Public exploit targets pre-auth root flaw in AnyDesk Linux

Security researchers have published a functional exploit for a severe vulnerability in AnyDesk Linux that allows attackers to execute code with root privileges before a user accepts a connection. The flaw, present in version 8.0.2, was patched by the vendor in June 2026, yet the release notes offered no security context or common vulnerability identifier.

The public release of the exploit code on October 8, 2026, raises immediate concerns for system administrators managing Linux endpoints with remote desktop tools. While the vendor claims the issue is limited to direct connections, the technical details suggest broader risks that require urgent attention from IT teams.

What happened

The vulnerability, dubbed AnyPwn by the researchers who discovered it, is a pre-authentication remote code execution flaw. It targets a heap buffer overflow in the session protocol handler of AnyDesk Linux. Rick de Jager of the V12 security team identified the issue using their proprietary code review engine. The exploit code was made publicly available on GitHub on October 8, targeting specifically AnyDesk Linux build 8.0.2.

AnyDesk addressed the vulnerability in version 8.0.3, released in June 2026. However, the company’s changelog merely stated that they "fixed a bug that could lead to a crash," omitting any mention of security implications, remote code execution, or privilege escalation. No CVE has been assigned to this specific flaw as of October 9, 2026, and the vendor has not issued a formal security advisory. Researchers noted that AnyDesk appeared to remove the download link for version 8.0.2 shortly after the proof-of-concept video was released.

Key details

• The exploit targets a heap buffer overflow in the mode-5 stream packet handler of AnyDesk Linux 8.0.2.
• Attackers can achieve root access without user interaction or connection approval.
• The vulnerability relies on an integer overflow where adding a 16-byte header to a large payload length wraps the 32-bit result to zero.
• Current public exploit code works only over direct TCP connections on port 7070 and is probabilistic in nature.
• Researchers indicate the same code path is reachable via relay servers, though a full exploit chain for relays has not been demonstrated.
• Windows and macOS versions of AnyDesk are not affected by this specific flaw.

Background

Heap buffer overflows occur when a program writes data beyond the boundaries of a memory block allocated on the heap. In this case, the AnyDesk session protocol handler calculates memory allocation size by adding a fixed header size to the user-supplied payload length. Because the calculation uses 32-bit arithmetic without checking for overflow, an attacker can specify a payload length near the maximum 32-bit value. When the header is added, the sum wraps around to a small number, causing the allocator to reserve a tiny buffer. The program then copies the large payload into this small buffer, corrupting adjacent memory structures.

This corruption allows attackers to manipulate internal pointers and execute a Return-Oriented Programming (ROP) chain. ROP is a technique where attackers reuse existing code snippets within the program to bypass security protections like non-executable memory. By chaining these snippets, the attacker can run arbitrary commands with the highest privileges on the system. This type of flaw is particularly dangerous in remote desktop software because these applications often run with elevated permissions to facilitate screen sharing and input control.

Why it matters

For teams that self-host or manage Linux workstations, this incident highlights the risk of relying on vague vendor communications. AnyDesk’s decision to describe a critical root-level remote code execution flaw as a simple "crash fix" prevents automated security scanners and administrators from prioritizing the update correctly. Without a CVE or clear advisory, many organizations may have delayed patching, leaving their systems exposed to the now-public exploit. This lack of transparency complicates compliance audits and incident response planning.

Furthermore, the uncertainty regarding relay server exploitation creates a significant blind spot. AnyDesk states the flaw is limited to direct connections, but researchers confirmed the vulnerable code path is accessible via relays. If an attacker can trigger the overflow through a relay, firewall rules blocking direct port 7070 access would not provide protection. Since many corporate networks restrict direct peer-to-peer connections and force traffic through relays, the actual exposure scope might be wider than the vendor admits. Administrators cannot assume safety based solely on network topology if the application logic itself remains vulnerable.

What you can do

• Update all Linux instances of AnyDesk to version 8.1.0 or at least 8.0.3 immediately.
• Verify that no systems are running the deprecated 8.0.2 build, which is no longer listed on the download page.
• Restrict inbound and outbound traffic on TCP port 7070 at the firewall level if immediate updating is not possible.
• Monitor logs for unusual crashes or restarts of the AnyDesk service, which may indicate attempted exploitation.
• Consider alternative remote access solutions that provide transparent security advisories and timely CVE assignments.
• Audit other remote desktop tools in your environment for similar lack of security disclosure practices.