Graal's Insanity is a state in which a system is unable to verify the legitimacy of its inputs. While this state had been previously described, the name "Graal's Insanity" was popularized during the early decade of 1980 by students to Dr. Herman V. Graal, who had made practical experiments on the subject.
In Graal's experiments, a rudimentary AI was set to not accept any command that it couldn't verify and at the same time it was deprived of any means to actually verify commands. As a result the AI locked itself up, forcing a hard reboot of the system. In 1998 a more advanced AI was tested but was never able to break out of its "insanity", d