CVE-2026-14898

Published Jul 6, 2026 CVSS 3.1 · 6.5 Medium EPSS < 1% KEV · Not listed
Description
The OpenAI Codex desktop app for macOS rendered remote images from Markdown in model responses. An attacker who could place an indirect prompt injection in content processed by Codex, such as a connected-tool result or another untrusted source, could induce the model to construct a remote image URL containing sensitive data. The app automatically fetched that URL when rendering the response, sending the embedded data to an attacker-controlled server without a separate user click. Successful exploitation could exfiltrate secrets and other information accessible in the Codex session, including API keys, source code, and data returned by connected tools. No direct integrity or availability impact was demonstrated, and there is no known exploitation in the wild.

Remediation

Vendor Solution

Upgrade to 26.527.31326 or later. The product fix disables loading remote images from Markdown.


Vendor Workaround

If an upgrade is not possible, avoid processing untrusted content and connected-tool data in affected versions.

Tracking

Sign in to view the affected projects.

Advisories

No advisories available
References
Link Providers
https://openai.com/codex/ cve-icon cve-icon
History
Type Values Removed Values Added
Title Remote Image Loading Enables Sensitive Data Exfiltration in Codex Desktop App
Type Values Removed Values Added
Title Remote Image Loading Enables Sensitive Data Exfiltration in Codex Desktop App
Type Values Removed Values Added
Title Remote Image Loading Exposes Sensitive Data in OpenAI Codex Desktop App
Type Values Removed Values Added
Title Remote Image Loading Exposes Sensitive Data in OpenAI Codex Desktop App
Type Values Removed Values Added
Title Remote image loading allows exfiltration of sensitive session data in OpenAI Codex desktop app
Type Values Removed Values Added
Title Remote image loading allows exfiltration of sensitive session data in OpenAI Codex desktop app
Type Values Removed Values Added
Title Exfiltration of Sensitive Data via Remote Image Loading from Untrusted Markdown in Codex Desktop
Type Values Removed Values Added
Title Exfiltration of Sensitive Data via Remote Image Loading from Untrusted Markdown in Codex Desktop
Type Values Removed Values Added
Title Exfiltration of Sensitive Data via Remote Image Loading in OpenAI Codex Desktop App
Type Values Removed Values Added
Title Exfiltration of Sensitive Data via Remote Image Loading in OpenAI Codex Desktop App
Type Values Removed Values Added
Title Remote image loading in Codex desktop leads to data exfiltration
Type Values Removed Values Added
Title Remote image loading in Codex desktop leads to data exfiltration
Type Values Removed Values Added
Title Remote Image Loading Enables Data Exfiltration in OpenAI Codex Desktop App
Type Values Removed Values Added
Title Remote Image Loading Enables Data Exfiltration in OpenAI Codex Desktop App
Type Values Removed Values Added
Title Exfiltration via Remote Image Loading in OpenAI Codex Desktop App
Type Values Removed Values Added
Metrics cvssV3_1

{'score': 6.5, 'vector': 'CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:N/A:N'}

ssvc

{'options': {'Automatable': 'no', 'Exploitation': 'none', 'Technical Impact': 'partial'}, 'version': '2.0.3'}

Type Values Removed Values Added
Title Exfiltration via Remote Image Loading in OpenAI Codex Desktop App
Type Values Removed Values Added
First Time appeared Openai
Openai codex Desktop App For Macos
Vendors & Products Openai
Openai codex Desktop App For Macos
Type Values Removed Values Added
Description The OpenAI Codex desktop app for macOS rendered remote images from Markdown in model responses. An attacker who could place an indirect prompt injection in content processed by Codex, such as a connected-tool result or another untrusted source, could induce the model to construct a remote image URL containing sensitive data. The app automatically fetched that URL when rendering the response, sending the embedded data to an attacker-controlled server without a separate user click. Successful exploitation could exfiltrate secrets and other information accessible in the Codex session, including API keys, source code, and data returned by connected tools. No direct integrity or availability impact was demonstrated, and there is no known exploitation in the wild.
Weaknesses CWE-200
References
Weaknesses

Sources

Status
PUBLISHED
Assigner
OAI
Published
Updated
Jul 7, 2026, 3:35 PM
Reserved
Jul 6, 2026, 7:35 PM
Status
Awaiting Analysis
Published
Jul 6, 2026, 8:16 PM
Modified
Oct 25, 2026, 8:10 PM