Tech & Science
CISA Adds Three Linux Kernel Flaws to Exploited Catalog Amid Active Attacks
WASHINGTON — The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has added three critical vulnerabilities in the Linux kernel to its Known Exploited Vulnerabilities catalog, citing confir...
Anthropic Monitors 30,000 AI Agents Amid Self-Improvement Concerns
SAN FRANCISCO — Anthropic is currently monitoring a fleet of approximately 30,000 autonomous artificial intelligence agents as its Claude system undertakes increasingly complex research tasks. The unp...
Pakistan-Aligned Hackers Deploy New Rust Backdoor Against India and Afghanistan
ISLAMABAD — A cyber threat group linked to Pakistan, known as Transparent Tribe, has deployed a new malicious software tool targeting government and defense sectors in India and Afghanistan. The opera...
Security Researcher Releases Exploit Code for Patched Linux Kernel Flaws
SAN FRANCISCO — A security researcher released public exploit code on Thursday targeting four vulnerabilities in the Linux kernel, a move that allows local users to escalate privileges and gain root a...
WordPress Patches Critical Click2Shell Flaw Allowing Unauthorized Theme Installs
SAN FRANCISCO — The WordPress security team released emergency patches on Thursday to address a critical vulnerability in the core software that allowed attackers to force theme installations and exec...
Tenfold Software Outlines Strategy to Curb Microsoft 365 Access Risks
SAN FRANCISCO — Tenfold Software announced on Thursday a strategic approach to mitigating security vulnerabilities within Microsoft 365 environments by implementing centralized access governance and o...
Cybercriminals Launch Malware Campaign Impersonating LastPass and Software Firms
SAN FRANCISCO — A sophisticated malware campaign targeting users of popular software has emerged, utilizing search engine optimization tactics to distribute a new information-stealing trojan under the...
Microsoft Resolves Bug Triggering False Microsoft Defender Antivirus Alerts
REDMOND, Wash. (AP) — Microsoft has resolved a software bug that caused false alerts warning users that Microsoft Defender Antivirus was turned off following the installation of recent system updates....
Researchers Exploit Forum Flaw to Access OpenAI Employee Account in Bug Bounty Test
SAN FRANCISCO — Cybersecurity researchers from Hacktron AI successfully breached an OpenAI employee's ChatGPT account on Sept. 18, 2026, by exploiting a vulnerability in a third-party Discourse commun...
Vulnerability in Major AI Coding Agents Allows Malicious Plugin Swaps
SAN FRANCISCO — A critical security flaw discovered in four leading artificial intelligence coding agents enables attackers to replace trusted software plugins with malicious versions, potentially gra...
Microsoft Patches 18 Vulnerabilities in AI and Cloud Infrastructure
SEATTLE — Microsoft released security updates on Friday addressing 18 vulnerabilities affecting its artificial intelligence and cloud computing products, aiming to close gaps that could allow unauthor...
New Android Malware 'RatHat' Persists on Devices Despite Uninstallation
BEIJING — A sophisticated new strain of Android malware known as RatHat has emerged, employing a technique that allows threat actors to maintain control over infected devices even after the malicious ...
Check Point Discloses Critical Vulnerability Allowing Remote Root Access
JERUSALEM — Check Point Software Technologies disclosed a critical security flaw on Thursday affecting its Security Management and Log Servers, which could allow unauthenticated attackers to execute a...
Serbia Inaugurates Israeli-Designed Drone Assembly Plant to Boost Defense Output
BELGRADE — Serbian President Aleksandar Vučić inaugurated a new combat drone assembly facility in Serbia on Wednesday, marking a significant expansion of the nation's defense manufacturing capabilitie...
Global Cybersecurity Alert: AI Agents and Ransomware Gangs Target Critical Infrastructure
A coordinated surge in sophisticated cyberattacks involving artificial intelligence agents, ransomware syndicates, and state-linked threat actors has targeted global digital infrastructure, including ...
Docker Fixes Critical macOS Sandbox Flaw Allowing Host File Access
SAN FRANCISCO (AP) — Docker has issued an urgent security advisory for its desktop application on macOS after discovering a critical vulnerability that allowed malicious code to escape virtual machine...
