/

CVE-2026-50775

Severity:

CRITICAL RISK

(9.8)

Blind SSRF (CWE-918) in DataHub Image Retrieval Path

CVE-2026-50775

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Blind SSRF and Potential Remote Code Execution

Severity:

CRITICAL RISK

(9.8)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

A blind server-side request forgery vulnerability in DataHub allows a remote attacker to cause the server to retrieve an image from a crafted URL without exposing the fetched content or errors directly.

Under deployments where the server can reach sensitive internal services or privileged endpoints, this behavior may enable internal network access and could contribute to arbitrary code execution.

Affected context

The vulnerability occurs in DataHub's image-retrieval functionality when externally influenced URL values are fetched by the server. Risk depends on whether unauthenticated or low-privileged remote users can reach the relevant endpoint and whether the DataHub server has network access to internal services, metadata endpoints, or administrative interfaces.

Exploit preconditions

  • Attack surface: A remotely reachable DataHub image-fetching or image-processing endpoint
  • Required access: Access sufficient to submit or influence an image URL; authentication requirements depend on deployment configuration
  • Trigger point: Server-side retrieval of the attacker-controlled image URL
  • Why exploitation is possible: The server performs outbound requests based on insufficiently trusted URL input, potentially reaching destinations that are inaccessible to the attacker directly

Why this happens

The underlying bug is server-side request forgery, mapped to CWE-918. An image retrieval component accepts attacker-influenced URLs and causes the application server to make outbound requests without adequately constraining destinations, protocols, or network reachability. Because the response is blind, content and errors may not be returned directly, but request side effects and access to internal services can still create security impact.

Potential impact

  • The server may access internal or cloud-provider metadata services
  • Attackers could probe internal network services through request behavior
  • Internal administrative endpoints may be reached from a trusted network location
  • Credentials or sensitive information could be exposed indirectly
  • Compromised internal services could enable arbitrary code execution
  • The vulnerability could support lateral movement or broader infrastructure compromise

General mitigation guidance

Mitigation generally focuses on reducing trust in user-supplied URLs and limiting the server's outbound request capabilities.

Common defensive patterns include:

  • Allowing only approved image sources through strict allowlists
  • Restricting accepted schemes, ports, hostnames, and redirect destinations
  • Resolving hostnames and validating final destinations against private, loopback, link-local, and other restricted address ranges
  • Applying egress network policies that prevent unnecessary access to internal services and metadata endpoints
  • Using a dedicated low-privilege image-fetching service with isolated credentials and network permissions
  • Limiting redirects, response sizes, connection duration, and request frequency
  • Avoiding exposure of fetched content, response details, or timing information where not required

How teams detect issues like this

Detection approaches may include:

  • Static analysis for outbound HTTP or image-fetch APIs that consume request parameters without URL validation
  • Code review checks for missing scheme, hostname, port, redirect, and private-address restrictions
  • Dynamic testing with controlled callback endpoints and blocked internal-address test cases
  • Egress monitoring for unexpected requests from DataHub to internal ranges, metadata services, or unusual external hosts
  • Policy enforcement requiring centralized URL validation and approved-domain controls for server-side fetch features
  • Automated security review tools that flag SSRF-prone data flows from remote input to network clients

Sources

SourceLink
datahub.comdatahub.com/
github.comgithub.com/Henkel-CyberVM/CVEs/tree/main/CVE-2026-50775

Frequently asked

questions

Yes, it may be remotely exploitable when an attacker can reach the affected image-retrieval functionality and influence the URL submitted to the server.

No. The server makes the outbound request, so the attacker may only need access to the DataHub endpoint and a URL accepted by its image-fetching logic.

Use an allowlist-based image retrieval design, validate URLs and redirect destinations, block private and metadata address ranges, and restrict outbound network access.

Treat all externally supplied URLs as untrusted, centralize SSRF protections, enforce egress policies, and review server-side fetch functionality with both static and dynamic security testing.

Find out what's already
exploitable in your codebase.