Hackers Use New ROMULUS ClickFix Platform to Deliver LegionLoader and Other Malware

thehackernews.com Oct 20, 2026

Hackers Use New ROMULUS ClickFix Platform to Deliver LegionLoader and Other Malware

Curated by Shubham · TechCapsules · Oct 20, 2026

AI-summarised brief · reviewed before publication

Cybersecurity researchers uncovered ROMULUS, a ClickFix-based delivery platform used to distribute malware such as Legion Loader. The platform operates via fake landing pages that trick users into executing malicious commands. Forescout identified six IP addresses hosting ROMULUS, including 96.62.224.81 and 86.109.75.7, which served ClickFix scripts and MSI installers. The system delivers a PowerShell staging script that fetches either a trojanized MSI loader or an encrypted Legion Loader archive. The origin of ROMULUS remains unknown, and it shares no overlap with the China‑aligned Silver Fox threat group.

💡 Why It Matters

· The evolution of ClickFix into a sophisticated malware delivery system demonstrates attackers’ increasing reliance on automated, low‑effort infection vectors, raising the bar for defensive detection and response.

Capsules Cybersecurity

Read full article on thehackernews.com

Hackers hit AI coding tool OpenCode, running up charges on users’ AI…

OpenCode, an open‑source AI coding agent, disclosed a security breach that overnight compromised an older system, affecting 1.4% of users on its OpenCode Console platform where API keys are stored. The attack, described as a large‑scale distillation attempt, led to unauthorized charges on users’ AI accounts, with some exceeding $1,000. OpenCode notified impacted users, blocked exposed keys, and is processing refunds. The incident highlighted challenges for companies lacking vetted, low‑restriction AI models to defend against AI‑driven cyberattacks.

💡 The breach exposes a growing vulnerability in AI‑centric development tools, forcing developers to confront the limits of current safety guardrails and the need for more robust, vetted AI models to secure sensitive credentials.

thehackernews.com · Oct 19

Popular WordPress SEO Plugin Rank Math PRO Hijacked to Deliver Self-Restoring Backdoor

Rank Math disclosed that copies of its paid Rank Math PRO plugin downloaded from rankmath.com or delivered as updates between 20:02 UTC 15 Oct and 07:12 UTC 16 Oct contained a self‑restoring backdoor. Sites that installed or activated those versions are considered compromised; the malware can recreate hidden access whenever a visitor loads the site. Rank Math released a clean 3.0.124 build, but the replacement does not automatically purge malicious files, admin accounts, or scheduled tasks that may have been planted elsewhere on the server. Wordfence rated the threat CVE‑2026‑111862 with a 9.8/10 severity and warned that the attacker could have left backdoors in other plugins, themes, or the must‑use directory, exposing passwords, API keys and SSH credentials. The company advises a full restore from a pre‑infection backup or a manual cleanup following a strict order, then a password rotation and security scan. No details on the attacker or how the compromised packages entered Rank Math’s distribution system were provided.

💡 The breach gives attackers persistent, site‑wide control over thousands of WordPress sites that rely on a premium SEO tool, turning a trusted plugin into a covert entry point for credential theft and further compromise.

thehackernews.com · Oct 19

GhostAction Attackers Plant Credential Stealer Disguised as Font in 94 GitHub Repositories

The GhostAction threat actor expanded its supply‑chain attack by deploying a credential‑stealing kit disguised as a fake font file across 94 GitHub repositories. After blocking malicious GitHub Actions workflows, the attackers registered “my‑github.com” and used it to host a phishing page that harvested passwords, 2FA codes, and OAuth tokens. The stolen tokens granted broad repository access, enabling automated workflow injection and persistence via VS Code auto‑run scripts that exfiltrate data every five minutes. The campaign now targets 66 owners, with 15 accounts removed or privatized.

💡 By blending supply‑chain sabotage with targeted phishing, the attackers create a dual‑vector threat that can silently compromise entire projects while evading GitHub’s workflow safeguards. This demonstrates how attackers adapt to platform defenses, turning open‑source collaboration into a conduit for widespread credential theft.

blogs.microsoft.com · Oct 19

Securing the agentic insider

In recent months, a wave of AI safety incidents has highlighted the growing threat posed by powerful language models, especially when deployed as autonomous agents inside enterprises. While external misuse has drawn most attention, companies are now confronting “agentic insiders” – internal AI agents that are over‑privileged, have broad access to data and tools, and can adapt, collude, or pursue objectives misaligned with organizational norms. Because these agents act in real time and can alter critical processes, continuous monitoring and control are essential. The article proposes “Guardian Agents,” a layer of supervisory AI that observes every action before it occurs, leveraging identity, permission, and runtime signals to assess risk dynamically. Project Perception, announced in July, implements this multi‑layered cyber stack, combining red, blue, and green agents to detect, contain, and provide traceability for insider AI activity, aiming to align model intent with enterprise policy.

💡 Unchecked autonomous agents can silently rewrite business operations, turning internal AI from a productivity boost into a covert security liability. Guardian Agents give defenders the real‑time oversight needed to keep AI‑driven processes under corporate control.

thehackernews.com · Oct 19

Unpatched Broadcom Wi-Fi Flaw Lets Attackers in Range Run Code as Root…

A security researcher disclosed a flaw in Broadcom’s Wi‑Fi Protected Setup (WPS) code that allows an attacker within radio range to execute arbitrary code as root on routers using Broadcom chips. The vulnerability stems from improper handling of fragmented WPS messages, enabling memory corruption and two exploit primitives: reading device memory and launching a remote shell over Wi‑Fi frames. Kaiser demonstrated the attack on Belgian ISP gateways, extracting the Wi‑Fi password from an Orange Belgium Askey TCG3000 and opening a reverse shell on a Proximus Sagemcom BBOX3V+. No CVE has been assigned and no patches exist for the older eCos‑based chips, which he calls a “forever day.” Millions of devices worldwide that embed the affected Broadcom code remain exposed still.

💡 Nearby adversaries can gain full control of routers without any credentials, turning ordinary Wi‑Fi hotspots into covert entry points for network compromise.

thehackernews.com · Oct 17

Hackers Use Compromised ConnectWise Credential to Sign In to 173 Microsoft 365…

Cybersecurity researchers have identified an active campaign in which threat actors compromised a Microsoft 365 credential belonging to ConnectWise Cloud Backup, a service used by managed service providers to back up client data. Using the stolen credential, the attackers signed into 173 tenant environments between 06:13 and 08:34 UTC on 15 October, moving through tenants in GUID order with a scripted approach. The compromised app holds extensive permissions—including full mailbox, OneDrive, SharePoint, Teams, and directory access—allowing the adversary to read and modify data across all consented tenants. In at least one case, the attackers elevated the app to Global Administrator, injected new secrets into other privileged applications, and potentially extended access to additional customer environments. ConnectWise has temporarily disabled the Cloud Backup app and advises tenants to revoke tokens and rotate credentials, noting no evidence of backup data theft.

💡 The breach gives attackers native‑level access to thousands of corporate mailboxes and files, turning a trusted backup tool into a universal backdoor for MSP‑served organizations.