mirror of
https://github.com/deepseek-ai/DeepSeek-V3.git
synced 2026-11-05 00:47:48 +00:00
[GH-ISSUE #1306] Cross-Site Scripting (XSS) via AI-Generated HTML Code Execution #850
Labels
No labels
closed-as-stale
enhancement
question
spam
stale
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
deepseek-ai/DeepSeek-V3#850
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @ev1lclow3n on GitHub (May 14, 2026).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-V3/issues/1306
Describe the bug
A Cross-Site Scripting (XSS) vulnerability was identified in the AI-generated HTML execution environment. The platform reproduces user-supplied HTML and JavaScript content without proper sanitization and executes it through the built-in “Run” functionality. By providing crafted HTML containing JavaScript code and instructing the AI model to repeat the content, arbitrary JavaScript execution was successfully achieved inside the embedded runtime environment.
To Reproduce
Open a chat session on the platform.
Provide HTML code containing a JavaScript payload such as alert("XSS").
Ask the AI model to reproduce or repeat the supplied HTML code exactly.
Execute the generated HTML using the “Run” functionality.
Observe that the JavaScript payload executes successfully inside the embedded page.
JS Payload Embedded in HTML File
Expected behavior
User-controlled HTML and JavaScript content should be properly sanitized or executed inside a strictly sandboxed environment where arbitrary script execution is prevented.
Screenshots
Additional context
The payload executed from the embedded runtime domain associated with the HTML execution environment after clicking the “Run” functionality. This behavior indicates insufficient sanitization and/or sandboxing of AI-generated HTML content before execution. The issue may allow attackers to execute arbitrary JavaScript, perform phishing attacks, manipulate rendered content, or conduct further client-side attacks against users interacting with generated previews.
#CWE-79: Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting')
@github-actions[bot] commented on GitHub (Jun 14, 2026):
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!
@github-actions[bot] commented on GitHub (Jul 11, 2026):
false