Skip to main content
Vulnerability Database/CVE-2024-10908

CVE-2024-10908: Lm-sys Fastchat Open Redirect Vulnerability

CVE-2024-10908 is an open redirect flaw in Lm-sys Fastchat Release v0.2.36 that enables attackers to redirect users to malicious sites for phishing and credential theft. This article covers technical details, exploitation risks, and mitigation.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2024-10908 Overview

CVE-2024-10908 is an open redirect vulnerability [CWE-601] in lm-sys/fastchat release v0.2.36. The flaw allows a remote unauthenticated attacker to redirect users to arbitrary external websites through a specially crafted URL. FastChat is an open platform for training, serving, and evaluating large language model (LLM) based chatbots, widely used in AI research environments.

Attackers can abuse the trusted FastChat domain to launch phishing campaigns, distribute malware, or harvest credentials. Because the redirect originates from a legitimate host, users and security tools often treat the initial link as benign.

Critical Impact

Unauthenticated attackers can craft URLs that appear to originate from a trusted FastChat deployment but redirect victims to attacker-controlled sites for phishing, malware delivery, or credential theft.

Affected Products

  • lm-sys fastchat version 0.2.36
  • FastChat web interface components handling redirect parameters
  • Deployments exposing the FastChat server to untrusted networks

Discovery Timeline

  • 2025-03-20 - CVE-2024-10908 published to NVD
  • 2026-06-17 - Last updated in NVD database

Technical Details for CVE-2024-10908

Vulnerability Analysis

The vulnerability is classified as an open redirect [CWE-601]. FastChat accepts a user-controlled URL parameter and uses it as the target of an HTTP redirect without validating that the destination belongs to an approved domain or path. The attack requires user interaction, typically a click on a crafted link, and executes with no authentication.

The impact crosses a trust boundary: the redirect appears to come from the FastChat origin but lands on an attacker-controlled host. This changes the security scope by influencing resources beyond the vulnerable application, which is consistent with the published CVSS scope change.

Root Cause

The root cause is missing or insufficient validation of a redirect target supplied through a URL parameter. The application forwards the value directly into a Location header or client-side navigation without enforcing an allowlist of trusted destinations. Any absolute URL submitted by the attacker becomes a valid redirect target.

Attack Vector

An attacker constructs a URL that points to the vulnerable FastChat endpoint and embeds a malicious destination in the redirect parameter. The attacker delivers the link through email, chat, or a malicious web page. When the victim clicks the link, FastChat issues a redirect to the attacker-controlled site, which may host a cloned login page or a drive-by malware payload.

For technical specifics, see the Huntr Bug Bounty Report.

Detection Methods for CVE-2024-10908

Indicators of Compromise

  • HTTP 3xx responses from FastChat endpoints containing Location headers that point to external, unexpected domains.
  • Access log entries showing redirect parameters that contain absolute URLs or encoded http:// or https:// schemes.
  • Referrer logs on external services showing inbound traffic originating from a FastChat host.

Detection Strategies

  • Inspect web server and reverse proxy logs for query parameters carrying full URLs and correlate with 302 or 303 response codes.
  • Deploy web application firewall (WAF) rules that flag redirect parameters resolving to domains outside the organization's allowlist.
  • Hunt for user-reported phishing links that share the FastChat hostname but terminate on third-party domains.

Monitoring Recommendations

  • Alert on bursts of redirect traffic from the FastChat service, which may indicate an active phishing campaign.
  • Monitor DNS and proxy telemetry for outbound traffic to newly registered domains referenced in FastChat redirects.
  • Track user-agent and geolocation anomalies on requests touching known redirect endpoints.

How to Mitigate CVE-2024-10908

Immediate Actions Required

  • Upgrade lm-sys/fastchat beyond release v0.2.36 once a patched version is published by the maintainers.
  • Restrict public exposure of FastChat deployments by placing them behind authenticated reverse proxies or VPNs.
  • Educate users about phishing links that may originate from internal FastChat URLs.

Patch Information

No vendor advisory URL is listed in the NVD record at the time of publication. Review the project repository and the Huntr Bug Bounty Report for the latest remediation guidance and release notes. EPSS data for this CVE is 0.784% at the 55.8 percentile as of 2026-10-26.

Workarounds

  • Enforce a server-side allowlist of permitted redirect destinations and reject any URL that does not match.
  • Replace absolute URL redirects with indirect references, such as numeric IDs mapped server-side to known-good destinations.
  • Add an interstitial warning page that notifies users before leaving the FastChat domain.
  • Strip or validate query parameters at a reverse proxy or WAF layer when the application cannot be patched immediately.
bash
# Example NGINX rule to block external absolute URLs in redirect parameters
if ($arg_redirect ~* "^(https?:)?//") {
    return 400;
}

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.