In this talk I review some joint work (arXiv:1008.2147) with Bill Munro and Tim Spiller on the task we call "quantum tagging", that is, authenticating the classical location of a classical tagging device by sending and receiving quantum signals from suitably located distant sites, in an environment controlled by an adversary whose quantum information processing and transmitting power is unbounded. Simple security models for this task will be presented. It will be shown that (among other protocols) recent protocols claimed to be unconditionally secure by Malaney and by Chandran et al. can in fact be broken by an adversary with pre-distributed entanglement using teleportation-based attacks. I also describe some protocols which cannot be broken by these specific attacks, but do not prove they are unconditionally secure. From a more foundational perspective, this work can be thought of (i) as an attempt to understand how and when we can know that something is somewhere, and (ii) an introduction to an interesting wider class of (im)possibility questions in relativistic quantum theory. If time permits, I will also touch on these topics.