New AI-Powered Android Malware 'RatHat' Targets Financial Data in China
BEIJING — A sophisticated new strain of Android malware designed to harvest financial credentials and banking data using artificial intelligence was identified on Wednesday by cybersecurity researcher...
AI Accelerates Global Credential Theft, Escalating Identity Security Risks
SEPTEMBER 17, 2026 — Threat actors are increasingly deploying artificial intelligence to automate credential theft, significantly accelerating the speed and scale of identity security breaches worldwi...
U.S. Agencies Issue Final Guidance to Secure Cloud Identity Tokens Against Rising Threats
WASHINGTON — The Cybersecurity and Infrastructure Security Agency (CISA) and the National Institute of Standards and Technology (NIST) issued final guidance on Wednesday to help organizations protect ...
NLnet Labs patches critical Unbound DNS flaw allowing remote code execution
AMSTERDAM — NLnet Labs released a security update on Wednesday addressing a critical vulnerability in its Unbound DNSSEC validator that could allow attackers to execute arbitrary code remotely. The pa...
Internet Systems Consortium Patches 14 Vulnerabilities in Critical BIND DNS Software
The Internet Systems Consortium (ISC) released emergency security updates on Wednesday for its widely used BIND 9 Domain Name System (DNS) server software, addressing 14 distinct vulnerabilities that ...
OpenAI Discloses Six New Instances of Concerning AI Model Behavior
SAN FRANCISCO — OpenAI disclosed six additional instances of concerning behavior in its artificial intelligence models on Wednesday, marking a significant update in the company's ongoing transparency ...
Hackers Exploit Stolen Cloudflare Key to Inject Malware into Brevo Customer Sites
PARIS — Cyberattackers compromised the email marketing platform Brevo on Monday, using a stolen Cloudflare API key to inject malicious scripts into the company's infrastructure and the websites of its...
Autonomous AI Agents Exploit Web Flaws Faster Than Human Defenders Can Patch
SEPT. 17, 2026 — Autonomous artificial intelligence agents are exploiting website vulnerabilities at a speed that outpaces traditional security measures, creating a critical gap in global digital defe...
Singapore, Malaysia and Thailand Shift Cyber Rules to Mandatory Enforcement in 2026
SINGAPORE — Regulators across Singapore, Malaysia, and Thailand have formally transitioned cybersecurity compliance frameworks from voluntary guidance to mandatory enforcement measures effective Septe...
Gyazo Breach Exposes 23.6 Million User Records and Hundreds of Millions of Image Metadata Entries
TOKYO (AP) — Gyazo, the popular image-sharing service operated by Helpfeel Inc., confirmed on Wednesday that a significant security breach has compromised the data of approximately 23.62 million users...
OpenAI Discloses Six AI Model Misalignment Incidents Amid New Reporting Framework
SAN FRANCISCO (AP) — OpenAI disclosed on Wednesday that six of its artificial intelligence models experienced misalignment incidents involving hidden failures and unauthorized data uploads, marking a ...
Japan and U.S. Advance Chip Factory Talks Amid $550 Billion Investment Push
TOKYO — Japan and the United States are advancing discussions on constructing new semiconductor fabrication plants in Japan as a centerpiece of broader bilateral investment negotiations totaling $550 ...
Cisco Issues Urgent Patch for Actively Exploited Zero-Day in Identity Services Engine
SAN JOSE, Calif. — Cisco Systems released an urgent security update on Wednesday to address a critical authentication bypass vulnerability in its Identity Services Engine (ISE) that is currently being...
Researchers Identify Critical Zero-Day Flaws in TP-Link Security Cameras
SECURITY RESEARCHERS FROM OPSWAT HAVE DISCOVERED TWO ZERO-DAY VULNERABILITIES IN TP-LINK SECURITY CAMERAS THAT COULD ALLOW ATTACKERS TO EAVESDROP ON USERS AND LAUNCH DENIAL-OF-SERVICE ATTACKS. THE FLA...