Anthropic MCP vulnerability
⚡ Weekly Recap: Chrome 0-Day, UniFi Exploits, macOS Stealers, VPN Flaw and More
Published
3 days agoon
By
Jon Tru
Cybersecurity activity this week paints a familiar but escalating picture: attackers are leaning heavily on unpatched systems, trusted platforms, and software supply chains. From a Chrome zero-day under active exploitation to compromised open-source repositories and enterprise VPN bypasses, the threat landscape shows no signs of slowing down.
Below is a structured breakdown of the most significant developments, emerging threats, and security tools from the past week.
⚡ Major Threat of the Week: Chrome Zero-Day Under Active Exploitation
Google has released emergency security updates to fix a critical vulnerability in Chrome’s V8 JavaScript engine, tracked as CVE-2026-11645 (CVSS 8.8).
The flaw, an out-of-bounds memory access issue, is already being actively exploited in the wild. Google confirmed real-world attacks but withheld technical details to ensure widespread patch adoption.
This marks the latest in a series of actively exploited Chrome zero-days this year, reinforcing the browser’s continued status as a high-value target for attackers.
🔴 Enterprise Breaches and High-Impact Exploits
ShinyHunters Exploit Oracle PeopleSoft Zero-Day
The threat group known as ShinyHunters (UNC6240) has been linked to exploitation of CVE-2026-35273 (CVSS 9.8) in Oracle PeopleSoft.
Attackers reportedly gained unauthorized access to enterprise environments, performed internal reconnaissance, and exfiltrated sensitive data. Universities were among the most heavily targeted organizations.
The vulnerability was later added to CISA’s Known Exploited Vulnerabilities (KEV) catalog, underscoring its severity.
Check Point VPN Authentication Bypass
A critical flaw in Check Point Remote Access VPN, CVE-2026-50751 (CVSS 9.3), has been actively exploited in limited attacks.
The issue stems from a certificate validation logic failure affecting systems using the outdated IKEv1 protocol. Attackers can bypass authentication entirely and gain VPN access without valid credentials.
Security researchers confirmed exploitation dating back to early May, with activity increasing in recent weeks.
UniFi OS Remote Code Execution Chain
Security researchers have confirmed active exploitation of a UniFi OS vulnerability chain involving:
- CVE-2026-34908
- CVE-2026-34909
- CVE-2026-34910
When combined, these flaws enable unauthenticated remote code execution as root, making them particularly dangerous for network infrastructure deployments.
🧨 Supply Chain Attacks Escalate
Arch Linux AUR Repository Compromise
Hundreds of abandoned packages in the Arch User Repository (AUR) were compromised and modified with malicious install scripts.
The campaign, dubbed “Atomic Arch,” delivered a malicious npm dependency called atomic-lockfile, designed to steal credentials, evade detection, and execute hidden payloads.
The number of affected packages has grown beyond 1,500, highlighting the scale of supply chain exposure in open-source ecosystems.
Malicious npm and PyPI Packages Target Developers
Multiple campaigns targeting developers have been identified across npm and PyPI ecosystems, including:
- Cryptocurrency credential theft tools
- Wallet and SSH key stealers
- Remote payload execution loaders
Security researchers warn that infected systems should be considered fully compromised, requiring immediate credential rotation.
📱 macOS Malware Campaigns Targeting Users
Attackers are increasingly using fake installers and SEO poisoning to distribute macOS infostealers.
These campaigns rely on:
- Fake software download pages
- Torrent and cracked software distribution channels
- Bypassing Apple Gatekeeper protections
More than 65% of new macOS malware in recent reports has been classified as information-stealing malware.
📡 Phishing-as-a-Service and AI-Driven Fraud
A major takedown effort targeted “Outsider,” a phishing-as-a-service platform responsible for millions of stolen payment cards and billions in fraud losses.
The platform offered:
- Ready-made phishing templates
- SMS-based smishing campaigns
- Subscription-based access for low-skilled attackers
At the same time, attackers are increasingly using AI-themed lures, impersonating tools like ChatGPT and Claude to steal credentials and financial data.
🧠 Emerging Malware and Ransomware Activity
The Gentlemen Ransomware Group
A ransomware operation known as The Gentlemen (Storm-2697) has claimed nearly 500 victims.
The group evolved from affiliate-based ransomware operations and now runs its own RaaS infrastructure, leveraging multiple ransomware families for double extortion attacks.
Residential Proxy Networks Expanding
Residential proxy usage has surged, enabling attackers to mask traffic through legitimate home devices.
These networks are increasingly used for:
- Credential stuffing
- Web scraping for AI training
- Bypassing anti-bot protections
🧩 Additional Global Threat Activity
AI-Themed Phishing Campaigns
Attackers are exploiting public interest in AI tools to deliver phishing kits, malware, and credential stealers disguised as:
- ChatGPT plugins
- Claude tools
- Fake AI installers
Android Trojan “MagicAd”
A new Android malware strain, MagicAd, has been discovered bypassing system restrictions to display background ads. It has been distributed through official app stores and third-party marketplaces.
Tax-Themed Phishing Campaigns
A sophisticated campaign targeting Indian users delivers multi-stage malware through fake tax notices, combining social engineering with memory-resident payload execution.
🛠️ Security Tools of the Week
SpooNMAP
A Python-based wrapper for Nmap and Masscan that simplifies port scanning and supports multiple scan profiles, including custom configurations.
CVE MCP Server
A centralized security intelligence tool that connects Claude-based workflows to CVE databases, EPSS scoring, exploit data, and threat intelligence sources.
🔐 Key Takeaway
This week reinforces a consistent reality in cybersecurity: attackers don’t need advanced exploits—they rely on trusted software, outdated systems, and supply chain blind spots.
Whether it’s a browser zero-day, a compromised open-source package, or a misconfigured VPN, the entry point is often something already sitting quietly inside the environment.
You may like
CISA Flags LiteSpeed cPanel Plugin Flaw Exploited for Root Privilege Escalation
Cisco Releases Security Updates for Actively Exploited SD-WAN Manager Flaw
Chinese Hackers Abused Google Workspace Rules to Steal Research and Defense Emails
North Korean Hackers Are Turning Developer Tools Into Malware Delivery Channels
Critical LiteLLM Security Flaws Enable Full AI Gateway Takeover via Privilege Escalation Chain
Cisco Releases Security Updates for Actively Exploited SD-WAN Manager Flaw
Critical LiteLLM Security Flaws Enable Full AI Gateway Takeover via Privilege Escalation Chain
North Korean Hackers Are Turning Developer Tools Into Malware Delivery Channels
CISA Flags LiteSpeed cPanel Plugin Flaw Exploited for Root Privilege Escalation
Chinese Hackers Abused Google Workspace Rules to Steal Research and Defense Emails

