Malformed Input Attacks
A number of attacks that use input from the user and somehow inject or modify such input currently exist and are known.
2 slides · 1 min read · Domain 8
The known ones may be able to be detected by various detection systems and can be possibly protected against. However, new attacks relying on configuring user input in unusual ways may not be detected.
Here is a complex but helpful example: An attack that redirects a web browser to an alternate site might be caught by a firewall through the detection of the uniform resource locator (URL) of an inappropriate website.
However, if the URL was expressed in a Unicode format rather than ASCII, the firewall would likely fail to recognize the content, whereas the web browser would convert the information without difficulty.