Recent research exposes critical flaws in Large Language Model (LLM) defense architectures through “role confusion” attacks, where prompt injections manipulate AI systems into misinterpreting instructions and security boundaries. The paper demonstrates how current safeguards fail when attackers expl
Brazil’s national emergency alert system was breached in a suspected cyberattack, sending fraudulent emergency notifications to millions of mobile devices nationwide. The unauthorized alert falsely warned citizens of an imminent threat, causing widespread panic and confusion. Authorities have launch
Google has released Gemini 3.5 Flash with computer use capabilities that enable autonomous AI agents to interact directly with systems, applications, and interfaces. While these features unlock powerful automation possibilities, they introduce significant security risks including unauthorized system
Artificial intelligence is fundamentally disrupting cybersecurity operations by discovering vulnerabilities faster than traditional security frameworks can handle. Organizations face an unprecedented challenge: AI-powered tools are identifying thousands of potential weaknesses daily, overwhelming ex
IBM has unveiled groundbreaking sub-1 nanometer chip technology, representing a significant leap in semiconductor miniaturization. While this advancement promises unprecedented computational power and efficiency, it introduces novel security challenges including increased vulnerability to side-chann
Cybercriminals are launching sophisticated phishing campaigns targeting website owners with fake domain renewal notices. These scams impersonate legitimate domain registrars, creating urgency around expiration dates to trick victims into paying renewal fees to fraudulent payment portals. The attacks
The Mistic backdoor has emerged as a sophisticated threat linked to the KongTuke threat actor, targeting multiple industry sectors through ClickFix social engineering techniques and ModeloRAT campaigns. This multi-stage attack employs PowerShell obfuscation, fileless execution, and advanced evasion
Cisco has confirmed active exploitation of CVE-2026-20245, a critical zero-day vulnerability in Cisco Catalyst SD-WAN Manager that allows unauthenticated remote attackers to gain root-level access. The flaw affects multiple Catalyst SD-WAN Manager versions and is being weaponized in the wild before
Nation-state threat actors have successfully infiltrated Australia’s critical infrastructure networks, establishing persistent access designed to enable destructive attacks at strategically opportune moments. The compromise represents a sophisticated pre-positioning campaign where adversaries have e
Google has released Chrome version 149 addressing multiple critical vulnerabilities, including several use-after-free flaws that could enable remote code execution (RCE) attacks. The update patches high-severity issues in core browser components including V8 JavaScript engine, Dawn graphics API, and