Security researchers at Calif, a Palo Alto-based cybersecurity firm, have used techniques derived from an early version of Anthropic’s secretive Mythos AI model to uncover two previously undocumented vulnerabilities in Apple’s macOS. The bugs were chained together into a privilege…
Tag: Cyber Security News
node-ipc npm Package with 822K Weekly Downloads Compromised in Supply Chain Attack
A widely used JavaScript inter-process communication library has been weaponized again. Socket and Stepsecurity have confirmed that three newly published versions of node-ipc, a package with over 822,000 weekly downloads, contain obfuscated stealer and backdoor payloads, marking the second major…
New Malware Framework Enables Screen Control, Browser Artifact Access, and UAC Bypass
A newly uncovered malware framework is raising serious alarms across the cybersecurity community. Researchers have identified a previously unknown implant called TencShell, a sophisticated tool capable of giving attackers full remote control over a compromised system. The discovery highlights how…
Dell Support assist Updates Forces Windows Systems to BSOD Loop
A faulty update to Dell’s SupportAssist Remediation service is sending thousands of Dell and Alienware laptop users into endless Blue Screen of Death (BSOD) loops, with systems crashing every 30 minutes and displaying the dreaded CRITICAL_PROCESS_DIED stop error. Dell Engineering…
New Critical Exim Mailer Allows Remote Attacker to Execute Arbitrary Code
A critical vulnerability in the widely used Exim mail server allows unauthenticated attackers to execute arbitrary code and fully compromise exposed servers. Federico Kirschbaum, head of the Security Lab at XBOW, discovered and reported the issue, which has been dubbed…
Amazon Quick Bug Exposed AI Chat Agents to Users Blocked by Custom Permissions
Imagine locking your organization’s sensitive data behind a heavy vault door, only to realize the locking mechanism is entirely missing. Security researchers at Fog Security recently uncovered a severe authorization bypass in Amazon Quick’s AI Chat Agents. This vulnerability allowed…
Microsoft Research Shows AI Can Generate Realistic Command Lines and Process Telemetry
Artificial intelligence is now capable of generating attack telemetry that looks and behaves like the real thing, and that is changing how security teams think about testing their defenses. In new work, Microsoft researchers show that large language models can…
Palo Alto PAN-OS 0-Day Exploited to Execute Arbitrary Code With Root Privileges on Firewalls
A critical vulnerability in Palo Alto Networks PAN-OS is putting enterprise firewalls at risk, allowing unauthenticated attackers to execute arbitrary code with root privileges. Tracked as CVE-2026-0300, the flaw affects the User-ID Authentication Portal (Captive Portal) and has already seen…
Hackers Abuse Legitimate HWMonitor Binary to Load Malicious DLL Payload
Hackers are once again turning familiar tools against the very users who trust them. A new attack campaign has been discovered in which threat actors weaponized HWMonitor, a widely used hardware monitoring utility developed by CPUID, to silently deliver a…
Critical GitLab Vulnerabilities Enables XSS and Unauthenticated DoS Attacks
Threat actors are constantly hunting for infrastructure weaknesses, and a newly discovered batch of vulnerabilities in GitLab just handed them a dangerous roadmap. On May 13, 2026, GitLab rolled out emergency security updates to address multiple high-severity flaws. These bugs…
OpenAI Hit with Class-Action Privacy Lawsuit for Sharing ChatGPT Data with Google and Meta
OpenAI Global LLC is facing a new class‑action complaint in the Southern District of California that accuses the company of quietly wiring its ChatGPT web interface with Meta’s Facebook Pixel and Google Analytics, turning highly sensitive chatbot conversations into monetizable…
Lyrie.ai Launches the Global Identity Standard for the AI Agent Age & Anthropic’s Cyber Verification Program
DUBAI, UAE — May 11, 2026 — As the internet transitions from a playground of chatbots to a workforce of autonomous agents, the question isn’t just what AI can do—it’s who the AI is. Today, OTT Cybersecurity LLC officially launched the Agent Trust Protocol…
Critical 18-Year-Old NGINX Vulnerability Enables Remote Code Execution Attacks
A critical heap buffer overflow vulnerability has been discovered in the source code of NGINX, present since 2008. This vulnerability has been publicly disclosed, along with a working proof-of-concept exploit that can enable unauthenticated remote code execution (RCE) against one…
Windows DNS Client Vulnerability Enables Remote Code Execution Attacks
A newly disclosed vulnerability in the Microsoft Windows DNS Client could let attackers silently execute malicious code across enterprise networks, exposing a massive attack surface. Officially designated as CVE-2026-41096, this critical security flaw carries a severe CVSS score of 9.8…
Seedworm APT Abuses Signed Fortemedia and SentinelOne Binaries for DLL Sideloading
Iran-linked hackers have been quietly breaking into networks around the world, and their latest campaign is more calculated than anything we have seen from them before. The group known as Seedworm, also tracked as MuddyWater, spent the first quarter of…
Packagist Urges Immediate Composer Update After GitHub Actions Token Leak
Packagist is sounding the alarm for PHP developers everywhere. A flaw in Composer, the widely used PHP dependency manager, briefly caused GitHub authentication tokens to leak into publicly visible CI logs, raising urgent concerns about credential exposure across thousands of…
Langflow CVE-2026-33017 Exploited to Steal AWS Keys and Deploy NATS Worker
Attackers are now abusing a fresh Langflow vulnerability to quietly steal cloud keys and turn victim systems into workers for a new NATS based botnet. This campaign shows how a single exposed AI workflow tool can become the start of…
The Gentlemen RaaS Leverages Fortinet and Cisco Edge Devices for Initial Access
A ransomware group that only surfaced in mid-2025 has already made a significant mark on the threat landscape. The Gentlemen, a ransomware-as-a-service (RaaS) operation, has quickly risen to become one of the most active ransomware programs in the world, with…
Critical MongoDB Vulnerability Allow Attackers to Execute Arbitrary Code
A newly disclosed critical vulnerability in MongoDB could allow threat actors to execute arbitrary code, potentially handing them complete control over affected servers and exposing millions of records to theft. The vulnerability, officially tracked as CVE-2026-8053, directly impacts MongoDB Server…
Critical 18-Year-Old NGINX Vulnerability Enables Remote Code Execution Attacks – PoC Released
A critical heap buffer overflow vulnerability, lurking in NGINX’s source code since 2008, has been publicly disclosed. Complete with a working proof-of-concept exploit capable of delivering unauthenticated remote code execution (RCE) against one of the world’s most widely deployed web…