ZDI-26-968: (Pwn2Own) OpenAI Codex GUI OpenProjectInCodex Argument Injection Remote Code Execution Vulnerability

Summary

ZDI-26-968 describes an argument injection flaw in the OpenProjectInCodex context menu command of OpenAI Codex, which allows remote attackers to execute arbitrary code as the current user. The issue stems from missing validation of a user-supplied string before it is used to build command-line arguments. Exploitation requires the target to open a malicious file, and the ZDI rates it CVSS 7.8, tracked as CVE-2026-110930.

Mitigation

OpenAI has issued an update to correct this vulnerability. More details are at https://openai.com/security/advisories/cve-2026-110930/

Related items

• MediumGHSA-68g8-gh8f-7rhg: pgEdge RAG Server discloses part of the configured provider credential through API error messagesSame vendor · GitHub Advisory Database
• InfoCVE-2026-119990 Azure OpenAI Elevation of Privilege VulnerabilitySame vendor · MSRC Security Update Guide
• HighCVE-2026-110930: Codex Desktop for Windows argument injection via quoted project pathsSame vendor · NVD/CVE Database
• MediumCVE-2026-111372: AI ChatBot for WooCommerce WoowBot plugin authorization bypass in RAG syncSame vendor · NVD/CVE Database
• HighCVE-2026-107685: req_llm WebSocket streaming client accepts untrusted TLS certificatesSame vendor · NVD/CVE Database