All systems fail at some point, no matter how much time and rigor are put into their design and development. AI is not immune, susceptible to attacks, exploitation and unexpected failures. This session will be broken into two presentations to explore: