Over the last five years, techniques for hiding information have become a rapidly growing research topic. Its better-known aspects range from inserting imperceptible copyright marks in digital audio and video, through auctions and elections, to de-identifying medical records for use in research. As we explore these applications, it becomes clear that the traditional protection goals of `confidentiality, availability and integrity' are insufficient. In this talk, I will discuss the role of anonymity in computer security. Anonymity mechanisms allow us to construct networks and file systems that are resistant to selective denial of service attacks, or which let their owners plausibly deny their existence (thus providing some protection against coercion). They can help us to meet protection goals such as personal privacy, location security and plausible deniability.