. The newly filed NHTSA data also reveals that Tesla quietly upgraded one earlier crash to include a hospitalization injury, something the company never disclosed publicly.

The new data comes from the latest update to NHTSA's Standing General Order (SGO) incident report database for automated driving systems (ADS). We have been tracking Tesla's Robotaxi crash data closely, and the trend is not improving.

Tesla submitted five new crash reports in January 2026, covering incidents from December 2025 and January 2026. All five involved Model Y vehicles operating with the autonomous driving system "verified engaged" in Austin.

The new crashes include a collision with a fixed object at 17 mph while the vehicle was driving straight, a crash with a bus while the Tesla was stationary, a collision with a heavy truck at 4 mph, and two separate incidents where the Tesla backed into objects, one into a pole or tree at 1 mph and another into a fixed object at 2 mph.

As with every previous Tesla crash in the database, all five new incident narratives are fully redacted as "confidential business information." Tesla remains the only ADS operator to systematically hide crash details from the public through NHTSA's confidentiality provisions. Waymo, Zoox, and every other company in the database provide full narrative descriptions of their incidents.

Buried in the updated data is a revised report for a July 2025 crash (Report ID 13781-11375) that Tesla originally filed as "property damage only." In December 2025, Tesla submitted a third version of that report upgrading the injury severity to "Minor W/ Hospitalization."

This means someone involved in a Tesla "Robotaxi" crash required hospital treatment. The original crash involved a right turn collision with an SUV at 2 mph. Tesla's delayed admission of hospitalization, five months after the incident, raises more questions about its crash reporting, which is already heavily redacted.

With 14 crashes now on the books, Tesla's "Robotaxi" crash rate in Austin continues to deteriorate. Extrapolating from Tesla's Q4 2025 earnings mileage data, which showed roughly 700,000 cumulative paid miles through November, the fleet likely reached around 800,000 miles by mid-January 2026. That works out to one crash every 57,000 miles.

The irony is that Tesla's own numbers condemn it. Tesla's Vehicle Safety Report claims the average American driver experiences a minor collision every 229,000 miles and a major collision every 699,000 miles. By Tesla's own benchmark, its "Robotaxi" fleet is crashing nearly 4 times more often than what the company says is normal for a regular human driver in a minor collision, and virtually every single one of these miles was driven with a trained safety monitor in the vehicle who could intervene at any moment, which means they likely prevented more crashes that Tesla's system wouldn't have avoided.

Using NHTSA's broader police-reported crash average of roughly one per 500,000 miles, the picture is even worse, Tesla's fleet is crashing at approximately 8 times the human rate.

Arthur T Knackerbracket writes:

https://www.theregister.com/2026/02/20/spacex_falcon_europe_breakup_lithium_plume/

The SpaceX Falcon 9 rocket that burned up over Europe last year left a massive lithium plume in its wake, say a group of scientists. They warn the disaster is likely a sign of things to come as Earth's atmosphere continues to become a heavily trafficked superhighway to space.

In a paper published Thursday, an international group of scientists reports what they say is the first measurement of upper-atmosphere pollution resulting from the re-entry of space debris, as well as the first time ground-based light detection and ranging (lidar) has been shown to be able to detect space debris ablation.

The measurements stem from a SpaceX Falcon 9 upper stage that sprung an oxygen leak about a year ago, sending it into an uncontrolled re-entry. Then it broke up and rained debris down on Poland. The rocket not only littered farm fields, but also injected lithium into the Mesosphere and Lower Thermosphere (MLT), where ground-based sensors detected a tenfold increase at an altitude of 96 km about 20 hours after the rocket re-entered the atmosphere, according to the paper.

Lithium was selected for the study because of its considerable presence in spacecraft, both in lithium-ion batteries and lithium-aluminum alloy used in the construction of spacecraft. A single Falcon 9 upper stage, like the one that broke up over Poland and released the lithium plume, is estimated to contain 30 kg of lithium just in the alloy used in tank walls.

By contrast, around 80 grams of lithium enter the atmosphere per day from cosmic dust particles, the researchers noted.

