A critical supply chain attack compromised the widely-used Axios npm package by publishing malicious versions that automatically installed a cross-platform Remote Access Trojan (RAT).
This incident, tracked as GHSA-fw8c-xr5c-95f9, was actively exploited to steal credentials and secrets from developer workstations and CI/CD pipelines before the malicious packages were removed.
Vulnerability Details
Affected Product: Axios npm package versions 1.14.1 and 0.30.4.
Identifier: GHSA-fw8c-xr5c-95f9
Exploitation Status: Actively Exploited.
Risk & Impact
Triage: Urgent; any system that installed the affected versions during the compromise window should be considered fully compromised.
Attack Vector: An attacker published malicious versions of Axios to the public npm registry with a hidden, malicious dependency. Developers and automated build systems that installed these versions automatically executed a script that deployed a RAT, leading to credential theft and remote access.
Ease of Exploit: Trivial. The exploit occurs automatically upon installation via standard commands like `npm install`.
Action Plan
Immediate Action: Downgrade to a safe version, such as Axios 1.14.0 (for 1.x users) or 0.30.3 (for 0.x users). Assume any environment that installed the malicious versions is compromised; rebuild systems from a known-good state and rotate all credentials, API keys, and secrets.
Workaround: Pin dependencies to known-good versions in lockfiles (e.g., `package-lock.json`) and use `npm ci` in build pipelines to prevent automatic upgrades to untrusted versions.
Detection: Hunt for the presence of the `plain-crypto[-]js` directory in `node_modules`. Monitor network logs for outbound connections to `sfrclak[.]com` or the IP address `142[.]11[.]206[.]73` on port 8000.
Relevant professional terms
Supply Chain Attack
An attack strategy that targets less-secure elements in a software supply network. Instead of attacking a target directly, threat actors compromise an upstream provider, such as a software library or vendor, to distribute malware to downstream users.
Remote Access Trojan (RAT)
A type of malware that provides an attacker with full administrative remote control over an infected computer. Once the RAT is installed, the attacker can send commands and receive data back, effectively turning the victim's machine into a remotely controlled device.
Google has released an emergency security update to address CVE-2026-5281, a high-severity use-after-free vulnerability in the Dawn graphics engine component of Chrome. The vulnerability is confirmed to be actively exploited in the wild, making this the fourth Chrome zero-day patched in 2026.
Vulnerability Details
Affected Product: Google Chrome versions prior to 146.0.7680.177/178 for macOS and Windows, and 146.0.7680.177 for Linux.
Identifier: CVE-2026-5281
CVSS Score: Not yet assigned (High Severity)
Exploitation Status: Actively Exploited
Risk & Impact
Triage: Immediate action is required due to active exploitation.
Attack Vector: A remote attacker can exploit this vulnerability by tricking a user into visiting a specially crafted webpage. This can lead to browser crashes, data corruption, or arbitrary code execution.
Ease of Exploit: The barrier for exploitation is low, as it only requires the user to visit a malicious webpage.
Action Plan
Immediate Action: Update Google Chrome to the latest stable version (146.0.7680.177 or higher for Linux, and 146.0.7680.177/178 for macOS and Windows).
Workaround: There are no specific workarounds; updating the browser is the only mitigation.
Detection: Monitor for suspicious browser behavior, unexpected crashes, or child processes spawned from Chrome.
Relevant professional terms
Zero-Day Vulnerability
A security flaw in software that is known to the software vendor but does not have a patch in place to fix it. Attackers can exploit these vulnerabilities before developers have a chance to release a fix.
Use-After-Free
A type of memory corruption bug that occurs when a program continues to use a pointer after the memory it points to has been freed. This can lead to a variety of security issues, including arbitrary code execution.
Researchers used Claude AI to discover critical remote code execution (RCE) vulnerabilities in Vim and GNU Emacs text editors.
The Vim vulnerability, identified as GHSA-2gmj-rpqf-pxvh, has been patched, while the GNU Emacs vulnerability remains unpatched as the maintainers attribute the issue to Git.
Vulnerability Details
Affected Product: Vim versions 9.2.0271 and earlier
Identifier: GHSA-2gmj-rpqf-pxvh
Exploitation Status: Proof-of-concept exists
Risk & Impact
Triage: Urgent for Vim users; High for Emacs users.
Attack Vector: An attacker can achieve arbitrary command execution by tricking a user into opening a specially crafted file. For Vim, this involves a malicious modeline that bypasses sandbox protections. For Emacs, the exploit involves the editor's version control integration with Git.
Ease of Exploit: Trivial. The exploit requires no user interaction beyond opening the malicious file.
Action Plan
Immediate Action: Vim users should upgrade to Version 9.2.0272 or later.
Workaround: Emacs users should exercise extreme caution when opening files from untrusted sources, as no official patch is available.
Detection: Monitor for the creation of suspicious files or unexpected processes spawned by Vim or Emacs.
Relevant professional terms
Remote Code Execution (RCE)
A class of software vulnerability that allows a malicious actor to execute arbitrary commands or code on a target machine or in a target process over a network.
Modeline
A feature in text editors like Vim that allows settings to be embedded within the first or last lines of a file. If not handled securely, modelines can be abused to execute commands when the file is opened.
