A new variant of the XCSSET malware, designed to infect Xcode projects, has been identified with key changes in browser targeting, clipboard hijacking, and persistence mechanisms. This variant employs sophisticated encryption and obfuscation techniques, uses run-only compiled AppleScripts for stealthy execution, and expands its data exfiltration capabilities to include Firefox browser data. It also adds another persistence mechanism through LaunchDaemon entries. The malware features a submodule for monitoring the clipboard and substituting wallet addresses. The infection chain consists of four stages, with modifications to the boot function and introduction of new modules. Changes include additional checks for Firefox browser, modified logic for Telegram existence check, and new info-stealer modules targeting Firefox data.
This analysis examines the Bookworm malware family and its connection to the Chinese APT group Stately Taurus. Using a structured attribution framework, the study evaluates tactics, tooling, operational security, infrastructure, victimology and timelines to establish a high-confidence link between Bookworm and Stately Taurus. Key evidence includes shared program database paths, overlapping command and control infrastructure, and consistent targeting of Southeast Asian governments. The framework assigns scores to each piece of evidence, resulting in an overall attribution confidence score of 58.4 out of 100, indicating strong confidence in the connection. This systematic approach aims to improve analytical rigor and collaboration in threat intelligence.
The SystemBC botnet, composed of over 80 C2s and 1,500 daily victims, primarily targets VPS systems from commercial providers. It creates proxies enabling high volumes of malicious traffic for various criminal threat groups. The network is used by multiple proxy services, including REM Proxy, which offers tiered packages for different cybercriminal needs. SystemBC's infrastructure allows for massive data transfers, with some bots generating over 16 GB of proxy data in 24 hours. The botnet is used for various malicious activities, including brute-forcing WordPress credentials, web-scraping, and supporting ransomware operations. The report highlights the evolving nature of proxy services in the cybercriminal ecosystem and their role in facilitating large-scale attacks.
This research explores the challenges posed by LLM-enabled malware, which can generate malicious logic at runtime. The study identifies characteristics of such malware, including embedded API keys and specific prompt structures. Notable cases like PromptLock and APT28's LameHug are examined. The researchers developed hunting strategies based on API key detection and prompt analysis, leading to the discovery of new samples, including 'MalTerminal'. The implications for defenders are discussed, highlighting both the adaptability and potential brittleness of LLM-enabled malware. The research also uncovered various offensive tools leveraging LLMs for operational capabilities.
An analysis of several Windows applications masquerading as legitimate utilities reveals a covert malware operation. These apps, including fake speed testers and AI search tools, install a Node.js runtime and execute obfuscated JavaScript via scheduled tasks. The malware communicates with a command and control server, potentially allowing arbitrary code execution. The operation's sophistication lies in its use of seemingly benign applications as cover for persistent background processes. The malware's capabilities include encoded network communications and the ability to receive and execute remote commands. This technique significantly expands the attack surface, as the malicious component operates independently from the visible application.
I’ve been seeing more and more cases where the SOC reports success, process killed, host isolated, dashboard green. Yet weeks later the same organisation is staring at ransom notes or data leaks.
The problem: we treat every alert like a dodgy PDF. Malware was contained. The threat actor was not.
SOCs measure noise (MTTD, MTTR, auto-contain). Adversaries measure impact (persistence, privilege, exfiltration). That’s why even fully “security-compliant” companies lose millions every day. Look at what's happening in the UK.
Curious how others here are approaching this:
* Do you have workflows that pivot from containment to investigation by default?
* How do you balance speed vs depth when you suspect a human adversary is involved?
* Are you baking forensic collection into SOC alerts, or leaving it for the big crises?
Full piece linked for context.
New Supermicro BMC flaws can create persistent backdoors
Source: https://www.bleepingcomputer.com/news/security/new-supermicro-bmc-flaws-can-create-persistent-backdoors/
Chinese Hackers RedNovember Target Global Governments Using Pantegana and Cobalt Strike
Source: https://thehackernews.com/2025/09/chinese-hackers-rednovember-target.html
Python Software Foundation Warns of Fake PyPI Login Site Stealing User Credentials
Source: https://hackread.com/psf-warn-fake-pypi-login-site-steal-credentials/
RedNovember, a Chinese state-sponsored threat group, has expanded its cyber-espionage activities globally. The group targets high-profile government, intergovernmental, and private sector organizations, focusing on defense, aerospace, and technology sectors. It uses the Go-based backdoor Pantegana and Cobalt Strike for intrusions, exploiting vulnerabilities in perimeter appliances. RedNovember's tactics include combining weaponized proof-of-concept exploits with open-source tools, allowing for scalable operations and attribution obfuscation. The group has shown particular interest in targets across the US, Taiwan, South Korea, and Panama, often aligning its activities with geopolitical events and Chinese strategic interests.
A new multi-stage ClickFix campaign, attributed to the Russia-linked APT group COLDRIVER, has been discovered targeting Russian civil society members. The campaign employs social engineering techniques to trick users into executing malicious commands, leading to the deployment of two new malware families: BAITSWITCH (a downloader) and SIMPLEFIX (a PowerShell-based backdoor). The attack chain involves a fake Cloudflare Turnstile checkbox, persistence establishment, and data exfiltration. COLDRIVER's tactics include using server-side checks, obfuscation techniques, and targeting specific file types for intelligence collection. The group's focus on NGOs, human rights defenders, and Russian exiles aligns with their known victimology.
New YiBackdoor Malware Shares Major Code Overlaps with IcedID and Latrodectus
Source: https://thehackernews.com/2025/09/new-yibackdoor-malware-shares-major.html
In late July 2025, an organized APT attack using shortcut files was discovered, attributed to the North Korean Kimsuky group. The attackers distribute decoy zip files containing password-protected documents and a disguised shortcut file. When executed, it connects to a C2 server, downloads encrypted payloads, and performs various malicious activities. These include collecting sensitive information from browsers, cryptocurrency wallets, messaging apps, and system files. The collected data is encrypted and sent to the C2 server, which can issue additional commands for remote execution. The attack employs anti-VM techniques and establishes persistence through registry modifications. It also includes a separate malicious DLL for browser process injection.
This report details a campaign involving NodeJS backdoors used to distribute proxyware and monetization schemes. The attackers employ Inno setup installers to drop PowerShell scripts that download and execute NodeJS packages with malicious JavaScript. The backdoors collect system information, communicate with command and control servers, and can execute various commands including PowerShell scripts and additional Node.js code. The campaign is associated with multiple proxyware applications like Infatica, Honeygain, earnFM, and PacketLab. The attackers also use browser extensions to track user navigation and potentially redirect to malicious URLs. The infrastructure involves numerous domains and cloud services for hosting malware and command and control.
