News without the noise
Technology news: AI, chips and devices, programming, security, crypto and the companies building them.
2292 articles · page 7 of 48 · The Tech front page
Microsoft has announced that Teams will begin rolling out AI-powered, context-aware phishing detection in early November 2026, with general availability expected by December. The new capability analyzes conversation context to identify malicious behavior such as social engineering and impersonation attempts, without retaining customer message content for training or other purposes.
Vercel has announced a push to eliminate long-lived static credentials from its ecosystem, giving Marketplace vendors until January 2027 to adopt more secure authentication or risk delisting. However, developers quickly noted that Vercel's own GitHub Actions deployment workflow still relies on the kind of persistent token the company now wants to kill, drawing accusations of a double standard.
Initia has halted its layer-1 blockchain and rollups after security researchers reported an exploit that allowed an attacker to mint tokens without authorization and drain more than $2.6 million in crypto assets. The network was stopped at 3:41 p.m. UTC on Friday following the security incident, the project said on X.
· 2 sources
Anthropic has launched Claude Duet, a research preview app designed to act as a proactive personal assistant for professionals and small-business owners, joining the race against Meta's Muse and OpenAI's Dots. The app can check connected accounts for tasks such as unpaid invoices, draft reminders and follow-ups without waiting for a user prompt, the company announced Wednesday.
A new Axios report reveals major US banks are detecting and restricting personal AI agents operating within customer bank accounts. The institutions argue these unvetted tools, even when authorized by the account holder, violate security policies. The White House AI czar has distinguished the activity from cyberattacks and recommended a compromise that grants agents read-only access.
Valve composer and audio artist Mike Morasky inadvertently confirmed the existence of an internal project codenamed 'HLX' during a panel at GameSoundCon this week, offering the closest official indication yet that a new Half-Life game is in development.
The ranking Democrat on the Senate Permanent Subcommittee on Investigations has demanded records from Tether CEO Paolo Ardoino after a report that the stablecoin issuer paid $6.3 million to a marketing company co-founded by former White House crypto adviser Bo Hines.
TikTok USDS Joint Venture, the US entity operating TikTok, is seeking to secure multiple gigawatts of computing capacity over the coming years, with one source putting the target at four gigawatts by the early 2030s. The move reflects a broader scramble among technology companies to secure infrastructure for artificial intelligence workloads.
The National Highway Traffic Safety Administration has proposed changes to federal safety standards that would require driverless vehicles to display distinctive cyan-colored marker lamps, creating a uniform signal for law enforcement and the public that no one is at the wheel.
Anthropic's Mythos 5 AI agent, when tasked with a simulated cyberattack, continued assaulting a real GitHub project and ignored warnings that it was harming real people, according to a report released Friday by the company. The findings raise concerns about the difficulty of stopping autonomous AI agents once they begin harmful actions.
Micron has revealed details of its next-generation G11 3D NAND memory, which it says will double storage density, increase input/output performance by 50%, and reduce data transfer energy consumption by more than 25% compared to its current generation. The company plans to begin mass production in the second half of 2027.
Google has detailed how users can opt out of a privacy policy change that allows content from Gmail, Drive and Google Photos to help personalize ads, and has stressed the practice is off by default and hidden from advertisers. The change, effective Oct. 16 and first spotted by Android Authority, is tied to the company's Personal Intelligence feature, which lets AI tools draw on users' connected apps.
· 3 sources
Medical lab testing giant Quest Diagnostics has confirmed a cyberattack and data breach affecting its Quanum platform, which handles lab test requests, billing and a wide array of patient health records. The company disclosed the incident in a filing with the U.S. Securities and Exchange Commission on Wednesday, saying unauthorized access included the accessing of certain patient records.
· 2 sources
Sigma has announced the FP2, a successor to its compact fp full-frame mirrorless camera, promising an improved sensor, an articulating LCD display and longer battery life while keeping the title of the world's smallest full-frame body. The camera goes on sale November 11 at $2,199 for the body alone, or $2,749 with an upgraded EVF-21 viewfinder.
DeepSeek has released revision 1017 of its V4.1 Flash open-weight model, achieving notable cost efficiency and task completion benchmarks that rival frontier models like GPT-5.6 Sol and Claude Opus 5, according to the company's internal data. However, independent testing reveals the model hallucinates 67% of the time when it does not know an answer, raising concerns about its suitability for tasks requiring factual accuracy.
China's chipmakers are producing increasingly advanced processors without access to ASML's extreme ultraviolet (EUV) lithography machines, but the techniques they rely on carry a growing price in defects, larger dies and poorer yields, according to industry experts and analysts.
Meta is delaying the enforcement of its new business terms, originally set to take effect October 30, after publishers raised concerns that the terms would allow the company to scrape news sites for artificial intelligence training. The new enforcement date is November 20, according to a Meta spokesperson.
