HabrAugust 18, 2026🇷🇺Translated from Russian

The Tale of Active Directory Domain Sabotage: Architect Plants DNS-Killing Task on Departing Branch Controllers

In a medium-complexity case study originally posted on Habr, an architect recounts events from an old Windows Server 2003 Active Directory domain that spanned 15–20 regional branches, each equipped with its own domain controller linked to headquarters via VPN tunnels, sometimes using GSM modems.

When two branches prepared to split off, the architect was asked by senior management to implement a discreet mechanism that would disrupt operations if the separation occurred. The solution involved placing a monitoring script deep in the file system of the branch controllers and scheduling it via Task Scheduler under an innocuous foreign name.

The script used alive.exe to ping a headquarters IP address twice daily. Successful pings reset a hidden registry counter; repeated failures incremented the counter. A second scheduled task running every 10–15 days checked the counter and, once it exceeded 15, executed net stop dns, causing the domain to become unusable for branch users arriving at work the next morning.

To reduce forensic traces, the architect created a temporary task running as SYSTEM that relocated the scripts, recreated the scheduled tasks under SYSTEM ownership, and then deleted the original task. After the branches departed, the DNS service was observed stopping three times; the local network engineer restarted it each time before finally calling the architect.

The architect concluded that Active Directory becomes inoperable the moment the DNS service stops and that the domain administrator is the single greatest danger to the domain’s existence because of the breadth of privileges available.

Related articles

BoletimSecVulnerabilities & Exploits

Apple Patches CVE-2026-43760 Screen Sharing Flaw Granting Root Access on macOS

A vulnerability in the macOS Screen Sharing feature allows remote attackers to execute commands with root privileges under specific configurations. Identified as CVE-2026-43760 and tied to the screensharingd service, the flaw affects systems with Screen Sharing or Remote Management enabled alongside the legacy VNC password option. Attackers who know the VNC password can exploit unauthenticated account binding to retrieve protected files such as /etc/sudoers or write policies into /private/etc/sudoers.d. This grants non-privileged accounts passwordless sudo rights. The issue stems from VNC-authenticated connections not being mapped to specific macOS accounts while file-transfer components retain root privileges. Apple addressed the vulnerability in macOS Tahoe 26.6 and macOS Sonoma 14.8.8, both released on July 27, 2026.

Security NEXTVulnerabilities & Exploits

CISA Adds Ray AI Framework Flaw CVE-2025-62593 to KEV Catalog After Confirmed Exploitation

US authorities have issued a warning that a critical vulnerability in the Ray framework for scaling AI and Python applications is being actively exploited in real-world attacks. The flaw, tracked as CVE-2025-62593, allows remote code execution through browsers such as Firefox and Safari, potentially enabling attackers to run malicious code simply by displaying an ad on a compromised page. CISA added the issue to its Known Exploited Vulnerabilities catalog on August 17, 2026, and directed federal agencies to apply mitigations by August 20. Originally disclosed in November 2025 and fixed in Ray version 2.52.0, the vulnerability carries a CVSS v4.0 base score of 9.4 and is rated Critical. Developers using Ray in environments that also run Firefox or Safari face additional risk from DNS rebinding attacks that can lead to code injection. Organizations relying on Ray are urged to verify patch status and monitor for signs of compromise.

Security NEXTVulnerabilities & Exploits

GitLab Issues Critical Security Updates Fixing Unauthenticated Project Modification Flaws

GitLab has released security updates for its Community Edition and Enterprise Edition platforms to address two vulnerabilities, including one rated critical. The flaws were reported through the company's bug bounty program and affect multiple supported versions. CVE-2026-19478 is a code injection issue that allows remote attackers to tamper with or delete public projects and user data via GraphQL directives without authentication under certain conditions. CVE-2026-19650 is a CSRF vulnerability stemming from insufficient validation in GraphQL multiplexed query handling, enabling mutation execution through GET requests. Both CVEs received high CVSS scores, prompting GitLab to urge immediate upgrades to versions 19.2.4, 19.1.6, 19.0.8, and 18.11.11.

AntiMalwareVulnerabilities & Exploits

VoLTE Video Call Vulnerability Grants Kernel Access on Unisoc Android Chips

Researchers from SSD Secure Disclosure have disclosed a two-stage attack chain that achieves remote kernel access on Android devices powered by Unisoc chipsets. The exploit begins with a remote code execution flaw in the modem firmware that is triggered by a specially crafted VoLTE video call. Once code executes on the modem, a second privilege-escalation issue abuses shared physical memory between the modem and application processor to reach Android kernel memory. The attack requires an attacker-controlled 4G network and VoLTE infrastructure, and the victim must answer the incoming video call. Vulnerable chip families include T606, T612 and T7250, found in devices such as the Motorola E13, Realme C33 and Xiaomi Redmi A5. No patch or CVE identifier has been issued by Unisoc, and the August Android security bulletin does not address the issue.