Cybersecurity Brief – 2026-02-25

Wed, 25 Feb 2026 00:00:00 +0000

Major Incidents or Breaches

Wynn Resorts confirmed an employee data breach following an extortion threat, with the ShinyHunters extortion group listing the company on its data leak site.
CarGurus suffered a data breach exposing information of 12.4 million accounts, with ShinyHunters publishing the stolen personal information.
ShinyHunters claimed responsibility for a breach at Dutch telecommunications provider Odido, reportedly affecting millions of user records.
A phishing campaign by the “Diesel Vortex” threat group targeted freight and logistics organisations in the US and Europe, stealing credentials via 52 domains.
Wynn Resorts, CarGurus, and Odido incidents all point to continued large-scale data theft and extortion targeting enterprises.
Roblox faces a lawsuit from Los Angeles County over alleged failure to protect children from predators.
A Taiwan security firm confirmed exploitation of a vulnerability in TeamT5 ThreatSonar Anti-Ransomware, likely by Chinese APTs.

SolarWinds patched four critical Serv-U remote code execution vulnerabilities that could allow attackers root access to unpatched servers.
VMware Aria Operations patched several vulnerabilities, including high-severity flaws enabling remote code execution.
CISA added FileZen CVE-2026-25108 to its Known Exploited Vulnerabilities (KEV) catalog, confirming active exploitation.
A vulnerability in GitHub Codespaces allowed attackers to seize repository control by injecting malicious Copilot instructions via GitHub Issues.
A new supply chain attack, “Sandworm_Mode”, was discovered in the NPM ecosystem, propagating like a worm, poisoning AI assistants, exfiltrating secrets, and containing destructive capabilities.
Open redirect vulnerabilities remain overlooked, with continued potential for abuse in phishing and redirection attacks.

The North Korea-linked Lazarus Group was observed using Medusa ransomware in attacks against US healthcare and Middle Eastern entities, also deploying the Comebacker backdoor, Blindingcan RAT, and Infohook infostealer.
UAC-0050, a Russia-aligned threat actor, targeted a European financial institution using a spoofed domain and RMS malware, likely for intelligence gathering or financial theft.
UnsolicitedBooker, a threat activity cluster, shifted focus to targeting telecommunications companies in Kyrgyzstan and Tajikistan, deploying LuciDoor and MarsSnake backdoors.
The ShinyHunters extortion group remains active, claiming responsibility for recent high-profile breaches.
“Diesel Vortex” threat group continues credential theft campaigns targeting freight and logistics sectors.
The “Arkanix Stealer” malware, written in C++ and Python, was observed exfiltrating system and browser data before disappearing shortly after its debut.
Arrests of Anonymous Fénix members in Spain disrupted part of the group’s operations.

Threat actors are increasingly using fake video meeting invites (Zoom, Teams, Google Meet) to trick users into installing remote monitoring and management (RMM) tools or surveillance software through social engineering.
1Campaign, a new cybercrime service, enables threat actors to run malicious Google Ads that evade detection and remain online for extended periods.
Attackers now require as little as 29 minutes to gain control of a network, aided by credential misuse, AI tools, and security blind spots.
GitHub Copilot can be abused to process malicious instructions embedded in issues, leading to repository takeover when launching Codespaces.
Social engineering remains a primary attack vector, with increasing sophistication in tactics.
Identity-first security is highlighted as critical, with AI agents now provisioning infrastructure and inheriting over-scoped privileges.
Password length recommendations are being revisited in light of AI and quantum computing threats.
Human risk management tools, such as the AIDA Orchestration Agent, are being introduced to address persistent social engineering risks.

The UK Information Commissioner’s Office (ICO) fined Reddit £14.47 million for unlawfully collecting and processing children’s data under 13 without adequate safeguards.
Ofcom and the ICO fined both Reddit and a US-based pornographic website for failing to protect children’s privacy and safety online.
Microsoft expanded data loss prevention (DLP) controls to restrict Copilot AI from processing confidential Word, Excel, and PowerPoint documents, regardless of storage location.