Remote code execution

TeamViewer clients have had remote code-execution and related flaws, including a URI-handler vulnerability where a crafted link caused the client to connect outward and leak or execute, and parsing and privilege issues in the desktop client. Against an affected version these give code execution or credential capture through the application rather than by guessing a password.

1 min di letturaaggiornato il 6 ott 2026florianamette1 contributore
Chiedi al tuo assistente (MCP)Modifica su GitHub

CVE-2020-13699 affected TeamViewer Desktop for Windows before specific fixed releases in versions 8 through 15. TeamViewer did not correctly quote arguments passed through its registered custom URI handlers. A malicious page could invoke a URI such as teamviewer10: --play with a UNC path, causing Windows to send an NTLM authentication request to an attacker-controlled SMB server. The attacker could capture the challenge-response for offline cracking or attempt relay where the destination and protections permitted it. This was credential coercion through TeamViewer's URI handling, not direct remote code execution in TeamViewer.

bash
# defensive verification: identify the installed TeamViewer version
(Get-Item 'C:\Program Files\TeamViewer\TeamViewer.exe').VersionInfo.ProductVersion
# compare the relevant major release with the fixed versions in CVE-2020-13699

Exploitation notes#

  • Exploitation requires the vulnerable Windows client to process a crafted URI, typically through user interaction with a malicious page or link.
  • The immediate result is an outbound NTLM authentication attempt. Any further effect depends on password strength and whether NTLM relay is possible; see NTLM relay.
  • Fixed releases differ by TeamViewer major version, so compare the complete installed version with the CVE record rather than using a single global threshold.
  • Do not generalize this CVE into an unspecified parsing or remote-code-execution class.

References#