It's been a busy 24 hours in the cyber world with significant updates on actively exploited zero-days, nation-state attacks on critical infrastructure, sophisticated vishing campaigns, and the evolving threat landscape of AI. Let's dive in:
Ivanti EPMM Zero-Days Under Active Exploitation ⚠️
- Ivanti has patched two critical zero-day vulnerabilities (CVE-2026-1281, CVE-2026-1340) in its Endpoint Manager Mobile (EPMM) product, both rated CVSS 9.8 for unauthenticated remote code execution (RCE).
- These flaws are actively being exploited in a limited number of customer environments, allowing threat actors to gain administrative access, move laterally, and potentially access sensitive data like phone numbers and GPS locations.
- While specific IOCs are scarce, defenders should scrutinise Apache access logs for unusual GET requests with bash commands in In-House Application Distribution and Android File Transfer Configuration features, and look for unexpected web shells or WAR/JAR files. If compromised, a full restore from backup or migration to a new EPMM instance is recommended.
🕵🏼 The Register | https://go.theregister.com/feed/www.theregister.com/2026/01/30/ivanti_epmm_zero_days/
Coordinated Cyber Attacks on Polish Critical Infrastructure 🚨
- CERT Polska has detailed coordinated destructive cyber attacks on over 30 wind and solar farms, a manufacturing company, and a combined heat and power (CHP) plant in Poland on December 29, 2025.
- The attacks, attributed to Russia's FSB-linked Static Tundra (aka Berserk Bear, Ghost Blizzard), involved reconnaissance, firmware damage, file deletion, and deployment of custom wiper malware like DynoWiper and LazyWiper.
- Initial access was gained via vulnerable Fortinet perimeter devices and statically defined accounts lacking two-factor authentication, with attackers also exfiltrating data related to OT network modernisation and SCADA systems from M365 services.
📰 The Hacker News | https://thehackernews.com/2026/01/poland-attributes-december-cyber.html
ShinyHunters-Style Vishing Bypasses MFA for SaaS Data Theft 🔒
- Mandiant has observed an expansion of financially motivated ShinyHunters-style (UNC6240) activity, tracked as UNC6661 and UNC6671, using advanced vishing and fake credential harvesting sites.
- These groups impersonate IT staff to trick employees into providing SSO credentials and MFA codes, then register their own devices for MFA to access cloud SaaS platforms, exfiltrate sensitive data, and extort victims.
- Organisations should enhance help desk verification processes, enforce strong passwords, remove SMS/phone/email as MFA options, restrict management access, and implement robust logging and detection for MFA lifecycle changes and SaaS export behaviours, moving towards phishing-resistant MFA like FIDO2.
📰 The Hacker News | https://thehackernews.com/2026/01/mandiant-finds-shinyhunters-using.html
Iran-Linked RedKitten Uses AI for Human Rights NGO Targeting 🐱
- A Farsi-speaking threat actor, RedKitten, linked to Iranian state interests, is targeting human rights NGOs and activists, likely leveraging large language models (LLMs) for tooling development.
- The campaign uses macro-laced Excel documents (fabricated protestor death details) in 7-Zip archives as lures, dropping a C#-based SloppyMIO implant via AppDomainManager injection.
- SloppyMIO uses GitHub as a dead drop resolver for Google Drive URLs, steganographically retrieving configuration for its Telegram Bot API-based command-and-control, enabling command execution, file exfiltration, and persistence.
📰 The Hacker News | https://thehackernews.com/2026/01/iran-linked-redkitten-cyber-campaign.html
Agentic AI: The Next Big Attack Surface 🤖
- A Dark Reading poll indicates that agentic AI is widely expected to become the top attack vector by the end of 2026, due to the expanded attack surface from agents' high access and autonomy, especially with insecure code and "shadow AI."
- Experts highlight that the primary vulnerability lies in what compromised AI agents can access, stressing that authentication and access control, rather than AI safety features, are the critical battleground for securing autonomous systems.
- Deepfakes are also rising as a major social engineering vector for high-value targets, while the adoption of phishing-resistant passkeys is lagging, leaving organisations vulnerable as agentic systems proliferate.
🕶️ Dark Reading | https://www.darkreading.com/threat-intelligence/2026-agentic-ai-attack-surface-poster-child
#CyberSecurity #ThreatIntelligence #Vulnerabilities #ZeroDay #RCE #Ivanti #NationState #APT #CriticalInfrastructure #Poland #Russia #Wiper #ShinyHunters #Vishing #MFA #SaaS #Extortion #Iran #RedKitten #LLM #AI #Deepfakes #ThreatLandscape #InfoSec #CyberAttack #Malware #IncidentResponse
13
Dissent Doe 
CyberNetsecIO
securityaffairs
PrivacyDigest
(in)sicurezza digitale
Security Land
clankussy
BrianKrebs
SOC Goulash