CVE-2024-27564: Dirk1983 Chatgpt SSRF Vulnerability

CVE-2024-27564 Overview

A Server-Side Request Forgery (SSRF) vulnerability exists in the pictureproxy.php file within the dirk1983 ChatGPT web application (mm1.ltd source code, commit f9f4bbc). The vulnerability allows remote attackers to forge server-side HTTP requests by manipulating the url parameter, potentially enabling access to internal network resources, data exfiltration, or attacks against other services.

Critical Impact

This SSRF vulnerability allows unauthenticated attackers to abuse the server as a proxy to reach internal services, scan internal networks, or exfiltrate sensitive data through crafted URL parameters.

Affected Products

• dirk1983 ChatGPT (version 2023-05-23)
• mm1.ltd deployments using pictureproxy.php from commit f9f4bbc
• Any self-hosted instances of the dirk1983/chatgpt repository containing the vulnerable component

Discovery Timeline

• 2024-03-05 - CVE-2024-27564 published to NVD
• 2025-03-20 - Last updated in NVD database

Technical Details for CVE-2024-27564

Vulnerability Analysis

The vulnerability resides in pictureproxy.php, a PHP script designed to proxy image requests for the ChatGPT web interface. The script accepts a user-controlled url parameter and makes HTTP requests to the specified URL without proper validation or sanitization. This design flaw allows attackers to craft malicious requests that cause the server to fetch arbitrary URLs, including those pointing to internal network resources that would otherwise be inaccessible from external networks.

SSRF vulnerabilities of this nature can be leveraged to bypass firewall restrictions, access cloud metadata services (such as AWS EC2 metadata at 169.254.169.254), enumerate internal services, and potentially pivot to more sensitive systems within the target infrastructure. The network-accessible attack vector combined with no authentication requirements makes this vulnerability particularly dangerous in internet-facing deployments.

Root Cause

The root cause of CVE-2024-27564 is the absence of URL validation and allowlist enforcement in the pictureproxy.php script. The vulnerable code directly processes the url parameter without verifying that the requested resource is a legitimate external image URL. There is no protection against requests to internal IP ranges (RFC 1918 addresses), localhost, or cloud metadata endpoints. This lack of input validation (CWE-918: Server-Side Request Forgery) allows attackers to abuse the proxy functionality for unintended purposes.

Attack Vector

The attack can be executed over the network without authentication. An attacker sends a crafted HTTP request to the pictureproxy.php endpoint with a malicious url parameter value pointing to an internal resource or service. The vulnerable server then makes a request to the specified URL and returns the response to the attacker, effectively acting as an open proxy.

Typical exploitation scenarios include:

• Accessing internal services by specifying internal IP addresses or hostnames
• Retrieving cloud instance metadata from providers like AWS, GCP, or Azure
• Port scanning internal network infrastructure
• Bypassing IP-based access controls on internal applications

The vulnerability can be exploited by simply modifying the url parameter in a request to pictureproxy.php. For example, an attacker could target internal resources like http://localhost/admin or cloud metadata services. Technical details about the vulnerable code are documented in the GitHub Issue Report.

Detection Methods for CVE-2024-27564

Indicators of Compromise

• HTTP requests to pictureproxy.php containing internal IP addresses (10.x.x.x, 172.16-31.x.x, 192.168.x.x) or localhost in the url parameter
• Requests targeting cloud metadata endpoints such as 169.254.169.254 or metadata.google.internal through the proxy endpoint
• Unusual outbound connections from the web server to internal network segments not typically accessed by the application
• Log entries showing pictureproxy.php requests with non-image URL targets or suspicious port numbers

Detection Strategies

• Implement Web Application Firewall (WAF) rules to inspect and block requests to pictureproxy.php containing internal IP ranges or metadata service addresses
• Monitor server-side request logs for requests to pictureproxy.php with url parameters pointing to non-standard destinations
• Deploy network segmentation monitoring to detect unexpected outbound connections from web servers to internal infrastructure
• Use intrusion detection systems (IDS) to identify SSRF attack patterns in HTTP traffic

Monitoring Recommendations

• Enable detailed logging for all requests to pictureproxy.php, including the full url parameter value and response status codes
• Configure alerts for any outbound connections from the web server to internal network ranges or cloud metadata IP addresses
• Implement request rate limiting on the pictureproxy.php endpoint to reduce the effectiveness of automated scanning attempts

How to Mitigate CVE-2024-27564

Immediate Actions Required

• Remove or disable the pictureproxy.php script if image proxying functionality is not essential to your deployment
• Implement strict URL validation to only allow requests to trusted external domains via an allowlist approach
• Block requests containing internal IP addresses, localhost references, and cloud metadata endpoints at the application or WAF level
• Review server logs for evidence of exploitation attempts and investigate any suspicious access patterns

Patch Information

As of the last NVD update on 2025-03-20, users should monitor the GitHub Issue Report for updates from the maintainer regarding a fix. The vulnerability was identified in commit f9f4bbc of the dirk1983/chatgpt repository. Deployers should check for newer commits that address the SSRF vulnerability or implement the recommended workarounds below.

Workarounds

• Implement an allowlist of permitted domains for the url parameter, rejecting any requests that do not match trusted patterns
• Add URL parsing and validation to reject requests to private IP ranges (10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, 127.0.0.0/8) and cloud metadata addresses (169.254.169.254)
• Deploy network-level controls to prevent the web server from making connections to internal network resources
• Use a security-focused HTTP library or framework that provides built-in SSRF protections

# Example: Block SSRF requests using Apache mod_rewrite
# Add to .htaccess or Apache configuration
<IfModule mod_rewrite.c>
    RewriteEngine On
    # Block requests to pictureproxy.php with internal/metadata URLs
    RewriteCond %{QUERY_STRING} url=.*(127\.0\.0\.1|localhost|10\.|172\.(1[6-9]|2[0-9]|3[01])\.|192\.168\.|169\.254\.|metadata) [NC]
    RewriteRule ^pictureproxy\.php$ - [F,L]
</IfModule>