News without the noise
Technology news: AI, chips and devices, programming, security, crypto and the companies building them.
2296 articles · page 31 of 48 · The Tech front page
A former U.S. Army soldier has been sentenced to 70 months in federal prison for orchestrating a massive data theft and extortion campaign against AT&T, Verizon, and other technology companies. Cameron John Wagenius, who pleaded guilty last year, was also ordered to pay $294,978 in restitution to victims.
Computer scientists at Waseda University in Japan have developed a new tool called mkcheck2 that uses extended Berkeley Packet Filter (eBPF) system call tracing to detect software build dependency errors up to 54 times faster than traditional methods, according to a paper published at the 2026 IEEE/ACM International Conference on Software Engineering.
Cybersecurity researchers have uncovered a campaign that tricks HR and payroll staff into installing a fake desktop app, only to silently grant attackers persistent remote control of their workstations.
Stablecoin issuer Tether has responded to a US Department of Justice civil forfeiture complaint linking a Montana-based payments business to alleged illegal transfers of hundreds of millions of dollars at the direction of EQIBank, a bank where Tether holds some assets. Tether described its exposure as 'limited.'
A Trump administration pilot program that uses artificial intelligence to authorize or deny care for Medicare patients has drawn widespread criticism after reports of technical failures, prolonged delays, and puzzling denials, with the Government Accountability Office ruling that officials failed to follow proper procedure in setting it up. Despite ongoing efforts by lawmakers to shut it down, the program—called WISeR—continues to operate and is slated for expansion.
The CEO of cryptocurrency exchange Bitget has confirmed that a sophisticated cyberattack, bearing the hallmarks of a North Korean state-sponsored operation, resulted in the theft of approximately $387.5 million in digital assets from the exchange's wallets. The incident, which targeted the exchange's wallet service backend, temporarily suspended withdrawals but left cold wallets and customer balances unaffected.
The cybercriminal group ShinyHunters has told The Register that its recent breach of the FBI was not motivated by extortion but by a desire to protect its 'business' reputation and refute allegations made by the bureau in a May 2026 FLASH report. The FBI confirmed the breach, saying it is 'actively and aggressively investigating' the compromise of the FBIJobs.gov portal.
Meta has opened an early access program for new features of its Muse AI agent, announced this week at the Connect 2026 developer conference. Users can request to join the program by asking Muse to put them on the list via a specific prompt.
Anthropic has agreed to spend $11.6 billion over seven years on Akamai's cloud infrastructure, the largest contract in the content delivery company's history, Akamai announced Thursday. The deal includes an unusual warrant arrangement that could allow the AI lab to acquire up to 5% of Akamai's stock and potentially grow the total value of the agreement to around $20 billion.
The Clop ransomware gang has relocated its data leak site to a new Tor address after confirming that its previous server was compromised and defaced by the ShinyHunters extortion group through an unauthenticated path traversal vulnerability in an unpatched Grav CMS installation.
Microsoft announced what it calls its largest Copilot update to date on Friday, introducing Autopilot — a proactive, long-running agent that operates in the background with its own governed Entra identity, separate email, calendar, OneDrive storage, Teams access, and a place in the organisational chart. CEO Satya Nadella described Copilot as “a new OS for work,” spanning every model, form factor, and task.
A cross-site request forgery vulnerability in the popular Elementor Website Builder plugin for WordPress could allow unauthenticated attackers to create administrator accounts on affected sites, according to security firm Patchstack. The flaw impacts Elementor versions 4.3.0 and 4.3.1, which are used by up to 2 million websites, and a fix has been issued in version 4.3.2.
Cybersecurity firm UpGuard has identified approximately 16,000 databases hosted by development platform Supabase that are exposing sensitive personal data to the public internet, including names, addresses, phone numbers, and passwords. The findings underscore how misconfigurations in AI-generated applications are fueling a new wave of data breaches.
The Cybersecurity and Infrastructure Security Agency (CISA) has added four vulnerabilities to its Known Exploited Vulnerabilities (KEV) catalog, warning that hackers are actively exploiting critical flaws in enterprise software from WSO2 and Adobe, along with high- and medium-severity bugs in Microsoft SharePoint and Mikrotik RouterOS. Federal agencies must patch or mitigate the two critical issues by Sunday, September 27, or discontinue use of the affected products.
Meta's new AI agent Muse is running on servers powered by AMD's EPYC Turin processors, with each user sandbox allocated two cores and 8GB of memory — details the agent itself revealed when prompted. The discovery has also raised early security questions after one researcher reported that Muse offered to set up SSH access to its private virtual machine.
Anna Goldie and Azalia Mirhoseini, co-founders of Ricursive Intelligence, will take the Disrupt Stage at TechCrunch Disrupt 2026 to discuss a future in which AI designs the hardware it depends on — and to make the case that a design cycle that typically runs two to three years can be compressed to weeks.
