Cybersecurity Brief – 2026-02-05

Thu, 05 Feb 2026 00:00:00 +0000

Major Incidents or Breaches

A hacktivist group claims to have breached the Mexican government, allegedly leaking 2.3 terabytes of data affecting 36 million citizens. The government disputes the sensitivity of the leaked data.
Notepad++ update infrastructure was compromised by suspected Chinese state-backed hackers, resulting in the distribution of a backdoored version of the application.
Ongoing web traffic hijacking campaigns are targeting NGINX servers and management panels such as Baota (BT), redirecting user traffic through attacker-controlled infrastructure.
The DEAD#VAX malware campaign is deploying AsyncRAT via IPFS-hosted VHD phishing files, using advanced evasion techniques to bypass traditional security controls.
Iranian threat actors continue credential theft operations targeting expatriates, Syrians, and Israelis using spear-phishing and social engineering.
Attackers are leveraging Windows .scr screensaver files to deliver malware and remote management tools, exploiting the fact that these executables often evade standard controls.

Multiple critical vulnerabilities (including CVE-2026-25049) have been disclosed in the n8n workflow automation platform, allowing arbitrary system command execution and full host compromise. Public exploits are available.
A critical vulnerability in DockerDash, the AI assistant for Docker, allows remote code execution and data theft due to insufficient instruction validation in the MCP Gateway architecture.
VMware ESXi is affected by a high-severity sandbox escape vulnerability now exploited in ransomware attacks, as confirmed by CISA.
A five-year-old GitLab vulnerability is being actively exploited, prompting CISA to order urgent patching by US government agencies.
Google Looker contained vulnerabilities (dubbed LookOut) that allowed cross-tenant remote code execution and data exfiltration, putting multiple GCP environments at risk.
A fresh critical vulnerability in SolarWinds Web Help Desk is being exploited in the wild, allowing unauthenticated remote code execution.
Malicious NGINX configurations are being used to facilitate large-scale web traffic hijacking and exploitation via React2Shell, with attackers deploying cryptominers and reverse shells.
The EnCase forensic software kernel driver, though long-revoked, is being abused by attackers in an EDR killer tool to disable 59 security products.

The Amaranth-Dragon group, linked to Chinese state-sponsored APT41, is exploiting the WinRAR CVE-2025-8088 vulnerability in espionage campaigns targeting government and law enforcement agencies across Southeast Asia.
Chinese threat actors hijacked Notepad++’s update infrastructure to distribute backdoored software.
Iranian state actors are conducting ongoing credential theft campaigns across the Middle East.
Ransomware group DragonForce is promoting a cartel model, increasing coordination among ransomware gangs.
Hackers are abusing legitimate but revoked drivers to disable endpoint detection and response (EDR) solutions.

AI-driven phishing attacks are increasing in sophistication, with attackers combining technical exploits and identity-based attacks to bypass email security solutions.
Leaked non-human identities, such as API keys and tokens, are becoming significant breach drivers in cloud environments, providing attackers with persistent access.
Attackers are increasingly abusing less-monitored file types (e.g., .scr screensaver files) and legitimate system features to evade detection.
Microsoft has developed a lightweight scanner for detecting backdoors in open-weight large language models (LLMs).
Microsoft is rolling out built-in Sysmon security monitoring in Windows 11 Insider builds, enhancing native endpoint visibility.
Orchid Security has introduced continuous identity observability for enterprise applications, addressing identity management beyond traditional IAM controls.
Prompt injection and data leakage risks have been identified in AI agent networks such as Moltbook.
Early incident response failures are often due to decision-making in the first moments after detection, not technical shortcomings.

CISA has made unpublicized updates to its Known Exploited Vulnerabilities (KEV) catalog, highlighting ransomware operators’ focus on network edge device vulnerabilities.
CISA has mandated the patching of actively exploited legacy vulnerabilities in products such as GitLab.
Mozilla Firefox is introducing an AI feature opt-out, reflecting a trend toward increased user control over AI-driven browser functions.

Blockchain intelligence firm TRM Labs raised $70 million in Series C funding to expand AI-driven disruption of criminal networks.
Varonis acquired AllTrue.ai for $150 million to enhance AI trust, risk, and security management capabilities.
Orion raised $32 million for data security product development and market expansion.