In GLPI before version 9.5.2, the `​pluginimage.send.php​` endpoint allows a user to specify an image from a plugin. The parameters can be maliciously crafted to instead delete the .htaccess file for the files directory. Any user becomes able to read all the files and folders contained in “/files/”. Some of the sensitive information that is compromised are the user sessions, logs, and more. An attacker would be able to get the Administrators session token and use that to authenticate. The issue is patched in version 9.5.2.

In GLPI before version 9.5.0, the encryption algorithm used is insecure. The security of the data encrypted relies on the password used, if a user sets a weak/predictable password, an attacker could decrypt data. This is fixed in version 9.5.0 by using a more secure encryption library. The library chosen is sodium.

In glpi before 9.5.1, there is a SQL injection for all usages of "Clone" feature. This has been fixed in 9.5.1.

In GLPI after 0.68.1 and before 9.4.6, multiple reflexive XSS occur in Dropdown endpoints due to an invalid Content-Type. This has been fixed in version 9.4.6.

In GLPI before 9.4.6, an attacker can execute system commands by abusing the backup functionality. Theoretically, this vulnerability can be exploited by an attacker without a valid account by using a CSRF. Due to the difficulty of the exploitation, the attack is only conceivable by an account having Maintenance privileges and the right to add WIFI networks. This is fixed in version 9.4.6.

GLPI before before version 9.4.6 has a vulnerability involving a default encryption key. GLPIKEY is public and is used on every instance. This means anyone can decrypt sensitive data stored using this key. It is possible to change the key before installing GLPI. But on existing instances, data must be reencrypted with the new key. Problem is we can not know which columns or rows in the database are using that; espcially from plugins. Changing the key without updating data would lend in bad password sent from glpi; but storing them again from the UI will work.

In GLPI before version 9.4.6 there are multiple related stored XSS vulnerabilities. The package is vulnerable to Stored XSS in the comments of items in the Knowledge base. Adding a comment with content "<script>alert(1)</script>" reproduces the attack. This can be exploited by a user with administrator privileges in the User-Agent field. It can also be exploited by an outside party through the following steps: 1. Create a user with the surname `" onmouseover="alert(document.cookie)` and an empty first name. 2. With this user, create a ticket 3. As an administrator (or other privileged user) open the created ticket 4. On the "last update" field, put your mouse on the name of the user 5. The XSS fires This is fixed in version 9.4.6.

In GLPI before version 9.4.6, there is a vulnerability that allows bypassing the open redirect protection based which is based on a regexp. This is fixed in version 9.4.6.

GLPI through 9.4.3 is prone to account takeover by abusing the ajax/autocompletion.php autocompletion feature. The lack of correct validation leads to recovery of the token generated via the password reset functionality, and thus an authenticated attacker can set an arbitrary password for any user. This vulnerability can be exploited to take control of admin account. This vulnerability could be also abused to obtain other sensitive fields like API keys or password hashes.

An issue was discovered in GLPI before 9.4.1. After a successful password reset by a user, it is possible to change that user's password again during the next 24 hours without any information except the associated email address.

Teclib GLPI before 9.4.1.1 is affected by a timing attack associated with a cookie.

An issue was discovered in GLPI through 9.2.1. The application is affected by XSS in the query string to front/preference.php. An attacker is able to create a malicious URL that, if opened by an authenticated user with debug privilege, will execute JavaScript code supplied by the attacker. The attacker-supplied code can perform a wide variety of actions, such as stealing the victim's session token or login credentials, performing arbitrary actions on the victim's behalf, and logging their keystrokes.

A remote code execution issue was discovered in GLPI through 9.2.1. There is a race condition that allows temporary access to an uploaded executable file that will be disallowed. The application allows an authenticated user to upload a file when he/she creates a new ticket via front/fileupload.php. This feature is protected using different types of security features like the check on the file's extension. However, the application uploads and creates a file, though this file is not allowed, and then deletes the file in the uploadFiles method in inc/glpiuploaderhandler.class.php.

SQL injection exists in front/devicesoundcard.php in GLPI before 9.1.5 via the start parameter.

front/backup.php in GLPI before 9.1.5 allows remote authenticated administrators to delete arbitrary files via a crafted file parameter.

GLPI before 9.1.5.1 has SQL Injection in the condition rule field, exploitable via front/rulesengine.test.php.

GLPI before 9.1.5.1 has SQL Injection in the $crit variable in inc/computer_softwareversion.class.php, exploitable via ajax/common.tabs.php.

GLPI before 9.1.5 allows SQL injection via an ajax/getDropdownValue.php request with an entity_restrict parameter that is not a list of integers.

GLPI before 0.85.3 allows remote authenticated users to create super-admin accounts by leveraging permissions to create a user and the _profiles_id parameter to front/user.form.php.

Unrestricted file upload in GLPI before 0.85.3 allows remote authenticated users to execute arbitrary code by adding a file with an executable extension as an attachment to a new ticket, then accessing it via a direct request to the file in files/_tmp/.