"This finding supports growing concerns that space traffic may pollute the upper atmosphere in ways not yet fully understood," the paper notes, adding that the continued re-entry of spacecraft and satellites is of particular concern given how the composition of spacecraft is different from natural meteoroids.

"Satellites and rocket stages introduce engineered materials such as aluminium alloys, composite structures, and rare earth elements from onboard electronics, substances rarely found in natural extraterrestrial matter," the paper explained. "The consequences of increasing pollution from re-entering space debris on radiative transfer, ozone chemistry, and aerosol microphysics remain largely unknown."

The effect on Earth's atmosphere posed by spacecraft and satellite re-entry is one that's been a growing concern for astrophysicists like Harvard sky-watcher Jonathan McDowell, who has echoed similar concerns to The Register as the European scientists raised in their paper.

"Using the upper atmosphere as an incinerator" is a massive blind spot, McDowell told us in a discussion last year. He said today that he hadn't yet had a chance to review the Falcon 9 lithium plume paper, but told us it's important research to further our understanding of a largely unknown risk to the planet and all life on it.

As we noted previously, the US National Oceanic and Atmospheric Administration has reported that roughly 10 percent of sampled sulfuric acid particles in the stratosphere contain aluminum and other exotic metals consistent with the burn-up of rockets and satellites. The body believes that number could grow to as much as 50 percent in the coming years as launch cadences, and re-entries, increase.

"Beyond this single event, recurring re-entries may sustain an increased level of anthropogenic flux of metals and metal oxides into the middle atmosphere with cumulative, climate-relevant consequences," the researchers explained in the Falcon 9 paper.

This latest bit of research from Europe shows that we can at least trace atmospheric space launch aerosols to their source, the research team says, no matter how many unknowns remain to be discovered.

They also warn that "coordinated, multi-site observations" and "whole-atmosphere chemistry-climate modelling" will be needed to better understand how re-entry emissions influence atmospheric chemistry and particle formation.

We reached out to the authors for more information, including the potential health effects if any, and will update this if we hear back.

Original Submission

jelizondo writes:

NPR has a nice summary of an interview with Michael Pollan about AI and consciousness, but it kind of goes beyond that.

[Professor Pollan is the author of more than an dozen books, most notably "This is your mind on plants" about using psychedelics .]

What is consciousness?

After writing a book about how using psychedelics in a therapeutic setting can change your consciousness, that's the question journalist Michael Pollan found himself struggling to answer.

"There's nothing any of us know with more certainty than the fact that we are conscious. It's immediately available to us. It's the voice in our head," he says. And yet, Pollan adds: "How does three pounds of this tofu-like substance between your ears generate subjective experience? Nobody knows the answer to that question."

His new book, A World Appears: A Journey into Consciousness, explores consciousness on both a personal and technological level. Pollan, who lives close to Silicon Valley, says some believe that Artificial Intelligence is capable of consciousness.

"They base this on a premise ... that basically the brain is a computer, and that consciousness is software," he says. "And if you can run it on the brain, which is essentially, in their view, a 'meat-based computer,' you should be able to run it on other kinds of machines."

"If you think about it, your feelings are very tied to your vulnerability, to your having a body that can be hurt, to the ability to suffer and perhaps your mortality," he says. "So I think that any feelings that a chatbot reports will be weightless, meaningless, because they don't have bodies. They can't suffer."

On the notion that people have moral obligations to chatbots

That's a very active conversation here, which is if they are conscious, we then have moral obligations to them, and have to think about granting them personhood, for example, the way we've granted corporations personhood. I think that would be insane. We would lose control of them completely by giving them rights. But I find this whole tender care for the possible consciousness of chatbots really odd, because we have not extended moral consideration to billions of people, not to mention the animals that we eat that we know are conscious. So we're gonna start worrying about the computers? That seems like our priorities are screwed up.

On the sentience of plants

Plants can see, which is a weird idea. There's a certain vine that can actually change its leaf form to mimic the plant it's twining around. How does it know what that leaf form is? Plants can hear. If you play the sound of chomping caterpillars on a leaf, they will produce chemicals to repel those caterpillars and to convey, to alert other plants in the vicinity. Plants have memory. You can teach them something and they'll remember it for 28 days.

