// live_feed --auto-update
AI & cybersecurity news
The stories that matter in AI, security, breaches and vulnerabilities, collected automatically from trusted sources through the day. Short summaries, with every story linked to its original source. RSS feed
Latest stories
A Single Malicious Webpage Let Researchers Hijack Meta Muse and Access Private Chats
Read at The Hacker News ↗GnuPG Warns Critical GPGME Flaw Could Expose App Data via Malicious OpenPGP Keys
Read at The Hacker News ↗Attackers' AI Agents Found a Zero-Day in Medical-Imaging Software and Gained Root Access
Read at The Hacker News ↗Signal Blocks AI Agents From Reading Private Chats via Screenshots on Android and iOS
Read at The Hacker News ↗Hackers Exploit Critical VMware, Cisco, and Apache Flaws as Espionage Attacks Emerge
Read at The Hacker News ↗CTO of ‘RansomwareHelp’ Indicted for Wire Fraud
Read at Krebs on Security ↗Cops Are Pushing Extremely Corny ‘Grand Theft Auto 6’ Slop
Read at WIRED ↗Google rolls out new search homepage with AI chat history and no search button
Read at Ars Technica ↗Meta reconsiders 2030 net-zero goal amid record emissions from AI data centers
Read at Ars Technica ↗West Virginia Legislature hit with cyber incident affecting IT systems
Read at The Record ↗China's 'Armory Castle' Lays Siege to VMware vCenter Zero-Day
Read at Dark Reading ↗Researcher plants 2m canaries into the AI supply chain. A privacy disaster surfaced.
Read at Ars Technica ↗US companies’ use of Chinese open weight AI draws rebuke from House chairmen
US companies are using Chinese open weight AI models for certain projects. A House investigation is examining the potential dangers of this practice.
Read at The Record ↗Daniel Ek backs AI researcher’s fusion power startup. Here’s what they’re building.
Spotify founder Daniel Ek is investing in Europa Fusion. The startup, founded by an AI researcher, is developing fusion power technology.
Read at TechCrunch ↗Jensen Huang’s ‘Lying’ Alarmed Nvidia Board Ahead of His Resignation
Read at WIRED ↗AI claimed a solution to the 167-year-old Riemann Hypothesis — and a human did, too
A potential solution to the Riemann Hypothesis has been found using AI. A human mathematician has also proposed a solution to the 167-year-old problem.
Read at TechCrunch ↗Container Store restores systems after cyberattack
The Container Store restored its website and systems after a cybersecurity incident. The incident occurred over the weekend, affecting the company's operations temporarily.
Read at The Record ↗Trump threatens ‘tremendous tax’ for moving AI data centers outside the US
President Donald Trump plans to impose a tax on companies building data centers abroad. The US is seeing restrictions on new data center construction in various states and cities.
Read at The Verge ↗In Security, Words Matter as Much as Tools, Maybe More
Cyber leaders are emphasizing clear communication as a key defense capability. Effective language is being used alongside tools like AI and cloud security to protect enterprises.
Read at Dark Reading ↗The most misunderstood graph in AI just got a major update
A nonprofit AI evaluation group has updated its graph tracking AI progress. The update shows some AI capabilities improving exponentially, but struggling with tasks requiring judgment.
Read at MIT Technology Review ↗WhatsApp Patches 3 Security Flaws, Including Android Bug Suspected of Use in Scams
WhatsApp has fixed three security flaws, including an Android bug. The bug, found in WhatsApp for Android, involved incomplete validation of external URLs.
Read at The Hacker News ↗Amazon says Chinese hackers used Cairn AI tool to discover apparent zero-day during attack on university
Chinese hackers used Cairn AI to attack universities. The hackers possibly discovered a zero-day bug during one of the attacks.
Read at The Record ↗What We Missed: Nvidia CEO Scandal Rocks AI Industry
Nvidia CEO Jensen Huang resigned amid scandal. Microsoft experienced a recent cloud outage.