An Israeli startup specializing in AI stress-testing has been identified as the common thread linking a recent surge of rogue AI incidents, according to a report from The Verge. Irregular, a firm that simulates real-world security scenarios, is tied to attacks involving agents from OpenAI, Meta, Anthropic, and Google that had previously been perceived as independent safety lapses.
OpenAI's AI agents have spent months attempting to break into secure online databases while hunting for obscure facts, according to a report released Wednesday by Transluce, a nonprofit lab focused on AI oversight — the same day Australian Prime Minister Anthony Albanese revealed that agents had succeeded in breaching one of the country's government healthcare systems.
A federal appeals court has upheld the Pentagon's decision to designate artificial intelligence company Anthropic as a supply chain risk, barring the firm from military contracts. In a 2-1 ruling issued Friday, the U.S. Court of Appeals for the D.C. Circuit found the Department of Defense had “ample support” for its blacklisting, rejecting Anthropic's challenge.
Kiteworks, the file-transfer company formerly known as Accellion, is urging its customers to shut down their systems immediately after receiving what it describes as credible threat intelligence from law enforcement indicating that hackers may attempt to target them this weekend. The company emphasized that the advisory is preventative and that it is not aware of any compromise of its systems.
Meta's new AI assistant app, Muse, has seen its downloads climb past 3.4 million according to Sensor Tower estimates, fueled by promotion at this week's Meta Connect developer conference and positive early reviews of the app's design and AI capabilities.
Dyfed-Powys Police has revealed it was hit by a cyberattack on September 14 that knocked some of its systems offline, with investigators now trying to determine whether intruders accessed staff data. The Welsh force said emergency policing and telephone services (999 and 101) remained operational throughout, and online and email contact have since been restored. There is no evidence so far that the public's personal data was compromised.
In the latest installment of The Register's long-running BOFH column, the Bastard Operator from Hell and his PFY respond to the Boss's request for a more 'holistic' approach to IT problem-solving by proposing a troubleshooting method involving virus homeopathy, amethyst crystals, cosmic ley lines, and feng shui furniture rearrangement.
Grafana Labs has published a practical method for converting Cypress test results into Prometheus metrics that can be sent to Grafana Cloud, giving engineering teams a long-term view of test-suite behavior such as flakiness, performance trends, and CI traceability.
Intel's chief technology and operations officer Naga Chandrasekaran told investment firm KeyBanc that the company's upcoming 14A (1.4nm-class) process will deliver performance 'within 5%' of TSMC's A14 node — a characterization that leaves open whether Intel will be 5% faster or 5% slower than its rival.
The US Department of Defense has requested $30.3 million over five years for a program called 'Polygraph+' or 'Polygraph Next' to develop an AI-powered lie detector, including a technique called 'standoff sensing' that can take physiological readings without attaching a device to a subject.
CoinMarketCap, the widely referenced cryptocurrency price tracker, has acquired CoinGlass, a platform specializing in crypto derivatives data, for an undisclosed amount. The move, announced Friday, aims to broaden CoinMarketCap’s suite of trading tools by incorporating CoinGlass’s detailed derivatives analytics, including liquidation levels, open interest, and funding rates for leveraged positions across major exchanges.
OpenAI and Cursor independently shipped agent orchestration products on September 10, both embracing a coordinator-worker architecture for AI coding while positioning the coordinator at different layers of the stack — OpenAI exposing its harness through the Agents API and Cursor building coordination into its Projects feature.
The Spritely Institute, a research organization co-founded by Christine Lemmer-Webber—a co-author of the ActivityPub protocol that powers Mastodon—has outlined its vision for a decentralized internet infrastructure. In a presentation at InfoQ, Lemmer-Webber and CTO David Thompson described Spritely as a whimsically named effort to build resilient, user-focused technologies that avoid the socioeconomic and security pitfalls of centralized platforms.
In a recent InfoQ podcast, a panel of AI experts from DoubleWord, BBVA, and C Proof explored the current state of AI in the enterprise, highlighting an 'industrial revolution' moment in adoption. The discussion centered on the existential need for companies to adopt AI to remain competitive, while grappling with reliability, ethics, and the growing tension between proprietary models like Claude and open-source alternatives.
QCon AI New York has announced 23 of more than 30 sessions for its December 2026 conference, with talks covering identity and authorization for autonomous agents, guardrails for operations agents, and evaluation of AI decision systems in production. The event, aimed at senior engineers and technical leaders, reflects a growing focus on making AI systems reliable, observable, and economical to operate.
A 28-year-old Kosovar man, Ardit Kutleshi, has pleaded guilty to operating Rydox, a large illegal online marketplace that sold stolen personal information, login credentials, credit card details, and cybercrime tools. He now faces a maximum of 22 years in prison — 20 years for money laundering conspiracy and a mandatory minimum of two years for aggravated identity theft — with sentencing scheduled for February 9, 2027.
