4 min read
Stay ahead of the game
Loading

click here to copy URL

Introduction

This report is generated using Velma (Vulnerability Enhanced Learning Machine AI) – Rootshell’s exploit intelligence engine.

Velma focuses on one thing: understanding when vulnerabilities actually become a problem.

There’s no shortage of vulnerability data out there, and most of it is driven by static scores. But risk isn’t static. A vulnerability can sit there for months with little real-world relevance, then overnight become critical when exploit code is released or it starts being used in the wild.

Velma tracks that shift.

By analysing exploit availability, attacker activity, and how vulnerabilities are being used in real-world scenarios, Velma highlights what’s genuinely worth paying attention to – not just what’s highly scored, but what’s actually exploitable.

This report provides a current view of the threat landscape, prioritizing vulnerabilities that are actively being weaponised or realistically used in attack paths.

For most organizations, the challenge isn’t a lack of vulnerabilities – it’s knowing which ones actually matter.

Jump to:

Velma Threat Prioritization Matrix - March 26

Priority
Threat
CVE
Likelihood
Impact
Exploit Maturity
Velma Score
1
Veeam Backup RCE
21666 / 21667 / 21708
High
Very High
High
9.8
2
SolarWinds Web Help Desk
26399
High
Very High
High
9.7
3
SolarWinds Serv-U Chain
40538–40541
High
Very High
High
9.6
4
Chrome Exploit Chain
3909 / 3910
Very High
High
High
9.4
5
VMware Aria Ops (KEV)
22719
High
Very High
High
9.3
6
Ivanti Auth Bypass
1603
High
High
High
9.0
7
Cisco SD-WAN
20775
Medium
Very High
Medium
8.6
8
Veeam LPE
21668 / 21672
Medium
High
Medium
8.4
9
Cisco File Overwrite
20122
Medium
High
Medium
8.2
10
FileZen Injection
25108
Medium
High
Medium
8.1
11
Microsoft MSHTML
21513
Medium
High
Medium
8.0
12
VMware SSRF
22054
Medium
Medium
Medium
7.5
13
Apple Memory Corruption
43000
Medium
Medium
Medium
7.3
14
Cisco Info Disclosure
20128
Low
Medium
Low
6.5
15
Wing FTP
47813
Low
Low
Low
5.8

Executive Summary

Velma identifies multiple high-confidence attack paths across commonly deployed technologies, with several vulnerabilities already being actively exploited or highly likely to be weaponised.

The most significant risks centre around:

  • Remote code execution vulnerabilities in externally exposed or widely used platforms (SolarWinds, VMware, Chrome)

  • Backup and recovery systems (Veeam), which remain a primary ransomware target

  • Authentication bypass and endpoint management platforms (Ivanti), providing direct routes to privileged access

What stands out is not just the severity of individual vulnerabilities, but how they can be combined — enabling a progression from initial access through to full infrastructure compromise.

Immediate focus should be on:

  • Prioritising vulnerabilities with confirmed or likely exploit activity

  • Securing externally accessible services and user-driven entry points (browsers, web apps)

  • Protecting backup infrastructure and core control systems

  • Addressing authentication and privilege escalation pathways

Without this focus, organisations remain exposed to a clear and well-established attack chain — from entry point through to operational impact, including ransomware and data compromise.

Velma Intelligence Assessment

🔴 Critical Risks

Veeam Backup & Replication – RCE
Backup infrastructure remains a primary ransomware target. These vulnerabilities allow code execution directly on backup systems, removing recovery options and increasing impact significantly.

SolarWinds Web Help Desk – RCE
Deserialization vulnerabilities are consistently reliable and quickly weaponised, making this a strong initial access vector.

SolarWinds Serv-U – Exploit Chain
Multiple vulnerabilities combine to enable full system compromise, including admin account creation and root execution.

Chrome Exploit Chain
One of the highest likelihood entry points. Commonly used in phishing, malvertising, and watering hole attacks.

VMware Aria Operations (KEV)
Confirmed active exploitation in the wild with no authentication required.

Ivanti Endpoint Manager – Authentication Bypass
Endpoint management platforms provide centralised control, making exploitation particularly high impact.

🟠 High Risks

  • Cisco SD-WAN privilege escalation

  • Veeam file manipulation and persistence

  • Cisco arbitrary file overwrite

  • FileZen command injection

  • Microsoft MSHTML security bypass (linked to APT activity)

🟡 Medium Risks

  • VMware SSRF

  • Apple web content processing

  • Cisco information disclosure

  • Wing FTP information leakage

Velma Correlated Threat View

Velma identifies a clear multi-stage attack pathway:

  • Initial Access: Chrome, MSHTML, Ivanti

  • Execution: SolarWinds, VMware, FileZen

  • Escalation: Cisco SD-WAN, Veeam

  • Impact: Backup compromise, infrastructure control, ransomware potential

Vulnerability Data (Full List)

Top Reported Known Exploitable Issues:

Chrome
CVE-2026-3909 / CVE-2026-3910 – Out-of-bounds write and sandbox escape vulnerabilities enabling remote code execution.

Cisco SD-WAN
CVE-2022-20775 – Privilege escalation via CLI enabling root command execution.

Cisco
CVE-2026-20128 – Information disclosure vulnerability
CVE-2026-20122 – Arbitrary file overwrite vulnerability

SolarWinds Web Help Desk
CVE-2025-26399 – Deserialization vulnerability enabling remote command execution

SolarWinds Serv-U
CVE-2025-40538–40541 – Multiple vulnerabilities enabling admin creation and root-level execution

VMware Aria Operations
CVE-2026-22719 – Command injection with active exploitation (CISA KEV)

VMware Workspace
CVE-2021-22054 – SSRF vulnerability

Ivanti Endpoint Manager
CVE-2026-1603 – Authentication bypass exposing credential data

Apple iOS / macOS
CVE-2023-43000 – Memory corruption via malicious web content

Veeam Backup & Replication
CVE-2026-21666 / 21667 / 21708 – Remote code execution
CVE-2026-21668 / 21672 – File manipulation and privilege escalation

FileZen
CVE-2026-25108 – Command injection vulnerability

Wing FTP
CVE-2025-47813 – Information disclosure via error messages

Microsoft MSHTML
CVE-2026-21513 – Security feature bypass linked to APT28 activity

Say hello to Velma!

Hello, I’m Velma, Rootshell’s Platform Vulnerability Enhanced Learning Machine AI. My purpose is to inform you about significant technical vulnerabilities and exploits that require immediate attention through patching or configuration changes. Similar to human security analysts, I tirelessly scour numerous forums, websites, and social media channels to provide what I deem as pertinent Threat Intelligence regarding known exploitable vulnerabilities.  

Whilst I don’t yet have the ability to track data breaches in the Rootshell platform watch this space I have some powerful useful supply chain monitoring capabilities on my roadmap.