Executive brief
gpt-researcher is an AI-powered autonomous agent used to conduct deep research by scraping and analyzing web data. A security vulnerability in its web socket endpoint allows attackers to manipulate the source URLs used for research. This could lead to a server-side request forgery (SSRF), potentially allowing an attacker to use the server to access internal resources or disrupt operations.
Technical details
A Server-Side Request Forgery (SSRF) vulnerability exists in the 'ws Endpoint' component of gpt-researcher through version 3.4.3. The vulnerability is rooted in the insufficient validation of the 'source_urls' argument, which can be manipulated by a remote attacker. By providing malicious URLs, an attacker can force the server to make outbound requests to unintended destinations, including internal network services that are not publicly accessible. This is a network-based attack that requires no authentication or user interaction. While a proof-of-concept has been publicly disclosed, the maintainers have marked the issue as 'wontfix' as of the latest advisory updates.
Affected products
- assafelovic gpt-researcher up to 3.4.3
Timeline
- 2026-03-23: disclosed: Issue reported on GitHub repository
- 2026-04-06: advisory: NVD/VulDB publication date