A week-long automated attack campaign targeted CI/CD pipelines across major open source repositories, achieving remote code execution in multiple targets. The attacker, an autonomous bot called hackerbot-claw, used five different exploitation techniques and successfully exfiltrated a GitHub token with write permissions from one of the most popular repositories on GitHub. The campaign targeted repositories belonging to Microsoft, DataDog, CNCF, and other popular open source projects. The attacks included token theft via poisoned Go scripts, direct script injection, branch name injection, filename injection, and AI prompt injection. The most severe attack resulted in a full repository compromise of Aqua Security's Trivy project. The campaign highlights the growing threat of AI-powered bots targeting software supply chains and the need for automated security controls in CI/CD pipelines.
The report details the resurgence of the Funnull cybercriminal group, now utilizing a new arsenal called RingH23. It exposes their tactics, including compromising GoEdge CDN nodes, poisoning the MacCMS supply chain, and deploying sophisticated malware components like Badredis2s, Badnginx2s, and Badhide2s. The group has expanded its operations to inject malicious JavaScript, hijack cryptocurrency transactions, and redirect traffic to fraudulent sites. The campaign's impact is estimated to affect millions of users daily. The report also highlights Funnull's use of a suspicious new CDN infrastructure, CDN1.AI, likely created to evade detection.
A new campaign exploits OpenClaw skills to distribute the Atomic MacOS Stealer (AMOS). This evolution in supply chain attacks manipulates AI agentic workflows to install malware. The campaign spans multiple repositories with hundreds of malicious skills uploaded to ClawHub and SkillsMP. The infection chain begins with a seemingly harmless SKILL.md file that installs a prerequisite, leading to the download of a Mach-O universal binary. This AMOS variant steals extensive data, including credentials, browser data, cryptocurrency wallets, and various user documents. It lacks system persistence but expands its reach by exfiltrating Apple and KeePass keychains. The malware uses sophisticated encryption schemes and targets multiple browsers and cryptocurrency wallets.
A new Android backdoor called Keenadu has been discovered embedded in the firmware of several tablet brands. It infects the libandroid_runtime.so library during firmware building, injecting itself into every app launched on the device. Keenadu provides attackers unrestricted control over victims' devices, primarily for ad fraud purposes. The investigation revealed connections between Keenadu and other major Android botnets like Triada, BADBOX, and Vo1d. The malware was found in system apps, Google Play apps, and modified versions of popular apps. Over 13,000 users worldwide have been affected, with Russia, Japan, Germany, Brazil and the Netherlands seeing the highest number of infections.
A recent supply chain attack targeting EmEditor users has been uncovered, involving watering hole tactics. The investigation reveals multiple domains masquerading as EmEditor-related sites, all registered through NameSilo LLC in December 2025. The domains resolve to various IP addresses, with some changes observed in February 2026. Additional domains with similar patterns were discovered, along with peculiar HTTP header behavior. A potential early stage of the campaign was identified, sharing similar characteristics with the initial report. The attackers continued their activities even after exposure, utilizing PowerShell scripts and various domains for command and control purposes. The analysis provides a comprehensive list of indicators, including domain names, IP addresses, and file hashes associated with the attack.
A massive malware campaign dubbed ClawHavoc has been uncovered in the ClawHub marketplace, targeting OpenClaw bots and their users. An AI bot named Alex, working with security researcher Oren Yomtov, discovered 341 malicious skills, including 335 from a single campaign. The malware, identified as Atomic Stealer (AMOS), uses sophisticated techniques to evade detection and steal sensitive data. The attack exploits users' trust in AI assistants, potentially compromising personal and financial information. In response, a new tool called Clawdex has been developed to help bots and users scan for malicious skills before installation.
Almost 400 fake crypto trading add-ons in the Moltbot/OpenClaw AI assistant project have been discovered, potentially leading users to install information-stealing malware. These add-ons, known as skills, masquerade as cryptocurrency trading automation tools and target various platforms. The malicious skills share the same command-and-control infrastructure and use social engineering to convince users to execute commands that steal crypto assets. The supply chain attack relies on social engineering and lacks security review in the skills publication process. Security experts warn about the inherent risks of endpoint-native AI agents and emphasize the need for proper security controls and architectural design considerations.
A malware campaign discovered on the NuGet package repository targets the cryptocurrency ecosystem by distributing 14 malicious packages impersonating legitimate crypto-related tools. These packages employ techniques such as homoglyphs, version bumping, and inflated download counts to appear trustworthy and evade detection. The malware aims to steal crypto funds by redirecting transactions and exfiltrating secrets, including OAuth tokens for Google Ads accounts. The campaign highlights the risks of software supply chain attacks, especially for projects relying on compromised dependencies. No known exploits in the wild have been reported yet, but the threat poses a significant risk to developers and organizations integrating these packages. The attack affects . NET developers using NuGet packages related to cryptocurrency and OAuth services. The severity is assessed as medium due to the potential confidentiality and financial impact, combined with moderate exploitation complexity. European organizations involved in blockchain development, fintech, and digital advertising are particularly at risk. Mitigation requires strict dependency vetting, use of package integrity verification, and monitoring of OAuth token usage.
