Kalakshar US

Cybersecurity company Dragos has flagged malware that can attack industrial control systems (ICS), tricking them into malicious behavior like turning off the heat and hot water in the middle of winter. TechCrunch reports that’s precisely what the malware, dubbed FrostyGoop, did this January in Lviv, Ukraine, when residents in over 600 apartment buildings lost heat for two days amid freezing temperatures.

Dragos says FrostyGoop is only the ninth known malware designed to target industrial controllers. It’s also the first to specifically set its sights on Modbus, a widely deployed communications protocol invented in 1979. Modbus is frequently used in industrial environments like the one in Ukraine that FrostyGoop attacked in January.

Ukraine’s Cyber Security Situation Center (CSSC), the nation’s government agency tasked with digital safety, shared information about the attack with Dragos after discovering the malware in April of this year, months after the attack. The malicious code, written in Golang (The Go programming language designed by Google), directly interacts with industrial control systems over an open internet port (502).

The attackers likely gained access to Lviv’s industrial network in April 2023. Dragos says they did so by “exploiting an undetermined vulnerability in an externally facing Mikrotik router.” They then installed a remote access tool that voided the need to install the malware locally, which helped it avoid detection.

The attackers downgraded the controller firmware to a version lacking monitoring capabilities, helping to cover their tracks. Instead of trying to take down the systems altogether, the hackers caused the controllers to report inaccurate measurements — resulting in the loss of heat in the middle of a deep freeze.

Dragos has a longstanding policy of neutrality in cyberattacks, preferring to focus on education without assigning blame. However, it noted that the adversaries opened secure connections (using layer two tunneling protocol) to Moscow-based IP addresses.

“I think it’s very much a psychological effort here, facilitated through cyber means when kinetic perhaps here wasn’t the best choice,” Dragos researcher Mark “Magpie” Graham told TechCrunch. Lviv is in the western part of Ukraine, which would be much more difficult for Russia to hit than eastern cities.

Dragos warns that, given how ubiquitous the Modbus protocol is in industrial environments, FrostyGoop could be used to disrupt similar systems worldwide. The security company recommends continuous monitoring, noting that FrostyGoop evaded virus detection, underscoring the need for network monitoring to flag future threats before they strike. Specifically, Dragos advises ICS operators to use the SANS 5 Critical Controls for World-Class OT Cybersecurity, a security framework for operational environments.

This article originally appeared on Engadget at https://www.engadget.com/russia-linked-hackers-cut-heat-to-600-ukrainian-apartment-buildings-in-the-dead-of-winter-researchers-say-171414527.html?src=rss 

Regulators in the US and Europe have laid out the “shared principles” they plan to adhere to in order to “protect competition and consumers” when it comes to artificial intelligence. “Guided by our respective laws, we will work to ensure effective competition and the fair and honest treatment of consumers and businesses,” the Department of Justice, Federal Trade Commission, European Commission and the UK’s Competition and Markets Authority (CMA) said.

“Technological inflection points can introduce new means of competing, catalyzing opportunity, innovation and growth,” the agencies said in a joint statement. “Accordingly, we must work to ensure the public reaps the full benefits of these moments.”

The regulators pinpointed fair dealing (i.e. making sure major players in the sector avoid exclusionary tactics), interoperability and choice as the three principles for protecting competition in the AI space. They based these factors on their experience working in related markets.

The agencies also laid out some potential risks to competition, such as deals between major players in the market. They said that while arrangements between companies in the sector (which are already widespread) may not impact competition in some cases, in others “these partnerships and investments could be used by major firms to undermine or co opt competitive threats and steer market outcomes in their favor at the expense of the public.”

Other risks to competition flagged in the statement include the entrenching or extension of market power in AI-related markets as well as the “concentrated control of key inputs.” The agencies define the latter as a small number of companies potentially having an outsized influence over the AI space due to the control and supply of “specialized chips, substantial compute, data at scale and specialist technical expertise.”