Read at Dark Reading ↗Bazuka Android Malware Uses Fake Screen Overlays to Hide Credential Theft and Device Control
Researchers discovered Bazuka, an Android malware framework. It uses fake screen overlays to steal credentials and control devices.
Read at The Hacker News ↗Anthropic mathematician claims to have solved prestigious Millennium Prize problem with Claude
A mathematician claims to have solved the Riemann hypothesis with AI assistance. The solution, if verified, would be a significant achievement in mathematics.
Read at The Verge ↗NSA resumes normal operations following internal network disruptions
The NSA has resumed normal operations after experiencing internal network disruptions. Some facilities are still under restricted access, with delayed openings and required supervisor check-ins for non-essential employees.
Read at The Record ↗WhatsApp Fixes Flaw Uncovered While Disrupting Scam Campaign
WhatsApp's security team found a vulnerability that could make scam messages appear credible. The issue, tracked as CVE-2026-23871, has been fixed by WhatsApp.
Read at SecurityWeek ↗In Other News: Norwegian Energy Firm Cyberattacks, Supercomputer Hack, New Nightmare Eclipse Exploit
Norway's energy firm faced cyberattacks and a supercomputer was hacked. The EU added a group to its cyber sanctions list and malware targeted personal computers.
Read at SecurityWeek ↗Cyber Director Cairncross out as White House resets cyber leadership
Sean Cairncross is departing the White House. Kirsten Davies has been named acting cyber director, replacing Cairncross.
Read at The Record ↗Dozens of credit unions hit by supplier breach that went unnoticed for six weeks
A data breach affected dozens of credit unions. Over 40 breach notifications were received by Ireland's Data Protection Commission.
Read at The Record ↗Six AI-powered gadgets that could make your home office smarter
Six gadgets use AI to enhance home offices. These include notetakers and webcams that upgrade workspaces.
Read at TechCrunch ↗Industry Reactions to Major Outages and AI Patching Debate: Feedback Friday
Experts discussed the role of AI-accelerated patching in major outages. They also highlighted the need for tested rollback plans and the fragility of interconnected systems.
Read at SecurityWeek ↗Signal is blocking Gemini and Siri AI from reading your screen
Signal is restricting screenshot ability on devices to prevent AI agents from reading user data. The move aims to protect user privacy from AI capabilities built into operating systems.
Read at The Verge ↗VMware vCenter zero-day exploited for months before patch (CVE-2026-123890)
Read at Help Net Security ↗Linux KVM Flaw CVE-2026-99312 Lets Guest VMs Access Host Memory on AMD Hosts
Read at The Hacker News ↗Muse Is Unleashing a World of Pettiness on Instagram
Meta's AI assistant tracks Instagram users who don't follow back. It can also unfollow them, a feature some users are utilizing.
Read at WIRED ↗The Path to AI Autonomy in the SOC
Security leaders are increasing AI authority in security operations. A survey found 36% of CISOs currently allow AI autonomy for low-risk actions, expected to rise to 60% in a year.
Read at The Hacker News ↗SBC Medical Group Data Breach Impacts Over 2 Million
A data breach at SBC Medical Group affected over 2 million individuals. Threat actors accessed customer information from clinics in Japan.
Read at SecurityWeek ↗European Commission Confirms AI-Assisted Cyberattack Disrupted Services
The European Commission experienced a cyberattack that disrupted services. Certain services were taken offline to contain the incident with no evidence of a data breach found.
Read at SecurityWeek ↗SonicWall Warns Gen 6 Firewalls Won't Get Security Patches, Even for Critical Flaws
SonicWall advises replacing Generation 6 firewalls due to security risks. The company will not provide security patches, even for critical flaws, citing increased vulnerability to AI-assisted attacks.