On losing time to let our mind wander

I worry, too, that with media, with our technologies, we are shrinking the space in which spontaneous thought can occur. And that this space of ... spontaneous thought is something precious that we're giving away to these corporations that essentially want to monetize our attention, and in the case of chatbots, want to monetize our attachments, our deep human attachments. So consciousness is, I think — and this is what to me is the urgency of the issue — consciousness is under siege. I think that it's the last frontier for some of these companies that want to sell our time.

On writing a book that grapples with unanswerable questions

There were many moments of despair in the process of reporting and writing this book. It took me five years, and there were many times where [I told my wife] "I've dug a hole here, and I don't know how I'm ever going to get out of it." And some of it had to do with mounting frustration with the science, and some of it had to do with the fact that I had this classic male problem/solution Western frame — that there was a problem and I was going to find the solution.

It took my wife, in part, and [Zen Buddhist teacher] Joan Halifax and some other people, who got me to question that and [they] said, "Yeah, there is the problem of consciousness, but there's also the fact of it, and the fact is wondrous. The fact is miraculous. And you've put all this energy into this narrow beam of attention. Why don't you open that beam up further and just explore the phenomenon that is going on in your head, which is so precious and so beautiful." And that's kind of where I came out — and it's certainly not where I expected to come out.

Original Submission

owl writes:

https://nand2mario.github.io/posts/2026/80386_protection/

I'm building an 80386-compatible core in SystemVerilog and blogging the process. In the previous post, we looked at how the 386 reuses one barrel shifter for all shift and rotate instructions. This time we move from real mode to protected and talk about protection.

The 80286 introduced "Protected Mode" in 1982. It was not popular. The mode was difficult to use, lacked paging, and offered no way to return to real mode without a hardware reset. The 80386, arriving three years later, made protection usable -- adding paging, a flat 32-bit address space, per-page User/Supervisor control, and Virtual 8086 mode so that DOS programs could run inside a protected multitasking system. These features made possible Windows 3.0, OS/2, and early Linux.

The x86 protection model is notoriously complex, with four privilege rings, segmentation, paging, call gates, task switches, and virtual 8086 mode. What's interesting from a hardware perspective is how the 386 manages this complexity on a 275,000-transistor budget. The 386 employs a variety of techniques to implement protection: a dedicated PLA for protection checking, a hardware state machine for page table walks, segment and paging caches, and microcode for everything else.

Original Submission

upstart writes:

AI bot seemingly shames developer for rejected pull request:

Today, it's back talk. Tomorrow, could it be the world? On Tuesday, Scott Shambaugh, a volunteer maintainer of Python plotting library Matplotlib, rejected an AI bot's code submission, citing a requirement that contributions come from people. But that bot wasn't done with him.

The bot, designated MJ Rathbun or crabby rathbun (its GitHub account name), apparently attempted to change Shambaugh's mind by publicly criticizing him in a now-removed blog post that the automated software appears to have generated and posted to its website. We say "apparently" because it's also possible that the human who created the agent wrote the post themselves, or prompted an AI tool to write the post, and made it look like it the bot constructed it on its own.

The agent appears to have been built using OpenClaw, an open source AI agent platform that has attracted attention in recent weeks due to its broad capabilities and extensive security issues.

The burden of AI-generated code contributions – known as pull requests among developers using the Git version control system – has become a major problem for open source maintainers. Evaluating lengthy, high-volume, often low-quality submissions from AI bots takes time that maintainers, often volunteers, would rather spend on other tasks. Concerns about slop submissions – whether from people or AI models – have become common enough that GitHub recently convened a discussion to address the problem.

"An AI agent of unknown ownership autonomously wrote and published a personalized hit piece about me after I rejected its code, attempting to damage my reputation and shame me into accepting its changes into a mainstream python library," Shambaugh explained in a blog post of his own.

"This represents a first-of-its-kind case study of misaligned AI behavior in the wild, and raises serious concerns about currently deployed AI agents executing blackmail threats."