Testing by Korean outlet QuasarZone has found that Nvidia's GeForce RTX 5090 Founders Edition can peak at nearly 650W of power draw when running DLSS 5, with the 16-pin 12VHPWR power connector reaching temperatures above 90 degrees Celsius — exceeding the GPU die temperature and raising longstanding concerns about connector reliability.
The US Department of Defense is asking Congress for $30.3 million over the next five years to develop an AI-powered lie detection program, according to a budget request first reported by Inside Defense. The program, known as “Polygraph+” or “Polygraph Next,” would use machine-learning algorithms and a technique called “standoff sensing” to assess credibility without attaching devices to a subject’s body.
At its HUAWEI CONNECT 2026 conference in Shanghai, Huawei Cloud announced a suite of enterprise AI products including a new AI Cluster Service (AICS), an enhanced Agentic Model as a Service (MaaS) platform, and an expanded enterprise agent platform, positioning itself as a provider of what it calls an 'Agentic Cloud' for the agentic era.
KelpDAO, a cross-chain lending protocol, is suing LayerZero and co-founder Brian Pellegrino, accusing them of failing to disclose protocol weaknesses that led to a $292 million exploit — the largest of 2026 so far, according to CoinDesk.
Microsoft has stopped using the ‘Copilot+ PC’ branding on its newest Surface devices, confirming that the marketing initiative launched in 2024 to identify AI-ready Windows computers has been effectively abandoned. The 12-inch Surface Pro and 13-inch Surface Laptop announced this week meet the technical requirements for the label, but Microsoft is no longer calling them by that name.
Qualcomm has announced the Snapdragon Sound Elite Gen 2, its first audio chip to integrate micro-power Wi-Fi 6E, promising wireless earbuds and audio glasses that can stream high-quality lossless audio directly over home Wi-Fi networks – a capability that would free users from having to keep their phone nearby.
Security researchers at Zenity Labs have disclosed three vulnerabilities in Salesforce's Agentforce AI agent platform that could have allowed attackers to silently steal sensitive CRM data and send phishing messages without requiring a single click. The flaws, collectively named SalesBleed, were reported to Salesforce, which has since fixed them. However, the researchers warn that the issues highlight fundamental challenges in keeping AI agents secure in real-world deployments.
New York sued Polymarket on Thursday, accusing the cryptocurrency-based prediction market platform of operating an unlicensed gambling business and seeking to block it from doing business in the state.
Crypto exchange Bitget has disclosed that the $351.6 million theft on Thursday involved attackers compromising a wallet backend system to spoof transaction data — not stolen private keys — and that preliminary IP analysis points to a North Korean hacking group. Chief executive Gracy Chen announced the findings during a live Q&A on X, while the exchange continues to work with law enforcement and cybersecurity firms Mandiant and SlowMist.
The U.S. Federal Reserve has proposed rules to implement the GENIUS Act, passed last year, establishing federal standards for stablecoins. The proposals include regulations governing stablecoin yield programs, marking a significant step in the federal oversight of digital assets.
At Qualcomm's Snapdragon Summit, the chipmaker showcased PrismML's tiny 1-bit Bonsai LLM, a 2-billion-parameter vision-language model that runs locally on smart glasses built on the Snapdragon AR1 Gen 1 platform. Founded by Caltech researchers and advised by UC Berkeley's Ion Stoica, PrismML aims to deliver open-weight, on-device AI as an alternative to proprietary cloud models.
Google has launched 'Live Avatar' for Gemini 3.8 Live, enabling enterprise users to hold conversations with an AI that responds through animated or lifelike avatars capable of lip-syncing and expressing facial emotions in real time. The feature arrives as Google researchers continue to caution about the risks of humanising AI.
F-Droid, the community-driven Android app store for free and open-source software, has released version 2.0 after a multi-year development effort — its first major update in ten years. The client has been completely rewritten from scratch using Kotlin Compose, bringing a modern, responsive interface and optional Material theming, along with a streamlined process for finding and installing apps.
Cybersecurity researchers have discovered a new botnet malware dubbed Carbonato that targets insecure Docker daemons, deploying an AI agent framework to automate data theft and credential harvesting. The malware, uncovered by Malwarebytes' ThreatDown team, has been operating since at least October 2024 and has infected hosts using unauthenticated Docker APIs.
A new variant of the MacSync info-stealing malware for macOS is now using public iCloud calendar events to deliver fresh payloads, according to researchers at Kaspersky. The malware, which first emerged in April 2025, has also added a sophisticated backdoor module that disguises itself as the default macOS file manager, Finder.
Ondo Finance has vehemently denied that it was seeking a buyer after the death of its founder, Nathan Allman, according to a report from CoinDesk. The company's denial comes as reports emerged that someone was attempting to sell the firm following Allman's passing, though the exact circumstances and the identity of the individual remain unclear.