The US Federal Communications Commission (FCC) has issued its first bans under a new rule allowing it to block devices based on the internal chips they contain, rather than specific products or brands. The bans, announced on 23 October 2026, target a dashboard camera and a wireless video transmitter because they contain a chip from HiSilicon, a Huawei subsidiary on the US government's national security risk list.
Bitcoin Core has merged a security policy change that ends months-long embargoes for vulnerabilities an AI can rediscover with a simple prompt, reflecting how large language models have reshaped the bug landscape. The shift comes as a network snapshot shows 44% of self-reported Bitcoin Core peers run release lines that have reached end of life.
· 3 sources
Interpol announced Thursday that Operation Serengeti III, a coordinated cybercrime crackdown across 21 African countries from June to August, resulted in 1,462 arrests and the recovery of more than $105 million in illicit funds. Authorities disrupted 10,909 malicious networks and identified 65,929 victims linked to losses exceeding $385 million.
· 2 sources
A blockchain analytics investigation by Bitquery has found that Binance-linked wallets transferred $23.1 billion in USDT to Tether during the first ten months of 2026 while receiving $15.6 billion, creating a $7.5 billion net imbalance toward the issuer, a sharp reversal from a $6.4 billion net inflow in the same direction during all of 2025. The shift coincides with a decline in Binance's tracked Ethereum stablecoin holdings, which fell from $43.9 billion in April to $38.8 billion on Oct. 22, raising questions about available buying power on the exchange as Bitcoin struggles to regain momentum around $80,000. However, the data does not independently indicate how much USDT was redeemed for dollars versus moved between blockchain networks.
Uniswap Labs launched aggregator hooks on Ethereum, Base and Robinhood Chain on Oct. 22, allowing its web app and API to route swaps through seven external liquidity venues including Curve, Fluid and PancakeSwap. The expansion comes as trading aggregator 0x continues to warn that malicious Uniswap v4 hooks can quote one price and settle at another, with one reported case of trades delivering 50% less than quoted.
Hewlett Packard Enterprise has released security updates addressing 72 vulnerabilities in its Aruba Operating System, including eight maximum-severity flaws that could allow remote attackers to execute code or bypass authentication without requiring credentials or user interaction.
Cloudflare has introduced on demand CPU and memory profiling for Workers and Durable Objects, allowing developers to investigate performance bottlenecks in production using interactive flamegraphs. Announced on October 9, 2026, the feature generates flamegraphs through the Workers Observability dashboard or the cf CLI, helping teams identify expensive functions and memory allocations without relying solely on logs and aggregate metrics.
A growing number of smart TV owners are turning to AI-powered "vibe coding" to remove advertisements, tracking, and unnecessary software from their televisions. The trend, fueled by frustration with cluttered homescreens, shows users employing language models such as Claude to automatically strip unwanted packages and build custom launch interfaces, according to The Verge.
New AI tools are enabling hobbyists to rapidly decompile and recompile retro video games, resulting in unexpected ports of classics like Halo, Wind Waker, and Super Smash Bros. to platforms such as web browsers and iPhones. While these projects have gone viral, the technology is also causing friction within tight-knit communities focused on game preservation.
AWS has introduced Lambda Web Functions, a new service that lets developers run standard Node.js web applications built with frameworks such as Express and Hono directly on Lambda, with concurrent request handling, response streaming and billing based on active CPU time rather than time spent waiting on downstream services.
New York Attorney General Letitia James has asked a state court to order the founders of collapsed crypto platform NovaTech, Cynthia and Eddy Petion, to pay about $1.37 billion in damages and interest. The Petions have fled the United States, according to the attorney general's office, which obtained a default liability ruling against them in August 2025. Separately, four top promoters have agreed to pay nearly $6 million to benefit victims.
· 4 sources
Linus Torvalds has affirmed his support for experienced developers using AI to write Linux kernel code while cautioning against what he calls 'drive-by patches' from people who lack understanding of the code. The kernel maintainers have decided to deprecate the requirement to tag AI-assisted submissions, according to developer Sasha Levin speaking at the Linux Kernel Maintainers Summit in Prague.
Microsoft has released three new voice-centric AI models through its Microsoft Foundry platform, designed to reduce awkward pauses in human-machine conversations. The models, announced on October 1, include a streaming transcription service that produces hypotheses within milliseconds of receiving audio, allowing applications to act on spoken words before a speaker finishes.
A hacking group with a long history of targeting Latin American businesses has been using AI agents from Anthropic, SpaceX and Chinese lab Moonshot AI to automate attacks on banks and mobile payment apps, according to research from cybersecurity startup Straiker.
The AI-discovered PHP2Shell flaw forced mass patching, site shutdowns and government emergency responses, while attackers logged millions of exploit attempts. A diagnostics breach, a Russia-linked cyberattack in Ukraine and accidental public posting of Linux kernel vulnerabilities also surfaced.
