An anonymous reader shares a report: A brain-inspired computer chip that could supercharge artificial intelligence by working faster with much less power has been developed by researchers at IBM in San Jose, California. Their massive NorthPole processor chip eliminates the need to frequently access external memory, and so performs tasks such as image recognition faster than existing architectures do -- while consuming vastly less power.

"Its energy efficiency is just mind-blowing," says Damien Querlioz, a nanoelectronics researcher at the University of Paris-Saclay in Palaiseau. The work, published in Science, shows that computing and memory can be integrated on a large scale, he says. "I feel the paper will shake the common thinking in computer architecture." NorthPole runs neural networks: multi-layered arrays of simple computational units programmed to recognize patterns in data. A bottom layer takes in data, such as the pixels in an image; each successive layer detects patterns of increasing complexity and passes information on to the next layer. The top layer produces an output that, for example, can express how likely an image is to contain a cat, a car or other object

A critical vulnerability that hackers have exploited since August, which allows them to bypass multifactor authentication in Citrix networking hardware, has received a patch from the manufacturer. Unfortunately, applying it isn't enough to protect affected systems. ArsTechnica: The vulnerability, tracked as CVE-2023-4966 and carrying a severity rating of 9.8 out of a possible 10, resides in the NetScaler Application Delivery Controller and NetScaler Gateway, which provide load balancing and single sign-on in enterprise networks, respectively. Stemming from a flaw in a currently unknown function, the information-disclosure vulnerability can be exploited so hackers can intercept encrypted communications passing between devices. The vulnerability can be exploited remotely and with no human action required, even when attackers have no system privileges on a vulnerable system.

Citrix released a patch for the vulnerability last week, along with an advisory that provided few details. On Wednesday, researchers from security firm Mandiant said that the vulnerability has been under active exploitation since August, possibly for espionage against professional services, technology, and government organizations. Mandiant warned that patching the vulnerability wasn't sufficient to lock down affected networks because any sessions hijacked before the security update would persist afterward.

The FCC has unanimously approved plans by several tech companies to use the 6GHz band for wireless devices. From a report: FCC Chair Jessica Rosenworcel proposed the new rules, which would authorize very low power (VLP) operations -- meaning their signals won't be able to go very far -- in about 850MHz of the spectrum, on September 27th. The rules will also allow devices to "use higher power levels" so long as they're geofenced to keep from interfering with actual licensed 6GHz usage, and the FCC will be taking comments on other ways it can expand 6GHz spectrum usage by technology devices.

A September Bloomberg report pointed to some of the kinds of devices the FCC's affirmative vote could open up, including in-car connections, mobile virtual or augmented reality devices, and more. The FCC originally opened up 1,200MHz of the 6GHz spectrum for unlicensed use by Wi-Fi routers and client devices (think smartphones or laptops), giving home networks far more wireless overhead than existing Wi-Fi standards already had. This new approval expands the spectrum for much more general use.

Finnish telecoms giant Nokia is to axe between 9,000 and 14,000 jobs by the end of 2026 to cut costs. From a report: The announcement was made as the company reported a 20% drop in sales between July and September. The company blamed slowing demand for 5G equipment in markets such as North America. It currently has 86,000 employees around the world, and has axed thousands of jobs since 2015. Nokia wants to cut costs by between $845m and $1.27bn by 2026, it said.

Its customers have been cutting spending amid high inflation and interest rates, it said. Advances in cloud computing and AI will need "significant investments in networks that have vastly improved capabilities," said chief executive Pekka Lundmark. "However, given the uncertain timing of the market recovery, we are now taking decisive action," he said. It said it wanted to "act quickly" by cutting costs by $422m in 2024, and $317m in 2025.

Jessica Lyons Hardcastle reports via The Register: The European Telecommunications Standards Institute (ETSI) may open source the proprietary encryption algorithms used to secure emergency radio communications after a public backlash over security flaws found this summer. "The ETSI Technical Committee in charge of TETRA algorithms is discussing whether to make them public," Claire Boyer, a spokesperson for the European standards body, told The Register. The committee will discuss the issue at its next meeting on October 26, she said, adding: "If the consensus is not reached, it will go to a vote."

