The Good

Amid the relentless tide of cyber threats, some positive strides are being made to fortify the digital ecosystem. OpenSSF has introduced the Open Source Project Security (OSPS) Baseline, a structured set of best practices designed to help developers mitigate risks and align with regulations like the EU’s Cyber Resilience Act and frameworks like NIST’s Secure Software Development Framework. Meanwhile, vlt has upped supply chain security in the JavaScript ecosystem with reproduce, a new tool that verifies whether npm packages can be faithfully rebuilt from their source code. This initiative marks crucial progress in securing the foundations of open-source development.

• The Open Source Security Foundation (OpenSSF) has launched new best practices to improve the security of open-source projects. The Open Source Project Security (OSPS) Baseline outlines tasks for developers to minimize risks and ensure compliance with regulations such as the EU’s Cyber Resilience Act, aligning with other security standards like the NIST Secure Software Development Framework.
• vlt, an open-source JavaScript package manager, has launched reproduce, a tool that verifies whether published npm packages can be accurately rebuilt from their source code. This tool aims to improve upon traditional methods of provenance in the JavaScript ecosystem. Its relevance is underscored by recent supply chain incidents, such as the compromise of the Ultralytics PyPI package and another incident involving a malicious version of the BoltDB package.
• OpenAI has taken down ChatGPT accounts used by Chinese threat actors, particularly for developing spying tools. A recent report from the company highlighted one operation called ‘Peer Review’, using ChatGPT accounts to create and share spying tools. These tools were designed to analyze social media posts about Chinese politics and to find calls for human rights demonstrations in China, which would be reported to Chinese authorities. The chatbot was also abused for a different China-linked operation that involved generating social media content written in English and long-form news articles written in Spanish.

The Bad

Cyber threats are escalating, with the Vo1d botnet infecting 1.6 million Android TV devices across 226 nations, enabling ad fraud and illegal proxy services. Meanwhile, GitVenom is deceiving gamers and crypto investors with hundreds of fake GitHub repositories, distributing stealers, RATs, and clipboard hijackers, netting attackers 5 BTC. On the espionage front, Lotus Blossom is targeting government and telecom sectors in the Philippines, Vietnam, Hong Kong, and Taiwan, using the Sagerunex backdoor and leveraging Dropbox, Twitter, and Zimbra webmail for covert operations.

