FastGPT Information Disclosure via Request ID Exposure

Executive Summary

This vulnerability exists in FastGPT versions prior to 4.15.0. The GET /api/core/ai/record/getRecord endpoint authenticates users but does not properly restrict access to data by team. It loads large language model (LLM) request and response traces based solely on a requestId without verifying if the requesting user belongs to the same team as the data. As a result, any authenticated user who knows a valid requestId can access another team's prompts, retrieved RAG chunks, and completions.

This issue was fixed in version 4.15.0 by adding proper team scoping to the data retrieval process.

Useful 0 Not Useful 0

Impact Analysis

This vulnerability can lead to unauthorized disclosure of sensitive information between teams using the FastGPT platform. An authenticated user could access confidential prompts, retrieved data chunks, and AI completions belonging to other teams if they know the requestId.

Such unauthorized access could result in data leakage, loss of confidentiality, and potential exposure of proprietary or sensitive business information.

Useful 0 Not Useful 0

Mitigation Strategies

To mitigate this vulnerability, upgrade FastGPT to version 4.15.0 or later, where the issue has been fixed.

Useful 0 Not Useful 0

Compliance Impact

The vulnerability allows any authenticated user to access another team's prompts, retrieved RAG chunks, and completions if the requestId is known, leading to unauthorized data exposure across teams.

Such unauthorized access to potentially sensitive information could lead to non-compliance with data protection regulations like GDPR and HIPAA, which require strict access controls and data confidentiality.

Useful 0 Not Useful 0