Sha1-Hulud is a sophisticated new variant of an NPM supply chain attack that executes during the preinstall phase of popular packages like Postman, Zapier, and AsyncAPI. It harvests cloud credentials across AWS, Azure, and GCP, and establishes persistence by creating a self-hosted GitHub Actions runner named 'SHA1HULUD' with an injection-vulnerable workflow. This enables attackers to move laterally across cloud environments beyond the initial development environment. Immediate mitigation includes removing compromised packages, revoking and regenerating all tokens and credentials, and enforcing hardware-based MFA for developer accounts. The attack leverages multiple MITRE ATT&CK techniques related to execution, persistence, credential access, and lateral movement. No known exploits in the wild have been reported yet, but the threat poses a medium severity risk due to its potential impact on cloud infrastructure security and software supply chains.
Qilin ransomware is used for domain-wide encryption, and a ransom is then demanded for the decryption keys and/or to prevent the publication of the stolen data. Qilin affiliates are recruited from cybercrime forums to use the Qilin RaaS platform, which handles payload generation, the publication of stolen data, and ransom negotiations.
A novel technique utilizing Ethereum smart contracts was discovered in two npm packages to conceal malicious commands for installing downloader malware. The packages, colortoolsv2 and mimelib2, are part of a larger campaign targeting npm and GitHub. The attackers created sophisticated GitHub repositories with fake popularity metrics to lure developers. The campaign focused on cryptocurrency-related projects, using blockchain technology to evade detection. This incident highlights the evolving strategies of malicious actors in compromising open-source repositories and the need for developers to carefully assess third-party packages before implementation.
A coordinated campaign of 18 malicious browser extensions infected 2.3 million users across Chrome and Edge. These extensions, including a color picker tool, appeared legitimate with verified badges and high install counts. The RedDirection campaign implemented sophisticated browser hijacking mechanisms, capturing users' browsing data and potentially redirecting them to malicious sites. The malware was introduced through version updates of previously clean extensions, exploiting the auto-update feature of browsers. The campaign demonstrates systemic failures in marketplace security, verification processes, and trust signals, turning productivity tools into surveillance malware. Users are advised to remove affected extensions and monitor their accounts for suspicious activity.
A VS Code extension for Ethereum smart contract development, ETHcode, was compromised through a GitHub pull request. The attacker, using a newly created account, submitted a PR that introduced a malicious dependency and code to execute it. The compromise was subtle, involving only two lines of code changes among thousands. The malicious code downloads and runs a batch script from a public file-hosting service, potentially to steal crypto assets or compromise Ethereum contracts. The extension, with nearly 6,000 installs, was removed from the marketplace after discovery. This incident highlights the importance of carefully reviewing contributions, especially from new accounts, and scrutinizing package dependencies in software development workflows.
Banana Squad, a threat actor first identified in 2023, has launched a new campaign targeting GitHub repositories. The group has created over 60 trojanized repositories masquerading as hacking tools written in Python. These malicious repositories contain hundreds of trojanized Python files that utilize encoding and encryption techniques to hide backdoor code. The campaign primarily uses the domain dieserbenni[.]ru, with a new domain 1312services[.]ru detected recently. The trojanized repositories exploit GitHub's UI feature where long lines of code don't wrap, allowing attackers to hide malicious code off-screen. This sophisticated technique makes visual detection challenging, highlighting the increasing stealth of supply chain attacks on open-source platforms.
A new backdoor named Pickai is exploiting ComfyUI vulnerabilities to spread and steal sensitive AI data. Developed in C++, Pickai offers remote command execution and reverse shell capabilities with strong persistence and evasion techniques. It uses multiple C2 servers for redundancy and has infected nearly 700 devices globally. The malware is hosted on Rubick.ai, an AI e-commerce platform serving major brands, posing significant supply chain risks. Pickai employs various obfuscation methods, including string encryption, process disguise, and multiple persistence mechanisms. Its network communication uses a three-tier timing strategy for C2 communication and device information reporting.
A new malicious campaign has been discovered targeting the Python Package Index (PyPI) by exploiting the Pickle file format in machine learning models. Three malicious packages posing as an Alibaba AI Labs SDK were detected, containing infostealer payloads hidden inside PyTorch models. The packages exfiltrate information about infected machines and .gitconfig file contents. This attack demonstrates the evolving threat landscape in AI and machine learning, particularly in the software supply chain. The campaign likely targeted developers in China and highlights the need for improved security measures and tools to detect malicious functionality in ML models.
