OpenAI Codex GUI argument injection allows remote code execution
Original title: ZDI-26-968: (Pwn2Own) OpenAI Codex GUI OpenProjectInCodex Argument Injection Remote Code Execution Vulnerability
CVE-2026-110930, disclosed from Pwn2Own as ZDI-26-968, is scored CVSS 7.8.
ZDI published an argument-injection flaw in the OpenProjectInCodex function of the OpenAI Codex GUI. Remote attackers can execute arbitrary code if the target opens a malicious file. The issue is CVE-2026-110930, labeled as found at Pwn2Own, with a CVSS rating of 7.8.
Unconfirmed: The advisory does not state a fixed version or whether attackers have used the flaw outside Pwn2Own.
- The vulnerable function is OpenProjectInCodex.
- The user must open a malicious file.
- No affected or fixed Codex version is stated.
Impact: Arbitrary code execution after opening a malicious file
| Product | Affected |
|---|---|
| OpenAI Codex | — |
- CVEs
- CVE-2026-110930
- ATT&CK
- T1204.002 Malicious File
- Vendors
- OpenAI
- Products
- Codex
- Organizations
- Zero Day Initiative
Vulnerabilities mentioned
All| CVE | Vulnerability | CVSS | EPSS | Signals | Published |
|---|---|---|---|---|---|
| CVE-2026-110930 | Command Injection in Codex Desktop for Windows via project paths | 8.0 | — |
From the original article
This vulnerability allows remote attackers to execute arbitrary code on affected installations of OpenAI Codex. User interaction is required to exploit this vulnerability in that the target must open a malicious file. The ZDI has assigned a CVSS rating of 7.
zerodayinitiative.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.