• A malware campaign named GitVenom has been targeting GitHub users, particularly gamers and crypto investors. The campaign involves hundreds of fake GitHub repositories created by threat actors, disguised as open-source projects such as automation tools, crypto bots, and hacking utilities. These repositories contain malicious code written in various programming languages. The attackers use AI-generated README files, tags, and fake commits to make the repositories appear legitimate. The malware aims to download further malicious components and execute them. The campaign has deployed a Node.js stealer, AsyncRAT and Quasar backdoors, and a clipboard hijacker, enabling the operators to earn around 5 BTC. The campaign may have been active for several years, with most infection attempts observed in Russia, Brazil, and Turkey.
• CYFIRMA researchers found SpyLend, an Android malware, downloaded 100,000 times from Google Play, disguised as the Finance Simplified app. It targets Indian users through unauthorized loan applications, leading to predatory lending, blackmail, and extortion. The app tricks users by showing fake loan applications instead of providing useful features. It directs users to download malicious apps from other sites, bypassing safety measures. Once installed, it steals sensitive information and can manipulate photos to create fake nudes for extortion.
• A large-scale malware campaign used a faulty Windows driver to avoid detection and spread a variant of Gh0st RAT. Thousands of harmful samples were found with the capability to disable EDR software using a Bring Your Own Vulnerable Driver (BYOVD) attack. Findings suggest the possible involvement of Silver Fox APT. The malware is distributed via fake applications on deceptive websites and messaging apps. It drops the driver and subsequent malware, which targets security processes. The modified driver allowed attackers to bypass Windows protections, leading to the installation of HiddenGh0st, used for remote control and data theft.
• A security flaw in Cyberhaven's browser extension allowed hackers to steal cookies when users visited rogue sites. The browser extension comprises a pop-up, a background script, and content scripts. While content scripts run alongside web pages, they can communicate with the background script. The flaw arose as the extension allowed websites to access user cookies without restrictions, leading to potential cookie theft. Hackers could exploit this by setting up a bogus site on subdomains linked to GitHub. Cyberhaven patched the flaw by revising its user identification to minimize cookie access and enhance security against these attacks.
• Since November 2024, a new variant of the Vo1d botnet has infected 1,590,299 Android TV devices in 226 nations. The botnet peaked on January 14, 2025, and has 800,000 active bots. It has evolved with advanced encryption and stealth features. The infected nations are Brazil, South Africa, Indonesia, Argentina, Thailand, and China. The botnet facilitates illegal activities by using compromised devices as proxy servers and engages in ad fraud. Android TV users must apply firmware updates and isolate IoT devices from sensitive data on the network.
• A new SQL injection vulnerability in the Exim mail transfer agent (CVE-2025-26794) can let attackers take control of email systems and compromise databases. This issue affects Exim version 4.98 installations that use SQLite to store hints databases, with risks including data theft and full system takeover. The flaw arises from how Exim processes SQL queries under certain configurations. Systems are at risk when compiled with the _USE_SQLITE_ option and when ETRN commands are enabled without proper safety measures. Attackers can inject harmful SQL commands through crafted email transactions, gaining access to sensitive database information.
• A malware campaign was aimed at the Royal Thai Police by the Chinese group Mustang Panda. It uses fake FBI documents to deliver a shortcut file that runs the Yokai backdoor. The attack involves a rogue RAR archive holding a shortcut file and a disguised PDF. The shortcut uses Windows FTP to execute commands from the PDF, leading to a trojanized PDF-XChange Driver Installer. This trojan employs evasion techniques, connects to a C2 server, and targets Thailand specifically. The campaign reflects Mustang Panda's TTPs, focusing on government bodies in Asia.
• A potential supply chain compromise has resulted in the hijacking of 16 Chrome extensions, endangering over 3.2 million Google Chrome users. The attackers injected harmful scripts to redirect web traffic for fraud. Hackers gained control via compromised developer accounts to push harmful updates to the Chrome Web Store, unnoticed by users. The malicious code could steal data, modify web activity, and inject ads without users' knowledge. Although Google has removed the affected extensions from the store, they remain installed for users.
• A threat actor named Lotus Blossom (aka Spring Dragon, Billbug, Thrip) conducted multiple cyber espionage campaigns against the government, manufacturing, telecommunications, and media sectors in the Philippines, Vietnam, Hong Kong, and Taiwan to deliver Sagerunex backdoor and other hacking tools for post-compromise activities. Persistence was gained using specific commands to install Sagerunex within the system registry and configure it to run as a service on infected endpoints. Thr group has also developed novel variants of Sagerunex that can use third-party cloud services like Dropbox, Twitter, and the Zimbra webmail as C2 tunnels.

New Threats

This week, new threats were also found propagating through GitHub, and vulnerabilities in Windows and WordPress. Over 1,100 rogue GitHub repositories spread Redox Stealer, stealing crypto wallets, browser cookies, and gaming credentials while promising game mods and cracked software. Meanwhile, LCRYX ransomware has resurfaced, encrypting Windows files, disabling security tools, and damaging the Master Boot Record (MBR) to pressure victims into paying a $500 Bitcoin ransom. Lastly, a critical WordPress flaw in the Essential Addons for Elementor plugin exposed over two million sites to XSS attacks, phishing, and malware distribution via malicious URL parameters.

