Description

Subscriber Arbitrary File Download in AI Hub <= 1.3.10 versions.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "7.7",
  "pubDate": "2026-08-13T14:17:04.320Z",
  "pubdate": "2026-08-13T14:17:04.320Z",
  "executiveSummary": "An arbitrary file download vulnerability has been identified in the AI Hub product affecting versions 1.3.10 and below. This security flaw enables authenticated users with subscriber-level privileges to retrieve sensitive files from the underlying server hosting the vulnerable application. The vulnerability poses a significant risk to confidentiality, potentially exposing system configuration files, source code, credentials, or other sensitive data residing within the file system scope accessible to the web application process. Exploitation of this vulnerability requires subscriber-level authentication, meaning an attacker must possess a valid, low-privileged user account within the targeted AI Hub instance. The attack flow involves manipulating input parameters related to file retrieval functionalities, bypassing insufficient path validation controls, and inducing the application to stream arbitrary server-side files back to the client. Remediation requires updating the AI Hub product to a patched version once available, implementing strict input sanitization, enforcing robust path traversal checks, and restricting file read operations to designated safe directories.",
  "technicalDetails": "The vulnerability resides in the file handling and retrieval mechanism of the AI Hub application, specifically within versions 1.3.10 and prior. The root cause stems from insufficient input validation and a lack of proper path normalization when processing parameters that dictate which files to read or download. Because the application fails to adequately sanitize user-supplied input, it is susceptible to path traversal techniques, allowing an attacker to traverse the directory structure outside of the intended web root or repository boundaries.\nAuthentication is required to exploit this vulnerability, specifically requiring at least a subscriber-level privilege within the AI Hub platform. Although the attack surface is restricted to authenticated users, the privilege boundary is crossing from a standard low-privileged subscriber into unauthorized data access of arbitrary server files. The vulnerability can be exploited over the network via standard HTTP requests interacting with the vulnerable endpoint responsible for file processing.\nThe step-by-step attack flow typically proceeds as follows: First, the attacker authenticates to the AI Hub application using a valid subscriber account to obtain session cookies or necessary authorization tokens. Second, the attacker identifies the vulnerable HTTP endpoint or function responsible for handling file downloads. Third, the attacker crafts a malicious HTTP request containing path traversal sequences (such as dot-dot-slash patterns) or absolute file paths directed at sensitive operating system files or application configuration files (e.g., database credentials, environment files). Fourth, the vulnerable component processes the unsanitized input without verifying whether the requested path falls within the permitted directory. Finally, the server reads the targeted file from the disk and includes its contents in the HTTP response body, granting the attacker unauthorized access to sensitive system data.\nThe post-exploitation impact includes the potential exposure of sensitive internal data, API keys, database connection strings, and system configuration parameters. Such information disclosure can facilitate further attacks, including lateral movement, privilege escalation, or complete compromise of the hosting infrastructure."
}