In addition, the CMA, DOJ and FTC say they’ll be on the lookout for threats that AI might pose to consumers. The statement notes that it’s important for consumers to be kept in the loop about how AI factors into the products and services they buy or use. “Firms that deceptively or unfairly use consumer data to train their models can undermine people’s privacy, security, and autonomy,” the statement reads. “Firms that use business customers’ data to train their models could also expose competitively sensitive information.”

These are all fairly generalized statements about the agencies’ common approach to fostering competition in the AI space, but given that they all operate under different laws, it would be difficult for the statement to go into the specifics of how they’ll regulate. At the very least, the statement should serve as a reminder to companies working in the generative AI space that regulators are keeping a close eye on things, even amid rapidly accelerating advancements in the sector.

This article originally appeared on Engadget at https://www.engadget.com/us-and-european-antitrust-regulators-agree-to-do-their-jobs-when-it-comes-to-ai-163820780.html?src=rss 

General Motors is putting the autonomous Cruise Origin shuttle van on ice. The company said that the embattled Cruise, of which GM is the majority owner, will now focus on making the next-gen Chevy Bolt. The automaker discontinued the previous Bolt last year due to a shift away from an older battery system but did not reveal plans for a new model at the time.

According to a letter that GM CEO Mary Barra sent to shareholders, the indefinite delay of the shuttle van “addresses the regulatory uncertainty we faced with the Origin because of its unique design.” Barra added that the per-unit costs of the next-gen Bolt will be much lower, “which will help Cruise optimize its resources.”

GM and Cruise were working on the Origin with Honda. The Origin — which does not have a driver’s seat, steering wheel or pedals — was supposed to debut in Japan in 2026.

In October, the California Department of Motor Vehicles suspended Cruise’s driverless vehicle permits over safety issues. Earlier that month, a pedestrian in San Francisco was dragged 20 feet by a Cruise vehicle and pinned under it after a hit-and-run by another car pushed her into the robotaxi’s path. Cruise later paused all driverless operations before temporarily halting production in November.

According to CNBC, former Cruise CEO Kyle Vogt at one point told staff that hundreds of pre-commercial Origin vehicles had been built. The company has resumed robotaxi operations in Phoenix, Houston and Dallas with human operators on board and is carrying out tests in Dubai. However, it hasn’t recommenced operations in San Francisco. It’s still under investigation for the October incident there.

Shelving the Origin is not a decision that GM and Cruise would have come to lightly. In GM’s second quarter earnings report, the automaker noted that it incurred around $583 million of Cruise restructuring costs. It said these resulted “from Cruise voluntarily pausing its driverless, supervised and manual [autonomous vehicle] operations in the US and the indefinite delay of the Cruise Origin.”

On the plus side, resuming work on the Bolt (which will presumably use GM’s Ultium battery tech the next time around) could be a boon for GM’s bottom line. As of 2023, the Bolt EV and EUV accounted for most of GM’s electric vehicle sales. It planned to make around 70,000 of them last year before ceasing production.

This article originally appeared on Engadget at https://www.engadget.com/gm-shelves-the-autonomous-cruise-origin-shuttle-van-144256801.html?src=rss 

Meta’s AI-powered assistant have been accessible on the Ray-Ban smart glasses for quite some time, but the company will only start rolling it out to its Quest headsets next month. The assistant will still be in experimental mode, however, and it’s availability will be limited to users in the US and Canada. Meta has revealed the update alongside its announcements for the Llama 3.1 and the new Meta AI capabilities. 

Users who get access to the assistant in August will be able to put its hands-free controls to the test. The company said Meta AI is replacing the current technology used for Voice Commands on Quest, so it will be the one controlling the headset whenever people use voice for navigation and the one answering their questions if they ask for information. They can ask the assistant for restaurant recommendations for an upcoming trip, as an example, or ask it for the weather those days, as well as suggestions on how to dress for it. 

