Researchers found a way to trick AI assistants into missing dangerous user instructions on a website.
First seen on securityboulevard.com
Jump to article: securityboulevard.com/2026/03/researchers-found-font-rendering-trick-to-hide-malicious-commands/