TETRA is the Terrestrial Trunked Radio protocol, which is used in Europe, the UK, and other countries to secure radio communications used by government agencies, law enforcement, military and emergency services organizations. In July, a Netherlands security biz uncovered five vulnerabilities in TETRA, two deemed critical, that could allow criminals to decrypt communications, including in real-time, to inject messages, deanonymize users, or set the session key to zero for uplink interception. At the time ETSI downplayed the flaws, which it said had been fixed last October, and noted that "it's not aware of any active exploitation of operational networks."

At the time ETSI downplayed the flaws, which it said had been fixed last October, and noted that "it's not aware of any active exploitation of operational networks." It did, however, face criticism from the security community over its response to the vulnerabilities -- and the proprietary nature of the encryption algorithms, which makes it more difficult for proper pentesting of the emergency network system. "This whole idea of secret encryption algorithms is crazy, old-fashioned stuff," said security author Kim Zetter who first reported the story. "It's very 1960s and 1970s and quaint. If you're not publishing [intentionally] weak algorithms, I don't know why you would keep the algorithms secret."

Amy Hawkins reports via The Guardian: A programmer in northern China has been ordered to pay more than 1 million yuan to the authorities for using a virtual private network (VPN), in what is thought to be the most severe individual financial penalty ever issued for circumventing China's "great firewall." The programmer, surnamed Ma, was issued with a penalty notice by the public security bureau of Chengde, a city in Hebei province, on August 18. The notice said Ma had used "unauthorised channels" to connect to international networks to work for a Turkish company. The police confiscated the 1.058m yuan ($145,092) Ma had earned as a software developer between September 2019 and November 2022, describing it as "illegal income," as well as fining him 200 yuan ($27). Charlie Smith (a pseudonym), the co-founder of GreatFire.org, a website that tracks internet censorship in China, said: "Even if this decision is overturned in court, a message has been sent and damage has been done. Is doing business outside of China now subject to penalties?"

AI startup Anthropic, writing in a blog post: Neural networks are trained on data, not programmed to follow rules. With each step of training, millions or billions of parameters are updated to make the model better at tasks, and by the end, the model is capable of a dizzying array of behaviors. We understand the math of the trained network exactly -- each neuron in a neural network performs simple arithmetic -- but we don't understand why those mathematical operations result in the behaviors we see. This makes it hard to diagnose failure modes, hard to know how to fix them, and hard to certify that a model is truly safe. Neuroscientists face a similar problem with understanding the biological basis for human behavior. The neurons firing in a person's brain must somehow implement their thoughts, feelings, and decision-making. Decades of neuroscience research has revealed a lot about how the brain works, and enabled targeted treatments for diseases such as epilepsy, but much remains mysterious. Luckily for those of us trying to understand artificial neural networks, experiments are much, much easier to run. We can simultaneously record the activation of every neuron in the network, intervene by silencing or stimulating them, and test the network's response to any possible input.

Unfortunately, it turns out that the individual neurons do not have consistent relationships to network behavior. For example, a single neuron in a small language model is active in many unrelated contexts, including: academic citations, English dialogue, HTTP requests, and Korean text. In a classic vision model, a single neuron responds to faces of cats and fronts of cars. The activation of one neuron can mean different things in different contexts. In our latest paper, Towards Monosemanticity: Decomposing Language Models With Dictionary Learning , we outline evidence that there are better units of analysis than individual neurons, and we have built machinery that lets us find these units in small transformer models. These units, called features, correspond to patterns (linear combinations) of neuron activations. This provides a path to breaking down complex neural networks into parts we can understand, and builds on previous efforts to interpret high-dimensional systems in neuroscience, machine learning, and statistics. In a transformer language model, we decompose a layer with 512 neurons into more than 4000 features which separately represent things like DNA sequences, legal language, HTTP requests, Hebrew text, nutrition statements, and much, much more. Most of these model properties are invisible when looking at the activations of individual neurons in isolation.

China aims to grow the country's computing power by more than a third in less than three years, a move set to benefit local suppliers and boost technology self-reliance as US sanctions pressure domestic industry. From a report: The world's second-largest economy is targeting more than 300 exaflops of computing capacity across its tech sector by 2025 from 220 this year, according to a joint statement from several agencies including the Ministry of Industry and Information Technology. The goal marks Beijing's latest attempt to construct digital infrastructure to spur a sluggish economy. China also plans to build an additional 20 smart computing centers in two years. Bigger optical networks and more advanced data storage will be installed in the years until 2025, the regulators said. The additional computational power will support manufacturing, education, finance, transportation, healthcare and energy, they added.