They will also be able to use the “Meta AI with Vision” feature, which will let them ask the assistant for information on what they’re seeing, while using Passthrough on the Quest. Passthrough lets users see their environment through a video feed while watching or doing something else on their headsets. A user can, for instance, ask the assistant to look at what’s inside the fridge and suggest what they can cook, or ask for tips on what kind of top would go with a skirt they’re holding up, all while watching a YouTube video.

This article originally appeared on Engadget at https://www.engadget.com/metas-ai-assistant-is-coming-to-quest-headsets-in-the-us-and-canada-150033530.html?src=rss 

Meta on Tuesday announced the release of Llama 3.1, the latest version of its large language model that the company claims now rivals competitors from OpenAI and Anthropic. The new model comes just three months after Meta launched Llama 3 by integrating it into Meta AI, a chatbot that now lives in Facebook, Messenger, Instagram and WhatsApp and also powers the company’s smart glasses. In the interim, OpenAI and Anthropic already released new versions of their own AI models, a sign that Silicon Valley’s AI arms race isn’t slowing down any time soon.

Meta said that the new model, called Llama 3.1 405B, is the first openly available model that can compete available rivals in general knowledge, math skills and translating across multiple languages. The model was trained on more than 16,000 NVIDIA H100 GPUs, currently the fastest available chips that cost roughly $25,000 each, and can beat rivals on over 150 benchmarks, Meta claimed.

The “405B” stands for 405 billion parameters, which are internal variables that an AI model uses to reason and make decisions. The higher the number of parameters an AI model has, the smarter we perceive it to be. OpenAI’s GPT-4 model, by comparison, is reportedly has roughly 1.5 trillion parameters, although the company has not disclosed the number so far. In addition, Meta also released upgraded versions of existing Llama models that contain 70 billion and 8 billion parameters each, claiming that the newer versions had stronger reasoning abilities among other things.

Developers can download Llama 3.1 from its official website, while regular users can play with it through Meta AI in WhatsApp or on meta.ai, the company’s website for its chatbot. “Llama 405B’s improved reasoning capabilities make it possible for Meta AI to understand and answer your more complex questions, especially on the topics of math and coding,” Meta’s blog post states. “You can get help on your math homework with step-by-step explanations and feedback, write code faster with debugging support and optimization.” (Editor’s note: Engadget will pit Llama 3.1 against the New York Times Spelling Bee and report back to you). For now, Meta AI on Facebook, Messenger and Instagram are still restricted to the smaller version of Llama 3.1 that uses 70 billion parameters.

Unlike OpenAI, Google, Microsoft and Anthropic that keep their AI models proprietary, Meta’s AI models are open source, which means that anyone can modify and use them for free and without sharing personal data with Meta. In a letter published on Tuesday, Meta CEO Mark Zuckerberg argued that an open source approach to AI development will ensure wider access to the technology’s benefits, prevent the concentration of power among a few big companies, and enable safer AI deployment across society. By open sourcing the company’s largest language model to date, Meta aims to make Llama the “industry standard” for anyone to develop AI-powered apps and services with, Zuckerberg wrote.

Open sourcing AI models and adding them to its existing products already used by billions of people could allow Meta to compete more effectively with OpenAI whose ChatGPT and DALL-E chatbots ignited an AI explosion when they launched in 2022. And it could also boost engagement — Meta announced today that users would soon be able to add AI-generated images directly into feeds, stories, comments and messages across Facebook, Messenger, WhatsApp and Instagram.

In his letter, Zuckerberg also criticized Apple and its closed ecosystem, arguing that the iPhone maker’s restrictive and arbitrary policies had constrained what Meta could build on its platforms. “[It’s] clear that Meta and many other companies would be freed up to build much better services for people if we could build the best versions of our producers and competitors were not able to constrain what we could build,” he wrote.

This article originally appeared on Engadget at https://www.engadget.com/llama-31-is-metas-latest-salvo-in-the-battle-for-ai-dominance-150042924.html?src=rss