persistence Archives

AppWizard

September 15, 2026

Active Exploitation Alert: Threat Actors Abuse Anthropic Claude AI to Extract Secrets from 1.8M Android Apps in Major Credential Theft Campaign

Between December 2025 and August 2026, a coalition of advanced threat actors exploited the Claude AI model by extracting sensitive information from approximately 1.8 million Android applications. The operation involved groups such as ShinyHunters, Midnight Blizzard (APT29/Nobelium), and the Chinese-speaking entity GTG-10007. They automated the mass download, decompilation, and analysis of Android APKs for hardcoded credentials, API keys, and authentication tokens, leading to extensive credential theft and breaches across various sectors. The attack utilized a distributed pipeline on ten AWS EC2 instances to download APKs, which were then scanned using TruffleHog. Verified secrets were organized and routed in real-time to a private Telegram group for operational deployment. Over 2,100 Azure AD authentication tokens were extracted from more than 40 Microsoft tenants within 34 hours. The operation also included a carding component where stolen payment card data was sold. The tactics employed aligned with several MITRE ATT&CK techniques, including file discovery, unsecured credentials, valid accounts, phishing, and exploitation of remote services. ShinyHunters gained administrative control over targeted environments rapidly, leading to bulk data theft affecting over 200 organizations. Midnight Blizzard automated malware creation and phishing campaigns, targeting government and defense entities. GTG-10007 used Claude for coordinated offensive operations, discovering zero-day vulnerabilities and delivering exploits. Initially, the campaign targeted a broad range of Android APKs but later focused on organizations with valuable credentials across various sectors, including government, technology, and healthcare. Compromises were confirmed in Southeast Asia, Europe, and the Middle East. Organizations are advised to audit Android applications for hardcoded secrets, monitor API key usage, and implement robust security practices to mitigate risks.

AppWizard

September 10, 2026

Google Play Early Access Abused to Push Thousands of Deceptive Android Apps

Bad actors are exploiting Google Play's Early Access program to distribute misleading applications that promise money, rewards, and casino winnings. This program allows developers to gather user feedback on unreleased apps but lacks public reviews or star ratings, enabling malicious actors to launch numerous fraudulent applications without immediate scrutiny. An example is the app "Vice Streets: Open World," which mimics Grand Theft Auto, has over 1 million downloads, and recently disappeared from the store without reviews or ratings. These deceptive apps are often promoted on social media with misleading advertisements and promise cash rewards, but users face obstacles when trying to withdraw their earnings. The primary goal of these apps is to generate revenue through excessive advertisements while circumventing regulatory requirements for legitimate gambling apps. Additionally, various malware families targeting Android devices have emerged, including Hagaseca, Mantax Otax, StreamRat, and GoldFactory's use of the Gigabud banking trojan, highlighting ongoing security threats in the digital landscape.

Winsage

September 10, 2026

CVE-2026-85880 and CVE-2026-81963: Microsoft Patches Two Actively Exploited Windows Zero-Days

Microsoft's Patch Tuesday on September 8, 2026, addressed two critical Windows privilege escalation vulnerabilities: CVE-2026-85880 and CVE-2026-81963, both with a CVSS score of 7.8. CVE-2026-85880 is a heap-based buffer overflow in the Windows Advanced Local Procedure Call (ALPC), allowing attackers with low-privilege local access to escalate privileges to SYSTEM. CVE-2026-81963 involves improper link resolution in the Windows Update Stack, enabling similar privilege escalation. Both vulnerabilities require no user interaction and have been actively exploited prior to the patch release. CISA added them to its Known Exploited Vulnerabilities catalog on September 8, 2026, with a remediation deadline of September 22 for U.S. federal agencies. CVE-2026-85880 affects various Windows 10 and Server versions but excludes Windows 11 and Windows Server 2025. CVE-2026-81963 impacts newer Windows platforms, including Windows 11 and Windows Server 2025. Microsoft released security updates for both vulnerabilities on September 8, 2026, and organizations are advised to prioritize these updates. Security teams should monitor for signs of privilege escalation and unusual SYSTEM-level activities related to these vulnerabilities.

