High

privategpt

Denial of Service via Multipart Boundary Manipulation

A Denial of Service (DOS) vulnerability was identified in version 0.5.0 of privateGPT. The issue arises when an attacker appends a large number of characters to the end of a multipart boundary during file upload, causing the system to become unresponsive. This vulnerability has not yet been patched.

Available publicly on Sep 29 2024

7.5

CVSS:

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

Credit:

mnqazi
Threat Overview

The vulnerability allows an attacker to exploit the multipart boundary handling mechanism in privateGPT. By appending a large number of characters to the end of a multipart boundary, the system is forced to process each character individually, leading to excessive resource consumption. This results in the service becoming unavailable for an extended period, effectively causing a Denial of Service (DOS) attack. The severity of the attack is proportional to the number of characters appended, with the potential to render the service inaccessible indefinitely.

Attack Scenario

An attacker could exploit this vulnerability by intercepting a file upload request to privateGPT using a tool like BurpSuite. They would then append a large number of characters to the end of the multipart boundary in the request and send it to the server. The server would then become unresponsive as it processes each character, leading to a prolonged or indefinite denial of service.

Who is affected

Users and organizations relying on privateGPT version 0.5.0 for their operations are affected by this vulnerability. This includes any deployment of privateGPT that allows file uploads via multipart forms.

Technical Report
Want more out of Sightline?

Sightline offers even more for premium customers

Go Premium

We have - related security advisories that are available with Sightline Premium.