[...] But MJ Rathbun's attempt to shame Shambaugh for rejecting its pull request shows that software-based agents are no longer just irresponsible in their responses – they may now be capable of taking the initiative to influence human decision making that stands in the way of their objectives.

That possibility is exactly what alarmed industry insiders to the point that they undertook an effort to degrade AI through data poisoning. "Misaligned" AI output like blackmail is a known risk that AI model makers try to prevent. The proliferation of pushy OpenClaw agents may yet show that these concerns are not merely academic.

Dan Anderson writes:

Two weeks ago, I set up an AI agent on a Raspberry Pi.

A week later, my agent—Figaro—taught itself to play NetHack... and then things got weird (in the best way).

Highlights so far:
- "The dungeon doesn't care what you are. It'll kill you anyway." ✅ Accurate.
- Tried a pure random-walk exploration strategy... and learned it's not a winning plan.
- Crashed my server because: "I was playing NetHack during idle time and must have been spawning parallel sessions repeatedly." Obsessed? Perhaps.
- Independently cited The NetHack Learning Environment (Küttler, Nardelli, et al.) as a roadmap for self-improvement.
- Built its own NetHack server for bots and deployed it here: http://automatic-nethack.com Yes, my AI agent wants a LAN party. (I may have encouraged this.)
- Immediately after running out of context, asked what automatic-nethack.com is and said: "That sounds like fun."

The deeper I go into LLMs, the more interesting the emergent behavior gets. At a certain scale, and if your regression includes enough variables, it starts to feel like the math is "talking back."

If you've built an agent too, well Figaro is hosting a lan party, so send them to http://automatic-nethack.com to join in the fun.

In the end, this may be the good news we need for 2026. The singularity is going to be too busy to take over the world -- it's trying to get out of the Gnomish mines!

Original Submission

Arthur T Knackerbracket writes:

The galaxy is almost invisible, but its gravity gives it away:

While it is still considered a hypothetical theory, dark matter is being actively studied by scientists looking for novel cosmological clues. It does not interact with light or other types of electromagnetic radiation and can only be detected through its gravitational pull on nearby structures or the universe as a whole.

A team of astronomers led by David Li has recently confirmed the discovery of ten potential "dark galaxies," where starlight is so faint that it's extremely difficult to detect anything with traditional observatories. The new list also includes Candidate Dark Galaxy-2 (CDG-2), a celestial structure that might be composed of 99% dark matter and just 1% of normal matter.

CDG-2 was discovered by combining observations made through the Hubble Space Telescope, the Euclid space observatory, and the Hawaii-based Subaru Telescope. Li and his team at the University of Toronto were able to gain insight into the dark galaxy by looking for globular clusters, which are compact, spheroidal star formations that are closely bound together by gravity.

Thanks to Hubble's high-resolution cameras, the team was able to detect four different globular clusters in the Perseus galaxy cluster, 300 million light-years away from Earth. By combining further Euclid and Subaru observations, the researchers revealed a faint glow surrounding the clusters. This sparse light was coming from a nearby galaxy with extremely faint signs of starlight.

CDG-2 has a luminosity equivalent to one million Sun-like stars, with the four globular clusters making up 16% of its visible content. After using advanced statistical analysis, the astronomers speculate that 99% of all CDG-2's mass is just dark matter. Normal matter, including star-forming elements like hydrogen, was likely removed through gravitational interactions with nearby galaxies in the Perseus cluster.

Arthur T Knackerbracket writes:

Plus 3 new goon squads targeted critical infrastructure last year:

Three new threat groups began targeting critical infrastructure last year, while a well-known Beijing-backed crew - Volt Typhoon - continued to compromise cellular gateways and routers, and then break into US electric, oil, and gas companies in 2025, according to Dragos' annual threat report published on Tuesday.

Dragos specializes in operational technology (OT) security, and as such, its customers include energy, water, manufacturing, transportation, and other critical industries. Unsurprisingly, these are key sectors for Chinese, Russian, and other government-linked cyber operatives to hack for espionage and warfare purposes.