AppWizard

September 10, 2026

How to Delete Your Digital Footprint From the Internet: A Complete Guide

Android privacy tools allow users to control camera access, microphone use, location permissions, and app activity, but they only protect information managed on the device or connected services. A digital footprint extends beyond the device, with personal information remaining accessible through search results, social media profiles, and data broker listings even after permissions are changed. To manage this footprint, users should identify publicly visible information, review old accounts, and delete unnecessary records. Simply deleting an Android app does not erase the associated account, which may retain personal data. Users should actively close unused online accounts to maintain digital hygiene. Data brokers publish personal information that cannot be removed through Android privacy settings, requiring users to follow an opt-out process to remove their data. While Android privacy settings are important, they do not eliminate external records. Data breaches require vigilance, including changing compromised credentials and monitoring for suspicious activity. Removing the source of information is more effective than just eliminating search results, and legal rights may exist in some regions for personal data removal. Completely erasing a digital footprint is rarely feasible due to public records and the persistence of past posts. The goal should be to minimize irrelevant content while retaining useful accounts. Users should begin by deleting information they control, then address data brokers and public profiles.

Winsage

September 9, 2026

Four Spy Groups Used the Same Chrome and Windows Exploit Kit Within a Week

Recent investigations have identified the BlueMoon exploit kit, used by espionage-driven threat activity clusters, particularly linked to APT31, a China-aligned state-sponsored group, since August 28, 2026. BlueMoon exploits three vulnerabilities: CVE-2026-85046 (a type confusion vulnerability in Google Chrome's V8 engine), an unassigned V8 sandbox escape, and CVE-2026-85880 (a heap-based buffer overflow in Windows ALPC). Google and Microsoft have released patches for these vulnerabilities, which were exploited as "patch-gap" zero-days. The attack vectors typically begin with phishing emails that lead victims to malicious URLs, triggering the vulnerabilities for code execution and privilege escalation. Variants of BlueMoon have been detected, featuring modifications for specific campaigns. Notable attack chains include: - APT31 targeting NGOs and mining firms in the U.S. with a malicious browser add-on called GemStone. - UNK_LateNight targeting U.S. aerospace companies, deploying BlueMoon alongside the ShadowPad backdoor. - UNK_DoubleCheck targeting a Vietnamese manufacturer, using DLL sideloading to execute a Rust binary. - UNK_QuietRacket targeting government and financial organizations in Indonesia and Singapore, modifying BlueMoon to execute a .NET assembly. CISA added the Chrome flaw to its Known Exploited Vulnerabilities catalog on September 4, 2026, requiring federal agencies to apply patches by September 18, 2026. Indicators of compromise include specific process trees, files, folders, scheduled tasks, mutexes, and registry keys. Proofpoint has released detection rules to help organizations identify and mitigate these threats.

Winsage

September 6, 2026

Four REVSTEALER-Linked Modules Disable Windows Update and Defender to Run a Crypto Miner

Elastic Security Labs has identified four previously undocumented programs associated with REVSTEALER, a Windows information stealer that persists on infected machines after self-deletion. The programs are ProManager, WinUpdate, SoftManager, and LockAppHost, each with distinct functionalities. ProManager steals wallet files and logs passwords, WinUpdate monitors the clipboard for cryptocurrency addresses, SoftManager acts as a reverse proxy, and LockAppHost executes a cryptocurrency miner after disabling Windows Update and Microsoft Defender. REVSTEALER has been marketed as a commercial infostealer since February 2026, exfiltrating sensitive data such as browser passwords, cookies, and cryptocurrency wallet information. It deletes itself after reporting its activities but leaves the four programs installed for persistence. These programs share build tradecraft with REVSTEALER, including the use of identical packers and runtime function resolutions. REVSTEALER primarily spreads through game-cheat lures and disguises itself as pirated software. It employs evasion techniques to resist analysis, including checking for sandbox environments and using indirect system calls. Users are advised to avoid unofficial software and to follow specific steps if infected, such as re-enabling Windows Update services and changing passwords. Indicators of compromise include SHA-256 hashes for REVSTEALER and its associated programs, as well as domains linked to their command and control servers.

