Remote Code Execution Vulnerability with Jupyter notebook markdown rendering in untrusted workspaces

VS Code - Remote Code Execution Vulnerability

A remote code execution vulnerability exists in VS Code 1.119.0 and earlier versions with the webview used by Jupyter notebooks. This could bypass rendered markdown sanitization in untrusted workspaces to run untrusted content inside the webview

The root cause is in an incorrect buffer being used for the internal protocol webviews use to load the VS Code controlled root webview content

Patches

The fix is available starting with VS Code 1.119.1. The fix (https://github.com/microsoft/vscode/commit/1dbe28533fe4204dcd3c3e30a05e22c6ba307145) mitigates this attack by making sure the correctly sized buffer is passed buffer to the webview protocol provider

Workarounds

Do not open notebook files from untrusted sources

References

• The patch for this can be found at https://github.com/microsoft/vscode/commit/1dbe28533fe4204dcd3c3e30a05e22c6ba307145
• MSRC details for this can be found at https://msrc.microsoft.com/update-guide/vulnerability/CVE-2026-41610