A team of scientists drove hundreds of miles through the steppes of Kazakhstan in search of what may be one of the largest and most diverse fungi ecosystems on Earth.

The Washington Post believes their efforts "could help make the planet more resilient to climate change." When these underground fungi come together, they form sophisticated systems known as "mycorrhizal networks...." Mycorrhizal fungi often form mutually beneficial relationships with plants. They trade essential nutrients such as phosphorus and nitrogen in exchange for carbon, and act as an extended root system, allowing plants to access water they can't reach. These networks may also prove to be invaluable for transporting carbon underground, a study published in June found. About 13 gigatons of carbon fixed by vegetation — equivalent to about one-third of all carbon dioxide emissions from fossil fuels in one year — flows through underground fungi, according to an analysis of nearly 200 data sets.

In the steppe, these plant-fungal benefits may be short-lived, however. While deserts are a natural part of Kazakhstan's ecosystem, more than half of the country's vegetation and drylands is at risk of becoming desert as well. The main drivers are large-scale intensive agriculture and increasingly warm and dry temperatures brought by climate change.... Knowing what species of fungi live here is key to understanding how to protect them, said Bethan Manley, project officer at the Society for the Protection of Underground Networks who was on the expedition. It will help determine "where we might be able to have the most effective measures of not poisoning them with fungicides or not having harmful farming practices," she said.

An anonymous reader quotes a report from SecurityWeek: Tens of thousands of Android devices have been shipped to end-users with backdoored firmware, according to a warning from cybersecurity vendor Human Security. As part of the global cybercriminal operation called BadBox (PDF), Human Security found a threat actor relied on supply chain compromise to infect the firmware of more than 70,000 Android smartphones, CTV boxes, and tablet devices with the Triada malware. The infected devices come from at least one Chinese manufacturer but, before they are delivered to resellers, physical retail stores, and e-commerce warehouses, a backdoor was injected into their firmware. "Products known to contain the backdoor have been found on public school networks throughout the United States," Human says.

Discovered in 2016, Triada is a modular trojan residing in a device's RAM, relying on the Zygote process to hook all applications on Android, actively using root privileges to substitute system files. Over time, the malware went through various iterations and was found pre-installed on low-cost Android devices on at least two occasions. As part of the BadBox operation that Human Security discovered, the infected low-cost Android devices allow threat actors to carry out various ad-fraud schemes, including one named PeachPit, which at its peak relied on 121,000 Android and 159,000 iOS devices infected with malware, and on 39 Android, iOS, and CTV-centric apps designed to connect to a fake supply-side platform (SSP).

One of the modules delivered to the infected devices from the command-and-control (C&C) server allows the creation of WebViews that are fully hidden from the user, but which "are used to request, render, and click on ads, spoofing the ad requests to look like they're coming from certain apps, referred by certain websites, and rendered" on specific devices. BadBox, Human Security notes, also includes a residential proxy module that allows the threat actors to sell access to the victim's network. Furthermore, they can create WhatsApp messaging accounts and Gmail accounts they can then use for other malicious activities. "Finally, because of the backdoor's connection to C2 servers on BadBox-infected smartphones, tablets, and CTV boxes, new apps or code can be remotely installed by the threat actors without the device owner's permission. The threat actors behind BadBox could develop entirely new schemes and deploy them on BadBox-infected devices without any interaction from the devices' owners," Human notes.

An anonymous reader quotes a report from MIT Technology Review: Thirteen US states are now implementing underground thermal energy networks to reduce buildings' carbon emissions as part of a nationwide push to adopt cleaner energy sources. Thermal energy networks use pipe loops that connect multiple buildings and provide heating and cooling through water-source heat pumps. Geothermal heat is commonly used in these networks, but it is also possible to bring in waste heat from other buildings through the sewer system. When installed, these networks can provide efficient, fossil fuel-free heating and cooling to commercial and residential buildings. Thanks to legislative backing and widespread support from utility companies and labor unions they're likely to become an increasingly significant part of the future energy mix in the US.