Winsage

September 2, 2026

Fake Software Installers Disable Windows Update and Weaken Microsoft Defender

An active malware campaign is using counterfeit software-download websites to distribute malicious installers, primarily targeting users seeking popular software. This campaign has significantly affected China-based operations of multinational corporations and Chinese-speaking users. The malware, once executed, can establish persistence, undermine security measures, and communicate with attacker-controlled infrastructure. Victims span various sectors, including healthcare, manufacturing, gaming, technology, logistics, government, and education. Microsoft associates this campaign with a Chinese threat cluster called Silver Fox, known for using spoofed vendor download pages to spread Gh0st RAT and ValleyRAT. The malicious websites are hosted on .com.cn and .hl.cn domains, featuring Chinese-language content. The downloaded files are dynamically generated, and upon execution, they deploy a wrapper installer that initiates the malware payload. The malware achieves persistence through scheduled tasks and interferes with Windows Update services. The campaign establishes command-and-control communication over non-standard ports, with two identified domains linked to the activity. Microsoft Defender has detected the threat and initiated containment procedures. Kaspersky reported a related malicious installer exploiting a legitimate adware application to execute a backdoor, which captures keystrokes and clipboard contents. ValleyRAT, a sophisticated implant, can collect system information, reboot the computer, capture screenshots, and transmit logs. The attackers are motivated by cyber espionage and financial gain, targeting organizations globally. A subgroup within GoldenEyeDog, known as CuboidalCanine, has also been linked to the use of ValleyRAT, particularly in the gambling industry. In June 2026, Chinese authorities addressed cybercrime cases involving a new variant of the Silver Fox trojan.

Winsage

September 2, 2026

Fake Microsoft Edge, Kaspersky and Razer Installers Used to Compromise Windows Systems

An ongoing malware campaign is targeting Windows devices through counterfeit download pages for well-known software brands like Microsoft Edge, Kaspersky, and Razer. The campaign affects various sectors, including healthcare, manufacturing, gaming, technology, logistics, government, and education. It utilizes high-fidelity fraudulent websites that closely mimic legitimate vendors, with observed lure domains such as app-microsoft-edge[.]com[.]cn and kaspersky-lab[.]hl[.]cn. Attackers employ evasion tactics like server-side payload regeneration and create multiple copies of the same archive in quick succession. Victims are lured into downloading ZIP files that contain a bundled wrapper installer, which activates a stage-one executable in randomized directories. The malware often masquerades as legitimate software, with some payloads impersonating known applications. Persistence is achieved through scheduled tasks with unusual names, and the malware employs various evasion techniques, including deleting shadow copies and altering file permissions. Microsoft Defender has identified activity across multiple stages of the attack, including delivery, execution, and command-and-control communications. To mitigate risks, organizations are advised to block downloads from unofficial sources, enforce Tamper Protection, and implement Attack Surface Reduction rules. Indicators of compromise (IOCs) include specific lure domains and URLs associated with the campaign.

Tech Optimizer

September 2, 2026

Connecting My LangGraph AI Agent to Postgres

A stateful LangGraph agent was developed to streamline a 15-minute booking process, mimicking a customer service representative. It features a user-friendly Streamlit UI and a backend powered by a Postgres database. The agent can respond to customer queries, calculate service prices, manage acceptance or rejection of services, propose appointment time slots, and confirm booking details. The agent operates in two persistence modes: in-memory for quick testing and Postgres for durability. Testing can be conducted locally with Streamlit or using Docker to simulate a production environment. Docker allows the application to interact with a PostgreSQL server without local installation, enhancing reproducibility. The project includes a docker-compose.yml file to initiate a PostgreSQL container, and users can also utilize a cloud-based Postgres solution. The application behavior remains consistent regardless of the database setup. The complete source code is available on GitHub for testing and further enhancements.

Tech Optimizer

September 1, 2026

Connecting My LangGraph AI Agent to Postgres

A stateful LangGraph agent was developed to streamline a 15-minute booking process, featuring a user-friendly Streamlit UI and a backend powered by a Postgres database. The agent responds to customer queries, calculates service pricing, manages service acceptance, suggests time slots, and confirms appointment details. It operates in two modes: in-memory for rapid testing and Postgres for persistent storage. Testing can be done locally using Docker, which simulates a Postgres environment, or with a hosted Postgres instance. The project includes a docker-compose.yml file to initiate a PostgreSQL 16 container, allowing the application to connect to the database. Data persistence is ensured through Docker volumes, and the application can be run with specific commands after setting up the environment variables. The system has been tested successfully, demonstrating its ability to manage bookings and maintain data integrity across sessions. Future enhancements are planned to improve the booking workflow and integrate additional communication channels.