Linux fréttir

Failure could've triggered a small explosion

NASA has revived a set of thrusters on the nearly 50-year-old Voyager 1 spacecraft after declaring them inoperable over two decades ago. …

An anonymous reader quotes a report from TechCrunch: A lawyer representing Anthropic admitted to using an erroneous citation created by the company's Claude AI chatbot in its ongoing legal battle with music publishers, according to a filing made in a Northern California court on Thursday. Claude hallucinated the citation with "an inaccurate title and inaccurate authors," Anthropic says in the filing, first reported by Bloomberg. Anthropic's lawyers explain that their "manual citation check" did not catch it, nor several other errors that were caused by Claude's hallucinations. Anthropic apologized for the error and called it "an honest citation mistake and not a fabrication of authority." Earlier this week, lawyers representing Universal Music Group and other music publishers accused Anthropic's expert witness -- one of the company's employees, Olivia Chen -- of using Claude to cite fake articles in her testimony. Federal judge, Susan van Keulen, then ordered Anthropic to respond to these allegations. Last week, a California judge slammed a pair of law firms for the undisclosed use of AI after he received a supplemental brief with "numerous false, inaccurate, and misleading legal citations and quotations." The judge imposed $31,000 in sanctions against the law firms and said "no reasonably competent attorney should out-source research and writing" to AI.

Read more of this story at Slashdot.

According to the Wall Street Journal (paywalled), Meta is delaying the release of its largest Llama 4 AI model, known as "Behemoth," over concerns that it may not be enough of an advance on previous models. "It's another indicator that the AI industry's scaling strategy -- 'just make everything bigger' -- could be hitting a wall," notes Axios. From the report: The Journal says that Behemoth is now expected to be released in the fall or even later. It was originally scheduled to coincide with Meta's Llamacon event last month, then later postponed till June. It's also possible the company could speed up a more limited Behemoth release.

Read more of this story at Slashdot.

AI footnote fail triggers legal palmface in music copyright spat

An attorney defending AI firm Anthropic in a copyright case brought by music publishers apologized to the court on Thursday for citation errors that slipped into a filing after using the biz's own AI tool, Claude, to format references.…

Longtime Slashdot reader mprindle writes: Nextcloud has been in an ongoing battle with Google over the tech giant revoking the All Files permission from the Nextcloud Android App, which prevents users from managing their files on their server. After a blog post and several tech sites reported on the issue, "Google reached out to us [Nexcloud] and offered to restore the permission, which will give users back the functionality that was lost." Nextcloud is working on an app update and hopes to have it pushed out within a week.

Read more of this story at Slashdot.

An anonymous reader quotes a report from Ars Technica: On Thursday, Telegram announced it had removed two huge black markets estimated to have generated more than $35 billion since 2021 by serving cybercriminals and scammers. Blockchain research firm Elliptic told Reuters that the Chinese-language markets Xinbi Guarantee and Huione Guarantee together were far more lucrative than Silk Road, an illegal drug marketplace that the FBI notoriously seized in 2013, which was valued at about $3.4 billion. Both markets were forced offline on Tuesday, Elliptic reported, and already, Huione Guarantee has confirmed that its market will cease to operate entirely due to the Telegram removal. The disruption of both markets will be "a big blow for online fraudsters," Elliptic confirmed, cutting them off from a dependable source for "stolen data, money laundering services, and telecoms infrastructure." [...] Elliptic reported that Telegram connected black markets with an audience of a billion users, noting that Telegram tried to remove several Huione Guarantee channels earlier this year, but "the marketplace was ready" with backups and remained online until this week. Wired suggested that Huione Guarantee "operated in plain sight" on Telegram for years. But Telegram suggested it just discovered it. Huione Guarantee is a subsidiary of Huione Group, which was recently sanctioned by the U.S. Treasury for supporting "criminal syndicates who have stolen billions of dollars from Americans." According to Reuters, that included allegedly laundering "at least $37 million in crypto from cyber heists by North Korea and $36 million of crypto from so-called 'pig butchering' scams."

Read more of this story at Slashdot.

Every shipment you make, every FLOP you generate, Uncle Sam will be watching you

Proposed legislation gaining steam in Congress this week would require high-end GPUs and AI chips to include location-tracking safeguards to ensure US-designed components don't end up in nations against Uncle Sam's wishes, with exporters on the hook for compliance.…

Surely Trump's anti-immigrant rhetoric has nothing to do with it

US immigration officials say they are winning the war on H-1B fraud – and say they've got the numbers to prove it.…

Uber's autonomous vehicle chief Andrew Macdonald predicted this week that the company will employ more human drivers in a decade despite aggressively expanding robotaxi operations. Speaking at the Financial Times' Future of the Car conference, Macdonald outlined a "hybrid marketplace" where autonomous vehicles dominate city centers while human drivers serve areas beyond robotaxi coverage, handle airport runs, and respond during extreme weather events. "I am almost certain that there will be more Uber drivers in 10 years, not less, because I think the world will move from individual car ownership to mobility as a service," Macdonald said. The ride-hailing giant has struck partnerships with Waymo, Volkswagen, Wayve, WeRide, and Pony AI. Robotaxis are already operational in Austin and Phoenix, with CEO Dara Khosrowshahi claiming Waymo vehicles in Austin are busier than "99%" of human drivers.

Read more of this story at Slashdot.