"Heat is the largest source of waste energy and it's an untapped resource," says Zeyneb Magavi, co-executive director at clean energy nonprofit HEET (Home Energy Efficiency Team). "Once we have a thermal energy network, we can tap into that resource by moving it to where we need it." While the projects are still at the planning and regulatory stage in most of the 13 states, construction is already underway in some. [...] The advantages of thermal energy networks extend beyond reducing carbon emissions. Scaling them up from a few buildings to a community or utility level can also help make the grid more resilient and efficient. Magavi says every time a "loop" of thermal energy network is added to the grid, its ability to predict and manage power flow becomes more accurate. This interconnectedness helps the system become more resilient in high-stress situations.

Long-time Slashdot reader RockDoctor writes: Regardless of their other potential benefits, modern cars, and modern electric cars in particular, involve complex networks of computer code, hardware, and servo systems cooperating (?) to deliver services to the user, like acceleration, steering and braking. Slashdot nerderati know better than most that such complex networks can never show unexpected, non-designed behavior, due to the infallibility of hardware, program coders and system designers... Yeah. Right. "I'll have some of what he's been smoking!" That's Musk-grade optimism.

On Sunday evening, a middle-aged driver in a "brand new" vehicle found it would not decelerate below 30mph (50kmph). He retained steering control, and avoided crashing until police vehicles "boxed in" his vehicle and helped him exit into a police van (most have sliding side doors) from the moving vehicle. The police then "carried out a controlled halt" on the unmanned vehicle, stopping it from driving away with the van's brakes until a roadside assistance technician arrived 3 hours later and managed to shut it down. "[W]hen the [technician] got to me [...] later, he plugged in the car to do a diagnostic check and there was pages of faults," said the "kidnapped" driver from Glasgow. "He said he had never seen anything like it and decided he was not willing to turn the engine on to see what was wrong."

By inference, the vehicle did not have a mechanical brake ("hand brake": English; "parking brake": American), which should have been able to keep the vehicle halted regardless of the motor's actions (even if a "clutch" did get burned out). From the only time I've been inside an electric car, I can't say if that is normal; it's certainly something I'll look for if I ever rent another. Had the failure happened at 10 a.m. in the morning, not 10 p.m. in the evening, the body count could have been ... substantial.

A dumb question, stemming from my only use of an electric car: do they have a weight sensor under the driver's seat that locks-out the main motor unless there is (say) 30kg in the driver's seat? Most have some such sensors -- they trigger the "seatbelt not fastened" alarm or silence it for empty seats -- but whether they can override the drive system ... ?

Europe's biggest telecoms companies have called on the EU to compel Big Tech to pay a "fair" contribution for using their networks, the latest stage in a battle for payments that has pitched the sector against companies such as Netflix and Google. From a report: Technology companies that "benefit most" from telecoms infrastructure and drive traffic growth should contribute more to costs, according to the chief executives of 20 groups including BT, Deutsche Telekom and Telefonica, who signed an open letter seen by the Financial Times. It will be sent to the European Commission and members of the European parliament. "Future investments are under serious pressure and regulatory action is needed to secure them," they warned. "A fair and proportionate contribution from the largest traffic generators towards the costs of network infrastructure should form the basis of a new approach."

They added that regulators need to take action to help secure future investment, with telecoms groups having to spend billions to support the rollout of 5G and upgrade to full-fibre networks. Signatories included Timotheus Hottges at Deutsche Telekom, Christel Heydemann at Orange, Jose Mara Alvarez-Pallete at Telefonica and Pietro Labriola at Telecom Italia. It was also supported by outgoing BT chief executive Philip Jansen, his successor Allison Kirkby, who is currently chief executive at Telia, as well as Vodafone's chief executive Margherita Della Valle. They suggested that a payment mechanism might only make demands on "the very largest traffic generators" with a focus on "accountability and transparency on contributions...so that operators invest directly into Europe's digital infrastructure."

Bill Toulas writes via BleepingComputer: Malicious advertisements are now being injected into Microsoft's AI-powered Bing Chat responses, promoting fake download sites that distribute malware. [...] Malicious ads spotted by Malwarebytes are pretending to be download sites for the popular 'Advanced IP Scanner' utility, which has been previously used by RomCom RAT and Somnia ransomware operators.

