The appeal of AI, despite its shortfalls, is growing our appetite for data collection and digitization. Data is vulnerable: no matter what regulations say, or what cyber security experts try, once a bomb is built, some idiot human (not a fictitious Skynet AI), some day, will explode it intentionally or accidentally; Chekhov’s Gun will always fire by the end of the play. Three notable experts on deriving value from data will share their thoughts on whether (and, if so, how) we can draw enough positive value from data collection to justify not just the risk but the certainty of its inevitable misuse.