School districts across the United States were woefully unprepared for ChatGPT's impact on education, according to thousands of pages of public records obtained by 404 Media. Documents from early 2023, the publication reports, show a "total crapshoot" in responses, with some state education departments admitting they hadn't considered ChatGPT's implications while others hired pro-AI consultants to train educators. In California, when principals sought guidance, state officials responded that "unfortunately, the topic of ChatGPT has not come up in our circles." One California official admitted, "I have never heard of ChatGPT prior to your email." Meanwhile, Louisiana's education department circulated presentations suggesting AI "is like giving a computer a brain" and warning that "going back to writing essays - only in class - can hurt struggling learners." Some administrators accepted the technology enthusiastically, with one Idaho curriculum head calling ChatGPT "AMAZING" and comparing resistance to early reactions against spell-check.

Read more of this story at Slashdot.

Scientists have successfully treated a 9.5-month-old boy with an ultra-rare genetic disorder using the world's first personalized gene-editing therapy. The patient, identified as KJ, has CPS1 deficiency -- a condition affecting just one in 1.3 million babies that prevents proper ammonia processing and is often fatal. The breakthrough treatment, detailed in the New England Journal of Medicine, uses base editing technology to correct KJ's specific DNA mutation. The therapy delivers CRISPR components wrapped in fatty lipid molecules that protect them in the bloodstream until they reach liver cells, where they make the precise edit needed. After three infusions, KJ now eats normal amounts of protein and has maintained stable ammonia levels even through viral illnesses that would typically cause dangerous spikes. His weight has increased from the 7th to 40th percentile. Dr. Peter Marks, former FDA official, called the approach "one of the most potentially transformational technologies" because it could be rapidly adapted for thousands of other rare genetic diseases without lengthy development cycles.

Read more of this story at Slashdot.

DragonForce-riding ransomware ring also has 'shiny object syndrome' so will likely move on to another sector soon

Interview The same miscreants behind recent cyberattacks on British retailers are now trying to dig their claws into major American retailers' IT environments – and in some cases possibly even deploying ransomware, according to Google.…

Apple has implemented conspicuous warning labels featuring red exclamation marks on EU App Store listings that use external payment systems. The company's new tactic targets apps like Instacar, a popular Hungarian vehicle valuation tool with thousands of positive reviews, displaying ominous warnings that the app "does not support the App Store's private and secure payment system." The associated support page cautions users that external payments require providing personal information directly to developers and third parties "based on their privacy and security controls." The move also follows the Epic vs Apple ruling that prohibits Apple from interfering with developers linking to alternative payment systems.

Read more of this story at Slashdot.

One expert tells us: 'It is the most unique breach disclosure I've ever seen'

Coinbase says some of its overseas support staff were paid off to steal information on behalf of cybercriminals, and the company is now being extorted for $20 million.…

Google has overtaken IBM to become the leader in generative AI-related patents and also leads in the emerging area of agentic AI, according to data from IFI Claims. Axios: In the patents-for-agents U.S. rankings, Google and Nvidia top the list, followed by IBM, Intel and Microsoft, according to an analysis released Thursday. Globally, Google and Nvidia also led the agentic patents list, but three Chinese universities also make the top 10, highlighting China's place as the chief U.S. rival in the field. In global rankings for generative AI, Google was also the leader -- but six of the top 10 global spots were held by Chinese companies or universities. Microsoft was No. 3, with Nvidia and IBM also in the top 10.

Read more of this story at Slashdot.

Sometimes, less information is more

In its latest gambit to reduce the noise of unnecessary security alerts, Socket has acquired Coana, a startup founded in 2022 by researchers from Aarhus University in Denmark that tells users which vulnerabilities they can safely ignore.…

The Federal Trade Commission has postponed enforcement of its consumer-friendly "click to cancel" rule from May 14 to July 14, giving businesses two additional months to comply. The regulation requires companies to make subscription cancellations as straightforward as the sign-up process, prohibiting practices like forcing customers who subscribed online to navigate through chatbots or call centers to cancel. The rule, established under former Democratic Chair Lina Khan, unsurprisingly has garnered support from consumer advocates while facing legal opposition from industry groups. A coalition including the U.S. Chamber of Commerce and organizations representing major telecom and media companies -- Charter Communications, Comcast, Disney Entertainment, and Warner Bros. Discovery -- has sued to block implementation, claiming the agency exceeded its authority.

Read more of this story at Slashdot.

Cryptocurrency exchange Coinbase said Thursday it is offering a $20 million reward for information leading to the arrest and conviction of criminals who attempted to extort the company for the same amount after stealing customer data. The criminals bribed customer support agents in overseas markets to access records containing addresses, phone numbers, government IDs, and partial bank and Social Security details of more than 80,000 customers. "It sucks but when we see a problem like this we want to own it and make it right," Coinbase Chief Security Officer Philip Martin told Fortune. The company will reimburse customers who fell victim to subsequent social engineering scams. No login credentials or wallet access were compromised in the breach. The extortionists had threatened to publish the stolen information unless paid $20 million in Bitcoin.

Read more of this story at Slashdot.

Devs told to swap raw results for LLM-generated summaries as August shutdown looms

Microsoft is retiring Bing Search APIs on August 11, directing customers toward AI products as an alternative.…

Lessons learned from last year's security snafu

interview Being the chief information security officer at Snowflake is never an easy job, but last spring it was especially challenging.…