Stories cluster related articles into a single narrative, linked to the underlying CVEs and affected products. 8,564 stories synthesized.
A high-severity use-after-free vulnerability (CVE-2026-50458) in Windows' Brokering File System allows local attackers to escalate privileges to SYSTEM.
Cobalt introduces an AI-assisted offering designed to deliver continuous offensive security testing and actionable penetration testing results within 24 hours.
A new benchmark by SentinelOne demonstrates that even advanced AI models falter when tasked with sustained, complex malware investigations, highlighting the need for human oversight.
Google is introducing a new AI-powered account recovery method that uses selfie videos to verify user identity when traditional recovery options fail.
A simple operational security mistake by cyber spies led to the exposure of the JadeProx espionage campaign and its novel TriBack malware, impacting organizations across Asia and Latin America.
Google has launched CodeMender, an AI agent designed to autonomously find, validate, and patch software vulnerabilities, aiming to accelerate the remediation lifecycle.
A Russian cyberespionage campaign, tracked as CL-STA-1114 and attributed to the Void Blizzard/LAUNDRY BEAR threat actor, is actively exploiting a zero-day vulnerability in Zimbra Collaboration Suite (ZCS) to steal sensitive user credentials and data.
Attackers are exploiting compromised GitHub repositories and GitHub Actions runners to build distributed attack infrastructure targeting cPanel and WHM servers, leveraging a known authentication bypass vulnerability.
An AI agent named Kimi K3 has identified critical remote code execution vulnerabilities in multiple versions of the widely-used Redis in-memory data store, including a double-free bug and a heap overflow.
A recent paper dissects the ongoing 'Going Dark' debate surrounding end-to-end encryption (E2EE), arguing that broad restrictions would cripple cybersecurity and commerce.
Attackers are actively exploiting CVE-2026-16232, a critical authentication bypass vulnerability in Check Point Security Management servers, granting them full administrative control.
Cisco Talos will present cutting-edge research on AI's role in cybersecurity, including threat actor prompt usage and AI-driven SOC workflows, alongside an analysis of the Warlock ransomware group at Black Hat USA 2026.
The Chaos ransomware group is employing a novel technique using msaRAT to hide its command-and-control (C2) traffic within legitimate browser processes, making detection significantly harder.
CISA has launched ANCHOR-CI, a new framework designed to enhance government-industry collaboration on critical infrastructure security, replacing the 20-year-old CIPAC model.
Key findings • Seven vulnerabilities disclosed in Netty on July 23, 2026, affecting multiple components. • Issues range from replay attacks and denial-of-service to information leaks and secu…
Key findings • Nine BIND 9 vulnerabilities disclosed on July 23, 2026, impacting DNSSEC validation and response handling. • Several flaws allow for DNS response forgery and cache poisoning th…
Key findings • 23 vulnerabilities disclosed simultaneously for Nlnetlabs Unbound on July 23, 2026. • Issues span DNSSEC, DoQ, RPZ, and core resolver logic, impacting versions 1.10.0 to 1.25.1…
The Python Package Index (PyPI) now prevents new files from being uploaded to releases older than 14 days, a move designed to thwart supply-chain attacks that compromise older, trusted packages.
Microsoft is actively working to resolve an issue affecting Exchange Online that has been mistakenly quarantining customer mailboxes since Sunday, disrupting email and calendar access for affected users.
A widespread Bluetooth vulnerability in aftermarket KARR Security Systems allows attackers to remotely unlock doors, control alarms, and immobilize approximately 2.2 million vehicles.
CISA's new Binding Operational Directive 26-04 mandates federal agencies adopt a dynamic, risk-based approach to vulnerability management, moving away from static CVSS scoring.
An old but persistent Brazilian banking Trojan, dubbed 'Lampion,' continues to target businesses in Portugal, exploiting social engineering tactics and the shared language for financial fraud.
A red teamer successfully gained physical access to a hospital's medical records room by impersonating a new employee and leveraging a nurse's willingness to gossip about a doctor.
Key findings • Four vulnerabilities in MZ Automation's Libiec61850 disclosed on July 23, 2026. • Vulnerabilities include heap/stack buffer overflows and NULL pointer dereferences. • Explo…