The researchers found that when you asked Bing Chat how to download Advanced IP Scanner, it would display a link to download it in the chat. However, when you hover over an underlined link in a chat, Bing Chat may show an advertisement first, followed by the legitimate download link. In this case, the sponsored link was a malvertisements pushing malware. [...] Unfortunately, Malwarebytes could not find the final payload for this malware campaign, so it is unclear what malware is ultimately being installed. However, in similar campaigns, threat actors commonly distribute information-stealing malware or remote access trojans that allow them to breach other accounts or corporate networks.

An anonymous reader quotes a report from Ars Technica: Hackers backed by the Chinese government are planting malware into routers that provides long-lasting and undetectable backdoor access to the networks of multinational companies in the US and Japan, governments in both countries said Wednesday. The hacking group, tracked under names including BlackTech, Palmerworm, Temp.Overboard, Circuit Panda, and Radio Panda, has been operating since at least 2010, a joint advisory published by government entities in the US and Japan reported. The group has a history of targeting public organizations and private companies in the US and East Asia. The threat actor is somehow gaining administrator credentials to network devices used by subsidiaries and using that control to install malicious firmware that can be triggered with "magic packets" to perform specific tasks.

The hackers then use control of those devices to infiltrate networks of companies that have trusted relationships with the breached subsidiaries. "Specifically, upon gaining an initial foothold into a target network and gaining administrator access to network edge devices, BlackTech cyber actors often modify the firmware to hide their activity across the edge devices to further maintain persistence in the network," officials wrote in Wednesday's advisory. "To extend their foothold across an organization, BlackTech actors target branch routers -- typically smaller appliances used at remote branch offices to connect to a corporate headquarters -- and then abuse the trusted relationship of the branch routers within the corporate network being targeted. BlackTech actors then use the compromised public-facing branch routers as part of their infrastructure for proxying traffic, blending in with corporate network traffic, and pivoting to other victims on the same corporate network."

Most of Wednesday's advisory referred to routers sold by Cisco. In an advisory of its own, Cisco said the threat actors are compromising the devices after acquiring administrative credentials and that there's no indication they are exploiting vulnerabilities. Cisco also said that the hacker's ability to install malicious firmware exists only for older company products. Newer ones are equipped with secure boot capabilities that prevent them from running unauthorized firmware, the company said. "It would be trivial for the BlackTech actors to modify values in their backdoors that would render specific signatures of this router backdoor obsolete," the advisory stated. "For more robust detection, network defenders should monitor network devices for unauthorized downloads of bootloaders and firmware images and reboots. Network defenders should also monitor for unusual traffic destined to the router, including SSH."

To detect and mitigate this threat, the advisory recommends administrators disable outbound connections on virtual teletype (VTY) lines, monitor inbound and outbound connections, block unauthorized outbound connections, restrict administration service access, upgrade to secure boot-capable devices, change compromised passwords, review network device logs, and monitor firmware changes for unauthorized alterations.

Ars Technica notes: "The advisory didn't provide any indicators of compromise that admins can use to determine if they have been targeted or infected."

Hollywood studios are expected to retain the right to train artificial-intelligence models based on writers' work under the terms of a tentative labor agreement between the two sides, WSJ reported, citing people familiar with the situation. From the report: The writers would also walk away with an important win, a guarantee that they will receive credit and compensation for work they do on scripts, even if studios partially rely on AI tools, one of the people said. That provision had been in an earlier offer from the Alliance of Motion Picture and Television Producers, the group representing studios, streamers and networks. The Writers Guild of America said Sunday it had reached a tentative agreement with the AMPTP to end a nearly five-month strike. Neither side has released the details of the agreement. The WGA said it plans to release the terms once its leadership votes on the deal, which could happen as soon as Tuesday.

The two sides have battled over issues ranging from wage increases to whether writers' rooms should have minimum staffing requirements. The use of generative AI by studios became a major issue, as advanced versions of the technology -- such as OpenAI's ChatGPT -- were released for public use over the past year. AI bots, which provide sophisticated, humanlike responses to user questions, are "trained" on large amounts of data. Entertainment executives didn't want to relinquish the right to train their own AI tools based on TV and movie scripts, since their understanding is that AI tech platforms already are training their own models on such materials, people familiar with the matter said.

An anonymous reader writes: Chase UK, JPMorgan's UK Bank, has told its customers that it will not carry out transactions related to crypto assets.