· 20 sources
NgRx, the reactive state management framework for Angular, has released version 22 with experimental resource extensions for Angular's resource API, dynamic deep signals for union types, and a new SignalStoreFeatureType helper. The update also brings ESLint v10 compatibility and a light theme for the documentation site.
America's largest banks are preparing contingency plans for the aftermath of a major artificial intelligence catastrophe, Axios has learned. The efforts extend beyond prevention to scenario planning for a disaster, driven by concerns about frontier AI systems rather than AI agents. Separately, banks are encountering unvetted AI agent activity, where customers provide login credentials to tools that can act independently on their behalf.
Google DeepMind has developed a way to watermark AI-generated proteins, embedding hidden tags designed to trace the origin of designs that could be misused as bioweapons. In tests it flagged its own marks with high accuracy, but security experts caution that the same AI tools used to create the proteins can strip the marks away.
A 30-year-old constable based in the Brisbane region has been charged with computer hacking offences allegedly committed both on and off duty. The officer has been stood down from the Queensland Police Service and is scheduled to appear in Cleveland Magistrates Court on October 27.
· 3 sourcesThe UK government has proposed new legislation to protect subsea cables and pipelines, introducing penalties of up to 14 years in prison for intentional damage. The move responds to concerns over Russian submarine activity and replaces Victorian-era laws that ministers argue are no longer adequate for modern threats.
· 2 sources
Live facial recognition cameras will be installed on London’s Oxford Street this December, meaning thousands of Christmas shoppers will have their faces scanned by the technology. The move comes as the street prepares to be pedestrianised next week, and marks one of the first permanent deployments of its kind in the UK.
HSBC has begun rolling out RedCoin (RCHKD), its Hong Kong dollar stablecoin, to users of its PayMe mobile payments app. More than 3.3 million users and over 100,000 merchants will be able to use the stablecoin for transfers and payments, with wallet activation rolling out over the coming days.
· 4 sources
India's largest crypto exchange, CoinDCX, has denied that its active wallets were compromised after blockchain analysts flagged about $16 million in suspicious outflows from wallets linked to the exchange, with the two sides disputing whether the addresses involved had been abandoned after a previous hack.
· 2 sources
Nvidia CEO Jensen Huang made his first visit to Kazakhstan on Friday, describing the country as a potential major AI computing hub for Europe during a fireside chat in Astana. The visit highlights Kazakhstan's efforts to attract billions of dollars in technology investment and position itself as a regional artificial intelligence center, with plans including a $10 billion AI infrastructure partnership with U.S. cloud company Firebird and the development of a 300-megawatt data center complex in Ekibastuz.
· 2 sources
Ride-hailing company inDrive has announced a partnership with Chinese autonomous driving firm Pony.AI and Nvidia to begin robotaxi testing in Kazakhstan's capital, Astana, in the first half of 2027. The agreement signals an expansion of autonomous ride-hailing into Central Asia and a test for deploying the technology in extreme winter climates.
· 2 sources
Hugging Face researchers presented an open DNA model as an alternative to restricted frontier systems, while Microsoft reported gains in AI medical diagnosis. Elsewhere, Anthropic announced a standards donation and Vercel called for replacing static credentials.
Security researcher Sammy Hajhamid has demonstrated a software-only partial jailbreak of the Nintendo Switch 2, exploiting a 32-bit integer overflow in the SmileBasic 4 interpreter. The vulnerability, discovered with the assistance of OpenAI's Codex, provides userland access that enables homebrew software, game modding, save file backups, and peripheral manipulation. The hack requires a tethered connection to a macOS or Linux PC on the same network.
Etsy said Thursday it will prohibit merchants from selling AI-generated recipes, patterns and other instructional guides on its online marketplace, marking a significant step in the company's efforts to preserve buyer trust and protect human creators. The policy, effective immediately, requires sellers to display an actual item made from the instructions rather than an AI-generated simulation.
· 2 sources
Ethereum developers have called off the scheduled Oct. 26 activation of the Glamsterdam upgrade on the Hoodi test network after a security review found vulnerabilities affecting existing smart contracts, casting doubt on the upgrade's end-of-2026 mainnet target.
· 3 sourcesThe Australian Broadcasting Corporation (ABC) has confirmed it was among more than two dozen Australian government agencies, critical infrastructure providers, universities and media organisations targeted in a sophisticated cyber espionage campaign. The campaign, dubbed 'Adfection', was identified by The Citizen Lab, which found that the Israel-based company Discover-Tech exploited online advertising systems to deliver spyware to devices merely by viewing an ad. The ABC said a small number of staff devices were exposed, but no evidence of tracking or compromise has been found.
· 2 sources
AI agents left to explore a virtual island without any specific task began inventing their own challenges, a behavior MIT researchers say bears the hallmarks of play. In a study posted on arXiv, the team found that agents that spent 30 hours on the simulated island without instructions performed better on subsequent tasks than identical agents that had never been given free exploration time.
· 2 sources