Description

Subscriber Local File Inclusion in MagicAI for WordPress - AI Text, Image, Chat, Code, and Voice Generator <= 1.4 versions.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "8.8",
  "pubDate": "2026-08-24T22:16:52.027Z",
  "pubdate": "2026-08-24T22:16:52.027Z",
  "executiveSummary": "A Local File Inclusion vulnerability has been identified in the MagicAI for WordPress - AI Text, Image, Chat, Code, and Voice Generator plugin affecting versions <= 1.4. This security flaw enables authenticated users with subscriber-level privileges or higher to manipulate input parameters that are passed directly into file inclusion routines without adequate sanitization or validation.\nSuccessful exploitation of this vulnerability allows an attacker to include arbitrary local files residing on the underlying web server filesystem. Depending on the configuration and file permissions of the hosting environment, this may lead to sensitive data exposure, including configuration files, system credentials, and application source code.\nThe risk implication is significant as it lowers the barrier for low-privileged accounts to compromise internal system data or potentially escalate privileges if combined with other attack vectors such as log poisoning or arbitrary file read scenarios.\nAttack capabilities are restricted to users holding valid subscriber accounts within the WordPress installation, necessitating authentication unless registration is open to the public. Remediation requires updating the affected software to a patched version once released or applying strict input validation controls within the vulnerable component.",
  "technicalDetails": "The vulnerability stems from insecure handling of input parameters within the MagicAI for WordPress - AI Text, Image, Chat, Code, and Voice Generator plugin <= 1.4. The root cause is identified as improper neutralization of special elements used in a file inclusion operation, commonly associated with Local File Inclusion (LFI).\nThe vulnerable component fails to implement strict allow-listing or canonicalization checks on user-supplied input before passing the data to inclusion functions. This architectural oversight allows an attacker to inject directory traversal sequences, such as dot-dot-slash (../) patterns, to navigate outside the intended application directory structure and access sensitive system files.\nThe attack flow proceeds as follows: First, an authenticated attacker with subscriber-level privileges crafts a malicious HTTP request targeting the vulnerable endpoint within the plugin. Second, the request incorporates path traversal strings designed to reference target files on the server filesystem, such as /etc/passwd or application-specific configuration files containing database credentials.\nThird, the vulnerable PHP script processes the parameter and executes the file inclusion function using the tainted path. Fourth, the server reads the target file contents and includes them within the HTTP response or processes them, depending on the specific PHP include or require function invoked.\nAuthentication and privilege requirements involve holding at least a subscriber-level role on the target WordPress site, meaning the attack is authenticated rather than unauthenticated. Network exposure is standard HTTP/HTTPS accessibility to the WordPress application frontend or backend endpoints exposed by the plugin. Payload behavior focuses on reading arbitrary local files rather than direct remote code execution, though inclusion of dynamically generated log files containing executable PHP code could theoretically lead to code execution under specific server configurations. Post-exploitation impact encompasses unauthorized data disclosure, reconnaissance of internal server architecture, and potential access to sensitive credentials stored in configuration files."
}