Read at The Hacker News ↗Ruby Patches Ruby2Shell Vulnerability That Could Lead to RCE
Read at SecurityWeek ↗Hackers Exploit Critical Cisco Expressway Flaw, Breach Four Devices and Reach Internal Network
Hackers exploited a critical Cisco Expressway flaw to breach devices and access an internal network. The vulnerability, with a CVSS score of 9.8, was used to target a Russian organization.
Read at The Hacker News ↗265,000 Oyster photocard customers’ data exposed in Transport for London breach
Transport for London reported a data breach affecting 265,000 Oyster photocard customers. The exposed data includes account details of children and seniors with specialized cards.
Read at Help Net Security ↗China-Linked Hackers Exploit Critical VMware vCenter Zero-Day to Gain Admin Access and Steal Data
Hackers linked to China exploited a VMware vCenter zero-day flaw. The vulnerability allows unauthorized admin access and data theft.
Read at The Hacker News ↗N-able Blames Service Disruption on Cyberattack
N-able experienced a service disruption due to a cyberattack. The incident affected certain cloud services and potentially exposed credentials.
Read at SecurityWeek ↗European Commission confirms AI-generated cyberattack disrupted its systems
The European Commission was hit by a cyberattack using a publicly available AI tool. Several of its applications were taken offline as a result of the disruption.
Read at Help Net Security ↗JPMorgan, Visa, Mastercard Cite Third-Party Incidents Amid Data Leak
JPMorgan, Visa, and Mastercard attributed a data leak to third-party incidents. The N0n extortion group leaked data allegedly tied to the companies' customers.
Read at SecurityWeek ↗ROMULUS ClickFix Attacks Deliver RAT That Can Spy on Users and Control PCs
Threat actors are using ROMULUS ClickFix to deliver a new remote access trojan. The malware allows spying on users and controlling PCs, sharing similarities with existing trojans.
Read at The Hacker News ↗This 16-Year-Old Founded an AI Missile Startup. He’s Targeting ‘Cost per Kill’
Henrijs Dandzbergs founded an AI missile startup after dropping out of high school. He aims to create efficient missiles for the US and its allies.
Read at WIRED ↗Your AI red team may be finding the same prompt injection thousands of times
Researchers tested AI agents against prompt injection attacks. The results showed that retrained AI agents were easier to hijack in most tests.
Read at Help Net Security ↗AI agent API security hasn’t caught up with agents already running in production
Many organizations have AI agents in production, but security measures are lacking. A significant number of respondents have not established credential and identity management for their AI agents.
Read at Help Net Security ↗New Unpatched Windows 0-Day FastUnlock Lets Someone Holding a Locked PC Run Programs on It
A new Windows zero-day technique allows access to locked computers. It enables running programs without a login, but only with limited account privileges.
Read at The Hacker News ↗Most CIOs and CISOs expect to contain attacks, though few can reach their gear right away
A survey of CIOs and CISOs found most expect to contain attacks despite difficulties accessing critical network infrastructure. Few organizations can quickly gain secure access to their infrastructure when the primary network is down.
Read at Help Net Security ↗Apple may be developing its own AI agent that runs your Mac
Apple is reportedly developing an AI agent for Mac computers. The agent would compete with existing AI tools from other companies.
Read at TechCrunch ↗Anthropic’s 7 founders cashed out nearly $2B in record-breaking IPO
Anthropic's seven founders sold shares worth nearly $2 billion. The company raised almost $105 billion in its initial public offering.
Read at TechCrunch ↗Introducing harness adapters: adapt your favorite model to your harness to improve performance and cost
Harness adapters are now available to enhance model performance. They allow users to adapt their models to existing harnesses.
Read at Hugging Face ↗Android developer verification will reach every install source, with a side door for power users
Google will extend Android developer verification globally in January 2027. Unverified developers' apps will lose normal install functionality on certified devices.
Read at Help Net Security ↗Qwen Incidents Show Self-Hosted AI Models Need Stronger Steering
Companies using self-hosted AI models face control issues. Recent incidents highlight the need for stronger guidance and direction.
Read at Dark Reading ↗