In our study of U.S. counterterrorism programs, we found that anthropology needs a mode of analysis that considers security as a form distinct from insecurity, in order to capture the very heterogeneity of security objects, logics and forms of action. This article first presents a genealogy for the anthropology of security, and identifies four main approaches: violence and State terror; military, militarization, and militarism; para-state securitization; and what we submit as “security analytics.” Security analytics moves away from studying security formations, (...) and how much violence or insecurity they yield, to identifying security forms of action, whether or not they are part of the nation-state. As a framework for anthropological inquiry, it is oriented toward capturing how these forms of action work and what types of security they produce. We then illustrate this approach through our fieldwork on counterterrorism in the domains of law enforcement, biomedical research and federal-state counter extremism. In each of our cases, we use security analytics to arrive at a diagnosis of the form of action. The set of conceptual distinctions that we propose as an aid to approaching empirical situations and the study of security is, on another level, a proposal for an approach to anthropology today. We do not expect that the distinctions that aid us will suffice for every situation. Rather, we submit that this work presents a set of specific insights about contemporary U.S. security, and an example of a new approach to anthropological problems. (shrink)