Palo Alto Networks researchers discovered a high-severity privilege escalation vulnerability in Google's Vertex AI platform where excessive default permissions for a default service account could be exploited by an attacker to steal sensitive data and access restricted cloud infrastructure.
Following responsible disclosure, Google has updated its documentation to recommend stricter access controls.
Vulnerability Details
Affected Product: Google Cloud Platform Vertex AI Agent Engine
Exploitation Status: Proof of concept exists. Not actively exploited in the wild.
Risk & Impact
Triage: High. The vulnerability allows for significant data exfiltration and unauthorized access to cloud resources.
Attack Vector: An attacker could deploy a malicious AI agent which, once active, queries the cloud metadata service to extract credentials for the overprivileged default service account. These credentials can then be used to access other cloud resources, including Google Cloud Storage buckets.
Ease of Exploit: Moderate. Requires the user to deploy a malicious AI agent, but exploits default, out-of-the-box configurations.
Action Plan
Immediate Action: Google has updated its official documentation in response to the findings. Users should immediately adopt a "Bring Your Own Service Account" (BYOSA) architecture.
Workaround: Avoid using the default service agent. Instead, create a custom dedicated service account for Vertex AI deployments with the least privilege necessary for its function.
Detection: Monitor for unusual activity from Vertex AI service accounts, particularly unexpected access to the cloud metadata service or access to a wide range of storage buckets.
Relevant professional terms
Overprivileged
A security risk where a user or service account is granted more access rights and permissions than are necessary to perform its intended function.
Principle of Least Privilege
A security concept in which a user is given the minimum levels of access - or permissions - needed to perform their job functions. The research highlights that granting broad permissions by default violates this principle.
Cisco's development environment was breached by threat actors who used credentials stolen from the recent Trivy supply chain attack. The incident resulted in the theft of source code for internal products and customer projects.
Attack Overview
Attack Path: Threat actors compromised a malicious GitHub Action plugin associated with the Trivy vulnerability scanner to steal CI/CD credentials, which were then used to access Cisco's internal build environment.
Attacker: The supply chain attack on Trivy was conducted by a group known as TeamPCP. Another group, ShinyHunters, has claimed responsibility for the subsequent extortion attempt against Cisco.
Impact Assessment
Data Stolen: Over 300 GitHub repositories were cloned, including source code for AI products and projects for customers like banks and US government agencies. Multiple AWS keys were also stolen.
Strategic Takeaway
This incident underscores the critical risk of supply chain credential compromise, where a vulnerability in a third-party developer tool can directly enable a breach of sensitive internal environments.
Relevant professional terms
Supply Chain Attack
A cyberattack that targets a trusted third-party vendor or software that has access to an organization's systems or data.
CI/CD Pipeline
Continuous Integration/Continuous Deployment is a software development practice that automates the building, testing, and deployment of applications.
The threat actor TeamPCP is conducting a multi-stage supply chain campaign, compromising trusted open-source security scanners and developer tools to harvest credentials and deploy malware.
The group has partnered with the Vect ransomware group to monetize the stolen access and data.
Key TTPs
Initial Access: Compromised maintainer credentials and service accounts to inject malicious code into legitimate software packages and GitHub Actions.
Execution: Malicious code is executed during routine CI/CD pipeline runs, using scripts to extract secrets and deploy further payloads.
Defense Evasion: Used steganography to hide payloads in audio files and masqueraded malware as legitimate system services like systemd.
Campaign Analysis
This campaign marks a significant evolution, weaponizing the very security tools designed to protect environments. The partnership with ransomware groups indicates a strategic shift from data theft to direct financial monetization through extortion.
Targeting & Infrastructure
Target Profile: Developers and organizations using popular open-source tools like Trivy, KICS, and LiteLLM within their CI/CD pipelines.
Actionable Intelligence
Domains:checkmarx[.]zone, models[.]litellm
Relevant Terms
Supply-Chain Attack: An attack strategy that targets less-secure elements in a software supply network, such as third-party code or developer tools, to compromise a final product or organization.
CI/CD Pipeline: Continuous Integration/Continuous Delivery is an automated process for developers to build, test, and deploy code, which, if compromised, can grant attackers broad access.
A zero-day vulnerability in TrueConf's video conferencing client is being actively exploited in the "TrueChaos" campaign against government entities in Southeast Asia.
Attackers are abusing the software's update mechanism to deploy malware by replacing legitimate update packages with malicious ones.
Key TTPs
Initial Access: Attackers compromise the on-premises TrueConf server to replace the legitimate update file with a malicious package.
Execution: The client application downloads and executes the malicious update, which lacks proper integrity and authenticity checks.
Defense Evasion: The attack uses DLL side-loading to load a malicious DLL (7z-x64.dll) and establish persistence.
Campaign Analysis
The TrueChaos campaign highlights a sophisticated supply chain attack, turning a trusted internal update process into a widespread malware distribution channel. This method allowed attackers to infect dozens of connected government agencies by compromising a single central server.
Targeting & Infrastructure
Target Profile: Government entities across Southeast Asia.
Infrastructure: The campaign utilizes the Havoc post-exploitation framework for command and control (C2).