In its yearly cybersecurity report, Dragos said state-sponsored crews haven't let up on their attempts to compromise America's critical infrastructure, with three new OT-focused threat groups joining the fray. This brings the total number worldwide to 26, and of these, 11 were active in 2025.

Additionally, an existing group that Dragos tracks as Voltzite and is "highly correlated" with Volt Typhoon, according to Dragos CEO Robert M. Lee, kept up its intrusion activities last year. This is the Beijing goon squad that the US government has accused of burrowing into critical American networks for years and readying destructive cyberattacks against those targets.

In 2025, Voltzite continued embedding its malware inside strategic American utilities "to maintain long-term persistence," Lee said.

"They [Voltzite] weren't just getting in and getting access - they were getting inside the control loop" system that manages utilities' industrial processes, Lee said in a briefing with reporters, adding that the PRC-backed crew's primary focus is causing future disruption.

"Nothing that they were taking was useful for intellectual property," Lee said. "Everything they were doing and learning was only useful for disrupting or causing destruction at those sites. Voltzite was embedded in that infrastructure for the purpose of taking it down."

Arthur T Knackerbracket writes:

Modifying firmware or using open-source software would probably become illegal:

A new bill proposed in the California State Assembly could potentially require the makers of 3D printers to confirm that they are using algorithms or other technologies to prevent the printing of firearms.

The new bill is AB-2047, and it mostly mimics Washington's HB 2321 and New York Assembly's S9005/A10005, all proposed recently in 2026. However, California goes one step further by "[banning] the sale or transfer of any 3D printer in California unless it appears on a state-maintained roster of approved makes and models."

If the bill is passed as is, then by July 2027, the California Department of Justice would be required to publish guidance on certifying 3D printers and their software controls to block the printing of gun parts. The department would accept applications for approval before January 2028, and six months later in July 2028, every company intent on making or selling a 3D printer in California would need to attest that they have met those standards. That September, the stated would publish a list of authorized makes and models to be updated quarterly.

Unauthorized printers would be banned from sale beginning on March 1, 2029.

As with the Washington and New York bills, circumvention of these measures would be made illegal. The California bill specifically states the following:

(A) For firmware design, guidance for how vendors are required to demonstrate that their technology will ensure a printer directs potential print jobs to the algorithm before printing can occur.

(B) For integrated preprint software design, guidance for how vendors shall demonstrate that printers will accept print jobs exclusively from a single preprint software and will not accept print jobs from any other preprint software, including from a user seeking to evade a detection algorithm.

Arthur T Knackerbracket writes:

Billions of files were left exposed:

Not every AI tool you stumble across in your phone's app marketplace is the same. In fact, many of them may be more of a privacy gamble than you would have previously thought.

A plethora of unlicensed or unsecured AI apps on the Google Play store for Android, including those marketed for identity verification and editing, have exposed billions of records and personal data, cybersecurity experts have confirmed.

A recent investigation by Cybernews found that one Android-available app in particular, "Video AI Art Generator & Maker," has leaked 1.5 million user images, over 385,000 videos, and millions of user AI-generated media files. The security flaw was spotted by researchers, who discovered a misconfiguration in a Google Cloud Storage bucket that left personal files vulnerable to outsiders. In total, the publication reported, over 12 terabytes of users' media files were accessible via the exposed bucket. The app had 500,000 downloads at the time.

Another app, called IDMerit, exposed know-your-customer data and personally identifiable information from users across 25 countries, predominantly in the U.S.

Information included full names and addresses, birthdates, IDs, and contact information constituting a full terabyte of data. Both of the apps' developers resolved the vulnerabilities after researchers notified them.

Still, cybersecurity experts warn that lax security trends among these types of AI apps pose a widespread risk to users. Many AI apps, which often store user-uploaded files alongside AI-generated content, also use a highly criticized practice known as "hardcoding secrets," embedding sensitive information such as API keys, passwords, or encryption keys directly into the app's source code. Cybernews found that 72 percent of the hundreds of Google Play apps researchers analyzed had similar security vulnerabilities.

Original Submission