The Financial Times writes:

Recent breaches of MGM's casino systems "were probably carried out by teens and young adults who have allied themselves with one of the world's most notorious ransomware gangs," writes the Washington Post's technology reporter.

Their alliance with the "Scattered Spider" group is described as "part of a trend that has alarmed security experts and defenders of corporate computer networks." The group is said to be "very active in the past two years, targeting large companies via stolen employee credentials and tricks such as convincing tech support employees that they have been accidentally locked out of their computers and need a new password." They moved from cryptocurrency thefts to targeting businesses that provide third-party business functions such as help desks and call center staffing, allowing them to infiltrate networks of many customers. And they extorted Western Digital and other technology firms after stealing internal data before heading for the jackpots in Las Vegas. But their willingness to deploy crippling ransomware while demanding money is a major escalation, as is their choice of a business partner: ALPHV, a hacking group whose affiliates include members of the former Russian powerhouses BlackMatter and DarkSide, the groups responsible for the Colonial Pipeline hack that awoke Washington to the national security risk of ransomware. ALPHV provided the BlackCat ransomware that the young hackers installed in the casinos' systems...

[According to new research presented Friday at the LABScon security conference] they came together through crimes enabled by SIM-swapping, which usually involves convincing phone company employees to hand over control of someone else's phone number. Because of poor security controls around those numbers, such gambits have allowed criminals to amass millions of dollars by beating SMS text-based two-factor authentication on cryptocurrency accounts. The extra money has made alliances possible with criminals who have different skills to bring to the table, including some who had hacked police servers and could send emails from purported officers demanding emergency disclosures of information on phone and internet customers. Worse, the researchers said, they have now attracted recruiters for the Russian gangs who want to combine their business savvy with the techniques and local knowledge of the native English speakers.

Cisco agreed to buy Splunk in a deal valued at about $28 billion, representing its biggest acquisition yet and a massive push into software and artificial intelligence-powered data analysis. From a report: Cisco will pay $157 a share in cash, the companies said in a statement Thursday, or a 31% premium to Splunk's previous closing price on Wednesday. Cisco has been expanding its software and services business in an attempt to rely less on its hallmark networking hardware. The Silicon Valley giant has traditionally generated the bulk of its revenue from equipment that forms the backbone of computer networks, but that's been changing. Last month Cisco outlined the headway it's making in artificial intelligence and security technology.

The deal is also arguably just as much a bet on artificial intelligence as it is on the business of software and data security. Earlier this year, Splunk announced a new line of AI offerings that it said would, among other things, allow companies to detect and respond to anomalies in their data faster. The merger offers the companies' AI products "substantial scale" and more visibility into data, they said in Thursday's statement.

An anonymous reader quotes a report from Ars Technica: A suspicious phone company is on the verge of having all its calls blocked by US-based telcos after being accused of ignoring orders to investigate and block robocalls. One Owl Telecom is a US-based gateway provider that routes phone calls from outside the U.S. to consumer phone companies such as Verizon. "Robocalls on One Owl's network apparently bombarded consumers without their consent with prerecorded messages about fictitious orders," the Federal Communications Commission said yesterday.

On August 1, the FCC sent One Owl a Notification of Suspected Illegal Robocall Traffic (PDF) ordering it to investigate robocall traffic identified by USTelecom's Industry Traceback Group, block all of the identified traffic within 14 days, and "continue to block the identified gateway traffic as well as substantially similar traffic on an ongoing basis." One Owl apparently hasn't taken any of the required steps, the FCC said yesterday. "One Owl never responded, and the [FCC Enforcement] Bureau is not aware of any measures One Owl has taken to comply with the Notice," an FCC order said.

Blocking robocall traffic from companies like One Owl is a bit like playing whack-a-mole. The FCC said it previously took enforcement actions "against two other entities to whom One Owl is closely related: Illum Telecommunication Limited and One Eye LLC. While operating under different corporate names, these entities have shared personnel, IP addresses, customers, and a penchant for disregarding FCC rules." If One Owl doesn't provide an adequate response within 14 days, all phone companies receiving calls from it "will then be required to block and cease accepting all traffic received from One Owl beginning 30 days after release of the Final Determination Order," the FCC said. "One Owl faces a simple choice -- comply or lose access to U.S. communications networks," FCC Enforcement Bureau Chief Loyaan Egal said in a press release.