• A malware campaign weaponized GitHub repositories disguised as game mods and cracked software, involving social engineering techniques and automatic data theft. Over 1,100 rogue repositories were detected spreading the Redox Stealer, which steals cryptocurrency wallet keys, browser cookies, and gaming platform credentials. Intruders use clever tactics to hide their activities, including fake descriptions and links to bypass detection. Many repositories continue to operate despite reports of malware indicating defects in monitoring.
• An ongoing PayPal phishing scam is tricking users by sending false Macbook purchase notifications that appear to come from PayPal's address settings. Recipients are then tricked into calling the scammer's fake PayPal support number. The scammers direct individuals to a website like pplassist[. ]com, where they must enter a service code provided by the fake employee. Doing so downloads a malicious program allowing the scammers to control the victim’s computer. Once they gain access, scammers can steal money, install malware, or collect personal data.
• A new technique lets hackers bypass Microsoft Outlook’s spam filters, delivering rogue ISO files directly to inboxes. They use hyperlink obfuscation to make rogue links look safe as ISO files can hide malware and bypass traditional detection. While Outlook usually flags risky files, hackers now disguise links to bypass these checks. This poses a significant threat to small and medium businesses as hackers can exploit user trust and evade protections. Firms must boost endpoint security until Microsoft updates its filters.
• A recent scam targeted job seekers in the Web3 field using fake interviews via a rogue app, "GrassCall," which installs malware to pilfer cryptocurrency wallets. To assist those affected, a Telegram group was created for malware removal on Mac and Windows devices. The campaign was run by a Russian-speaking team dubbed "Crazy Evil." The team created a fake online presence, including a website and social media profiles, mimicking a company named "ChainSeeker.io", and posted fake job listings on sites like LinkedIn and CryptoJobsList. Victims were tricked into downloading GrassCall, which installs malware to steal passwords and cryptocurrency wallets. In light of the scam, CryptoJobsList removed the bogus listings and advised users to check their devices for malware.
• A new malware campaign named PolarEdge has been targeting edge devices from Cisco, ASUS, QNAP, and Synology since late 2023. Attackers exploited a critical security flaw (CVE-2023-20118) in certain Cisco routers, which remains unpatched due to the devices' end-of-life (EoL) status. This flaw is used to deploy a TLS backdoor that allows command execution and client connection management. The malware has compromised over 2,017 unique IP addresses globally, mainly in the US, Taiwan, and other nations.
• The LCRYX ransomware, based on VBScript, reappeared in February 2025 after first showing up in November 2024 to target Windows devices. It encrypts files with the .lcryx extension and demands a $500 Bitcoin ransom for decryption. The ransomware runs with administrative rights, disables important system tools, and restricts access to system settings. LCRYX employs complex encryption, deletes backups, disrupts system processes, modifies registry settings, and executes commands to damage the Master Boot Record (MBR). It also avoids detection by disabling antivirus (AV) features and hiding its files. After encryption, it leaves a ransom note, urging victims to pay for decryption keys, and uses psychological tactics to pressure victims. Users should use strong security measures and keep regular backups.
• A critical security vulnerability, CVE-2025-24752, in the Essential Addons for Elementor plugin, used by over two million WordPress websites, has left sites vulnerable to script injection attacks through malicious URL parameters. The flaw, which scored 7.1 on the CVSS scale, allowed attackers to execute reflected cross-site scripting (XSS) attacks by exploiting insufficient input sanitization in the plugin’s password reset functionality. The vulnerability originated from the improper handling of the popup-selector query parameter in the plugin’s JavaScript code. Attackers could create URLs containing malicious scripts in this parameter, which were executed when users clicked the link, leading to potential session hijacking, phishing redirects, or malware distribution. 
• A new scam targets PayPal users with bogus search ads and fraudulent payment links. Scammers use Google search ads that look like legitimate PayPal links, especially on mobile devices. They abuse PayPal’s “no-code checkout” feature to create bogus payment links, misleading users into contacting their fake customer support contact to steal their personal information.