Log injection happens when untrusted data is stored in logs and does not escaped when used by the downstream component or rendered in the UI. An adversary can exploit this vulnerability to execute arbitrary command (see Log4Shell), JavaScript code (see second-order XSS or blind XSS), or insert fake log entries (see CRLF injection).

Remediation

• Restrict size and character set of the data that is stored in the logs.
• Perform contextual escaping of the logs before usage ore rendering.