Researchers have discovered serious vulnerabilities in Moodle, a popular open-source learning platform used by many top universities in the United States, the United Kingdom and other countries around the world.
Moodle updates released last week address a total of four vulnerabilities, including two that have a “serious” severity/risk rating and two classified as “minor.”
One of the serious flaws, tracked as CVE-2017-2641, can be exploited by an attacker to execute arbitrary PHP code on a vulnerable Moodle server. While the security hole is tracked as a single issue, Netanel Rubin, the expert who discovered the problem, says an attack is possible due to several smaller weaknesses affecting the platform.
The flaw can be exploited by an authenticated attacker to conduct an SQL injection attack via user preferences and add a new administrator user to the system. Once the attacker has an admin account on the system, they can execute arbitrary code by uploading a new plugin or a template to the server, Rubin said.
This vulnerability affects Moodle 3.2 to 3.2.1, 3.1 to 3.1.4, 3.0 to 3.0.8, 2.7.0 to 2.7.18 and older versions, and it has been fixed with the release of versions 3.2.2, 3.1.5, 3.0.9 and 2.7.19.
Moodle developers noted that the flaw can only be exploited in Moodle versions prior to 3.2 by users with manager or admin rights. In version 3.2, the attack works with any type of user account, including teacher and student accounts.
The second serious flaw patched by Moodle last week has been described as a cross-site scripting (XSS) flaw in the functionality that allows users to attach files for evidence of prior learning.
“Serving files attached to evidence of prior learning did not force download. When viewed by other users they would be opened in current moodle sessions,” Moodle wrote in its advisory.
The flaw, reported by a researcher who uses the online moniker “wez3” and tracked as CVE-2017-2645, only affects versions 3.2 to 3.2.1 and 3.1 to 3.1.4.
Jaymark Pestaño discovered a less severe XSS vulnerability in the evidence of prior learning section, and Nadav Kavalerchik identified an issue related to usernames being displayed in global search for unauthenticated users. Both issues are considered minor.