TorchServe is a tool for serving and scaling PyTorch models in production. TorchServe default configuration lacks proper input validation, enabling third parties to invoke remote HTTP download requests and write files to the disk. This issue could be taken advantage of to compromise the integrity of the system and sensitive data. This issue is present in versions 0.1.0 to 0.8.1. A user is able to load the model of their choice from any URL that they would like to use. The user of TorchServe is responsible for configuring both the allowed_urls and specifying the model URL to be used. A pull request to warn the user when the default value for allowed_urls is used has been merged in PR #2534. TorchServe release 0.8.2 includes this change. Users are advised to upgrade. There are no known workarounds for this issue.

PyTorch Serve version prior to 0.8.2 and versions using a default configuration are vulnerable to a Server-Side Request Forgery allowing an unauthenticated user to write a file to disk, that can lead to an arbitrary code execution.

The remote host contains a torchserve version that is prior to  0.8.2. It is, therefore, affected by a Server Side Request Forgery vulnerability.  TorchServe default configuration lacks proper input validation, enabling third parties to invoke remote HTTP download requests and write files to the disk. This issue could be taken advantage of to compromise the integrity of the system and sensitive data. This issue is present in versions 0.1.0 to 0.8.1. A remote attacker can exploit this and load the malicious model of their choice from any URL.

Note that Nessus has not tested for this issue but has instead relied only on the application's self-reported version number.

The remote host contains a torchserve version that is prior to  0.8.2. It is, therefore, affected by a Server Side Request Forgery vulnerability.  TorchServe default configuration lacks proper input validation, enabling third parties to invoke remote HTTP download requests and write files to the disk. This issue could be taken advantage of to compromise the integrity of the system and sensitive data. This issue is present in versions 0.1.0 to 0.8.1. A remote attacker can exploit this and load the malicious model of their choice from any URL.

ID	Name	Product	Family	Severity
114228	PyTorch Serve Server-Side Request Forgery	Web App Scanning	Component Vulnerability	critical
184081	PyTorch TorchServe < 0.8.2 SSRF	Nessus	Artificial Intelligence	critical
184080	PyTorch TorchServe SSRF (CVE-2023-43654)	Nessus	Artificial Intelligence	critical

Detections

Analytics

CVE-2023-43654

critical

Tenable Plugins

critical

critical

critical